One of the biggest inspirations for this work was Zakia Sewell, a writer, broadcaster and DJ on a quest to discover Albion: an alternative national story and canvas rooted in myth, folklore and enchantment.

When so many people seem to be looking for meaning, and at a time when so many conversations about national identity are about who to exclude, I found it a hugely inspiring project and proposal about what it is we could all be inspired by, proud of and connected to.

Now, having written her new book, Finding Albion, it was an enormous privilege to hear more about what she discovered on her journey.

Thanks to the Centre for British Progress for supporting this podcast, to Podcast House for their production support, and to Simon Panayi for original music.

How to find Albion, with Zakia Sewell

• Zakia Sewell is a writer, broadcaster and DJ from London. She is the host of Dream Time on BBC 6 Music and the author of Finding Albion, which explores British national identity, folk culture and myth.

[Transcript lightly edited by Claude.]

So Zakia, welcome to the show. Let’s start with an easy one — what does Albion mean to you? How did you come to it?

[01:26] An easy one, he says. Albion is the ancient name for Britain, and it’s a nebulous word or concept that means lots of different things to different people. I can’t pinpoint the first time I came across it — it was probably emblazoned on a pub sign, or when I lived in Stoke Newington, around the corner from Albion Road.

It’s probably seeped into my consciousness slowly over the years rather than in one particular memorable moment. But to me, it represents this alternative spirit of the nation. In some of the early texts and histories of Britain, the island is referred to as “the isle formerly known as Albion,” so it represents this kind of time before. And yet at the same time, we don’t really know anything about what the word really means or where it came from. There are guesses about the etymology, but it’s all a bit unclear — shrouded in mystery.

I think the fact that we can’t quite pin it down has meant it’s rich and fertile for dreaming about Britain and alternative visions of the nation.

That really came through in your work. But it’s funny, because when I started writing about Albion I had this bittersweet moment where it became clear the word has some contested connotations — someone sent me a message from a person who said “oh, you’ve got to beware, this is textbook fascism.” Which I found funny, because what I was writing was basically the opposite — rather than a national identity based on who we exclude, it’s about what we should all actually be connected to. But it did make me wonder: did you have a blank slate here? How do you think about the different connotations?

[03:31] It was quite interesting through the process of researching the book — you end up in all sorts of weird corners of the internet, and I came across a book called Dark Albion.

I bought myself a copy. I really regret buying it. I’ve given £7.50 to this nasty guy, whoever wrote it. It was basically this negative fantasy about how the country has been colonised by immigrants. One of the potential roots of the word Albion is the Latin albus, meaning white — and some people see that as a reference to the white skin of the “true inhabitants” of the nation. So absolutely, people with very different views and visions of Albion to mine have been drawn to it.

However, I didn’t really feel the need to reclaim it as such, because for me the most powerful association with the word is via William Blake. Blake, the incredible polymath, for whom Albion was a kind of giant — a personification of Britain. He had this very complex mythology, written across various books and prophetic works, in which Albion represented the soul of the nation that had fallen into a deathly slumber, walking in a kind of state of death. Blake saw this as a consequence of the hyper-materialistic, empiricist, irreligious state of his age — and yet he promised that the spirit of Albion would rise again. That is my most dominant association with Albion, and it eclipses some of those darker visions.

That’s interesting, because you’ve talked about this in a way that I think many people will resonate with — rationalism has its benefits, but it’s not the only way of seeing the world, and there’s something about synthesising the rational with the woo-woo. There’s an element there that takes it beyond just the nation state. I’m curious: on your travels throughout the quest, was there a particular folk festival or custom that you were especially drawn to?

[06:12] The most powerful experience for me personally was Stonehenge. Which might also be because it was the final place I visited — the book starts at the Spring Equinox and finishes at Stonehenge on the following Spring Equinox. It was also a gorgeously sunny day.

I’ve been to Stonehenge before, but I’d never been for one of the Solstices or Equinoxes, when the barriers come down and you can get up close with the stones. What I found really special was that it was genuinely diverse — much more diverse than I expected. And I had this sense, as we were all trudging through the mud towards the stones at 5am, of disparate tribes and nations and peoples from all over coming together just to watch the sunrise. Something so simple and yet so deep and profound.

Being there, I was walking around thinking about what it all means — and in recent times the evidence has revealed that the building of Stonehenge was exactly that: stones from the northernmost point of Scotland, England, and Wales, lugged for miles to this particular site. That sense of gravity, that idea of unity and coming together in our difference — it was a really positive and hopeful note to end my personal quest, and hopefully the book, on. It was a quiet ritual. There was a bit of a Druid session happening, but really it was just that simple act of watching the sunrise.

How much do you go in for the woo-woo? I remember talking to people in my family and they’re like, “why are you talking about paganism?” — the Woo-Woo Taboo. How have you navigated that?

[09:12] The Woo-Woo Taboo — I love that, I’m going to steal that.

I would say I’m definitely interested in spirituality, and I’m quite open to mystery — quite open to not having to know it all. I like the fact that there are things we can’t know or understand. And yet at the same time, I would say I’m quite a rational, practical, pragmatic person. I’d like to think I’m a healthy balance of the two.

I know there will be people who are put off by talking about spirituality or paganism, and it’s not an attempt to convert anybody to that way of viewing the world. Within folk culture, a lot of folk traditions really aren’t woo-woo at all — it’s like blokes carrying flaming tar barrels on their backs, not necessarily to bring about a good harvest or for any fertility reason. I like that it’s a mix of the two.

I hope that anyone put off by the more woo-woo aspects of my journey for Albion will find more rational, historical arguments in the radical histories and other aspects of the book.

Because there’s the woo-woo, crystals end of the spectrum, and then there’s folk traditions that are of, by, and for the people — and there’s a quite radical politics here. This is not about the state; it’s about communities building up their own customs across the country. In some ways you’re doing a kindness by making these paths and customs more legible to people. But on the other hand, a lot of people might look at this and say: “this feels like Merrie England cosplay — middle class, National Trust, Morris dancing, cream teas.” How should we think about making it more accessible? Is that your mission?

[11:37] I don’t see myself as some sort of folk evangelist — or a Priestess of Albion, just a humble follower.

I’ve just had this strange fascination for a really long time that I thought was a very personal, quirky thing about me. It’s only in recent years that I’ve realised it has legs, that there is an appeal for people outside the folky scene — of which I was never really part.

I’m not out here to convert people to the virtues of folk, but rather to say: within folk culture we have this fertile source of alternative stories about Britain. Traditions like Morris dancing, old songs passed down through the ages, strange seasonal customs performed all around the country — these have traditionally been the expressions of disenfranchised people: the working classes, women, people of colour here in Britain and across the empire, who didn’t have access to high forms of culture or were excluded from the national imaginary. In that sense they’re a fertile source of alternative stories, alternative information that we can excavate and use.

That’s particularly helpful when we’re thinking about alternative visions to balance out the exclusionary, racist visions of Britishness that are extremely dominant at the moment. So my mission is more: here is a rich seam of our culture that’s been buried. And yes, that could be slightly twee May Day celebrations in Oxford, but it could also be traditions practised across the empire that are fusions of British and West African customs — which tell a very different kind of story about our history and heritage.

I think it’s those radical and subversive histories that I’m personally much more drawn to.

There’s definitely something I want to come back to on new folk customs, but just to follow up on that point about contested national identity. I think a lot of people look back at some of the awful things that have happened in British history and don’t want any association with them. It has complicated the ability to feel national pride for many people. How did you explore this through your quest?

[15:14] This book is really an expression of my very specific identity. My dad is White, English and Welsh. My mum is Black Caribbean, born here.

Growing up between those two cultures has at times been quite complicated, because in many ways I feel very at home in Britain — it’s where I was born, it’s where I’ve always lived, and there are lots of things I love about this country. When I go to the Caribbean, I feel more British than ever. And yet, as a person of colour, as someone with very direct links to the devastating impacts of colonialism — not just on the Caribbean island my family are from, but within interpersonal relationships in my own family — it’s quite a difficult place to be. How can I reconcile these two aspects of my identity?

So the idea of pride — I still don’t even know if that’s something I aspire to. More, a sense of peace about who I am and where I live. That’s where I’ve arrived through writing this book.

Something I’m grappling with within it is: how can we name and be honest about those really dark aspects of our past and present? We only need to look at the horrific behaviour of the far right in the last year or so to know this is a very much living, breathing problem. How can we balance that — come up with practical ways collectively to address these dark aspects of our history, our heritage, and our present — without letting go completely of hope? Can we find a national story that can hold the dark and the light? That is the question. I don’t know. I hope so, but I suppose my book is an attempt to start doing that work.

There’s something to learn from how Germany has engaged with its darkest history — it allows you to decouple and say: here are the awful moments, and also the moments of real invention and ambition and greatness that have nothing to do with the bad. And when I look at Britain now and see a country that’s stagnating, or feels like a bit of a museum — parts of the country just not doing as well as they might — there’s a tension there. I want us to aspire to greatness again, but not the kind that manifested so badly before. Can those two things be reconciled?

[19:22] I think firstly it’s important to analyse what we mean by greatness. The greatness I’m critical of in the book is the greatness of empire — of feats of engineering, of military, of monarchy. A self-aggrandising, pompous greatness. It’s about superiority. And that is an aspect of our national story and identity that I and many others find very difficult to contend with, especially when my ancestors were enslaved by the British. I know what that greatness was built on. I know the dark underbelly of it.

That’s not to diminish the incredible creative achievements and works of British people throughout the ages — it’s just that story’s already had a lot of airtime. We’ve got Shakespeare, for god’s sake. He’s not about to be toppled from his plinth. My hope is just to bring things into balance a little bit.

It was also an invitation to ask people: what do you actually think is great about Britain, if we’re rejecting those inflated visions of Britishness latched onto by the far right, by Nigel Farage, Boris Johnson and so on? What are the aspects of Britishness we’re actually proud of, and that we want to integrate into a more balanced, healthy, and nuanced national story? The darkness doesn’t have to eclipse the light — but nor should the light totally conceal the darkness.

That’s a good way of looking at it. There’s also an angle here about class — this is a story of national history, but it’s also one of class struggle. You write really interestingly about the Right to Roam campaign as a modern manifestation of that. And maybe it’s a bit of a stretch, but the one permanent thing throughout time has been the land. There’s something in rituals that connect you to land over time, and something in Right to Roam that feels like a modern enfranchisement — because the enclosures severed a connection to land that so many people had. How do you see the next phase of rebuilding our connection to the land and our national story?

[23:17] In the chapter where I was talking to Nadia Shaikh, one of the Right to Roam campaigners, it was really my first significant exposure to Britain’s radical history. Right to Roam are a modern incarnation of the people throughout the ages who have fought to make Britain a fairer place — from the Tolpuddle Martyrs to the Chartists, Kett’s Rebellion, the Diggers. To me, this is such an important aspect of our history and culture that should be taught at school. It could be a really inspiring way to think about who we are and who our folk heroes are. Why are they not on plinths? Because the powers that be don’t want us to rebel.

One of the things that really stuck with me from my conversation with Nadia was this relationship between our disconnection from the land, enclosure, and the erosion of local identities. The acts of enclosure that shoved people off the land and into wage labour in the cities eroded those long-term relationships that distinct local communities had with their landscapes, and the traditions and dialects that went along with them.

And the kind of blanket Britishness — the contemporary visions of national identity that we have today — came in to replace those more rooted, localised identities that were really about specific communities and their histories. I suppose one of the radical possibilities of folk culture that I see is this: if we came up with new traditions, new ways of rooting stories about who we are in our local, specific landscapes, reinvigorating that local sense of community identity — then perhaps people wouldn’t need to cling to vast, unwieldy ideas of Britishness. You could be British, but with all the feeling and emotion located in your local identity. I think a lot of people feel that way already. But it’s something that could be nurtured a lot more.

That’s really interesting. I find it easier to describe myself as British than English — and the more I’ve understood some of these alternative customs and practices, across England and across Britain including Wales, Scotland, and Northern Ireland, the more I’ve been able to extract from that a kind of emergent national connection. But it’s interesting that you can go one of two ways: you can go bigger, or you can go much more local. Which brings me to new folk customs — you write so interestingly about Notting Hill Carnival and others. How do you think about what counts as folk culture today? Is UK Garage folk? It could only have existed here.

[28:58] A lot of people talk about grime as a kind of folk expression. I’ve thought a lot about it. I think one of the problems with thinking about grime as a whole as folk culture is its relationship to the music industry — because usually when we’re talking about folk culture, we’re talking about things that have existed outside of the academy, the institutions, or commercial enterprise.

So maybe those early grime freestyles — I’m thinking of a specific stairwell, D Double E doing a freestyle back before these artists were releasing on labels. It was absolutely DIY: little camcorders uploaded to YouTube, people speaking about their experience in Bow and across the city. That could count. But there’s a crossover point where those things become part of the music industry and are perhaps no longer folk.

Do you think there’s a tension between folk culture becoming commodified or trendy, and still preserving the meaning while being able to scale it to a new generation?

[30:53] There’s been a lot of talk about the contemporary folk revival, lots of newspaper pieces — “Oh, Britain’s new cool folk wave.” There definitely has been a resurgence, to the point where Morris dancing was on the Brit Awards, which I’m sure few people had on their cultural predictions list.

One of the interesting things about this contemporary folk wave is that a lot of it is happening online, and it is largely a revival of the aesthetics of folk in particular. Because it’s happening on Instagram and TikTok, it’s the traditions that look a certain way that are getting more views and likes.

I spoke to a die-hard folky in the book, a lady called Chloe Middleton-Metcalfe, who grew up going to folk festivals, is a Morris dancer, and is raising folky children. She’s not on social media, so she didn’t even realise there was a folk revival happening. We had this conversation about: yes, you can see the hard data for a revival in social media followers, but what about membership to folk clubs?

I think one of the wonderful things about this revival — even aided by journalists and brands being interested in it, and Simone Rocha using pagan imagery on the catwalk — is that so many people from different walks of life are being exposed to these traditions and becoming curious about them. It’s so much more diverse than folk revivals of the past, not just because of the demographic shift in Britain but because of the visibility. You don’t have to get up at 5am and turn up to some town in the Midlands — you can watch it in bed on your phone.

Just recently, the first Black British folk club has been set up. I don’t know if that would have happened without social media and this slightly more sheeny, trendy aspect of the revival. Blessings and curses. I think the people on the ground who’ve been preserving these traditions out of love will carry on doing what they’re doing. And I hope that even once the journalists have lost interest and the brands have moved on, there will still be a more diverse, wider range of people who’ve got the folk bug — and who will continue being interested, preserving things, and creating new traditions in times to come.

There’s something more fundamental and meaningful about where these customs come from and what they represent. There’s been a decline of religion, people still feeling something’s missing — and it feels like there’s a gap that folk could and should fill. I’m interested in how you think about why there’s been this revival and how much deeper it is than just an aesthetic.

[35:22] I think there are a few layers to it. Church attendance is at its lowest ever — but that doesn’t take away people’s yearning for a deeper sense of meaning. Why are we floating around on a rock in space? I don’t have the answer, but these questions are fundamental and stay with us. So people are turning to alternative sources for meaning: alternative forms of spirituality, astrology — maybe that’s going too far.

I’m drawing the line at astrology.

[35:53] It’s lighthearted, but it speaks to a genuine curiosity. Also, we’ve lived through the peak of Enlightenment thinking and the materialist, empiricist hopes and dreams for humanity — and we’re seeing that project is somehow incomplete. There are many ways of understanding the world, ways of being, that were eclipsed by the Enlightenment’s bright light, that people are now trying to retrieve from the shadows.

Then there’s the environmental dimension, which is huge. We’re faced with climate catastrophe. So the idea of honouring the landscape with traditions, getting out into a field at six in the morning and doing a funny ritual with friends — whether you believe there’s any great spiritual meaning to it or not — it’s about a connection to the landscape. Following the pagan wheel of the year, or even just noticing the seasonal shifts, the Equinoxes, the Solstices, May Day. This is about becoming more aware of the natural world and its cycles. For me personally, that’s been a massive re-connector.

And then there’s everything we’ve discussed about Britishness and the identity crisis we’re going through in this country. Folk culture seems to offer a different flavour of Britishness — one that’s eccentric and strange and weird, but a bit radical and aligned with people’s politics, and also mischievous and silly. I think people are being drawn to folk as an antidote to some of those more toxic visions.

Folk in its broadest sense is just the culture of the people — people getting together and singing or dancing or coming up with a community event that they practise again and again. That is something fundamental that can look completely different according to where you are in the world, or who you are. We need to de-twee folk a little bit, to make it more appealing to the masses.

There’s also something in the choice to believe — or the choice to sit with uncertainty. You can try on some practice without really knowing whether you believe in it, and it can still connect you to something. I went wild camping in Dartmoor and walked through a forest where all the trees were swaying in a particular way, and I thought: are there spirits here? There might be spirits here. I choose to believe the Dartmoor spirits were talking to me. Beyond this individual level, there’s also a decline of communities of place — and you write so nicely about how communities of place can fill a hole that communities of interest, no matter how valuable, can’t always fill. You’ve gone on this journey, explored the Wheel of the Year, travelled the country, spoken to so many different people. When you ask whether a vital collective reimagining of our folk customs and festivities could help revive the communal spirit that’s been lost in Britain — what did you conclude?

[41:19] I have a dream. I’ve been reading about the Festival of Britain after the war — this great collective, nationwide series of events all about bringing community together. Obviously it had a particular flavour; it was of its time. But I thought: what would it look like to have a Festival of Albion? One where it was just local people, in their town halls or wherever, coming together and asking: what does our community need? What event could we put on?

And if that’s in Rochdale, it’s going to look very different to if it’s in Stroud. But actually it would be a way of bringing the contemporary, current communities of Britain together — whether you’ve got a large Muslim population, elderly locals who’ve been there for generations, and new young families who’ve moved in from London or wherever. Everyone coming together and bringing their unique perspectives, perhaps bringing their own traditions and customs, and collectively coming up with new traditions that could bind the community. That’s my utopian dream.

This is ultimately what these folk customs and traditions have always been about: bringing people together. Usually it’s the depths of winter, or living somewhere really gruelling and miserable, and just coming up with something fun for your community — not to make money, not because there’s a rational basis for it, but because it makes you feel good to commune with your neighbours. Even if you’re not a great singer or artist or sculptor, to use your creative faculties to conjure something fun and beautiful for the people around you.

When I went to Cornwall and got involved in the Montol parades in Penzance, I’d thought I might experience some deep ritualistic spiritual feeling while processing through the streets. But actually that’s when I really understood what these traditions are about. It’s just about people — connection to people, connection to place, and the sense of passing something on to those who will come next.

I would love to see the Festival of Albion ripple across the country, with people recognising that yes, our customs and traditions and songs and stories might be different in flavour, but ultimately they are an expression of that human desire to tell stories about who we are — and that is something that crosses faiths, cultures, continents. It unites us, coming together in our difference.

Obviously this is a utopia under a Reform government, so it’s quite unlikely — but that’s my dream.

I want to make the Festival of Albion happen. Any sponsors listening, get the chequebook out. Thanks so much for coming on, Zakia. That was great.

[45:05] You’re very welcome. Thank you.

What if it’s more like the brain, a heterogeneous architecture with varied regions capabilities and properties that enables incredible intelligence and incredible efficiency?

And what if this holds the key to redistributing leverage globally away from the biggest labs and countries to rebuild optionality across the rest of the world with their energy constraints, with their compute constraints?

That’s the thesis behind today’s guest. Dan Akarca is the co-founder and CEO of Callosum, a new AI infrastructure company that’s just raised over $10 million to build what they call the orchestration layer for heterogeneous compute. That means they’re making AI systems work across a range of different chip architectures, models, and other hardware, rather than simply depending on a single monolithic stack.

Dan and his co-founder Jascha did computational neuroscience PhDs at Cambridge, studying how the brain builds these exquisite, efficient, intelligent systems, and they’ve taken those principles and they’ve applied them to arguably one of the most important infrastructure problems of the next decade.

If they’re right, the implications go well beyond efficiency. In a world where artificial intelligence is quickly becoming a critical input to production, akin to energy, breaking our dependency on today’s dominant paradigm -- where value accrues to the biggest labs and the biggest countries -- could be hugely consequential, both for Callosum to build a deeply ambitious global company themselves and to redistribute leverage globally.

How to break the AI hardware monopoly, with Dan Akarca (CEO, Callosum)

• Dan Akarca is co-founder, with Jascha Achterberg, of Callosum, a new AI infrastructure company unlocking a world of heterogeneous compute. They just raised $10.25m led by Plural.

[Transcript lightly edited by Claude.]

Daniel, welcome to the show.

Thank you for having me.

How good an analogy is the brain for AI?

The earliest development in AI was inspired by the brain — all the way from how we analogise neural networks. This was deeply inspired by the weights of synapses, going back to the forties and fifties.

When people were thinking about some of the most interesting questions impacting our society — building intelligent systems — they believed the brain was the perfect existence proof of intelligence, and that we should effectively emulate it.

At the same time, we’ve now realised that’s not quite true. I’d describe it as a Venn diagram: there are shared principles between artificial and biological intelligence, and there are things that are totally different. The obvious one being that the hardware we run AI systems on is quite different from the messy brain.

The thing that I think is most deeply overlapping is a principle about where efficiency comes from in intelligent systems. The brain evolved subject to selection pressures that AI hasn’t been — energy constraints. One of my core theses is: what does putting an energy constraint on intelligent systems lead to? If you put energy constraints on, you get the brain, or something like it. If you don’t, you get massive data centres of huge stacks of homogenous GPUs.

The biggest difference is that the brain is not a monolithic, homogenous structure — it’s very varied at multiple scales of analysis. I believe that heterogeneity in our computing systems is also where we’ll find all the efficiencies we can gain beyond single-substrate digital silicon.

This sits within a wider trend about how we build AI systems, and it’s what led me and my co-founder Jascha — having done our PhDs at Cambridge thinking about these questions for many years — to build Callosum. Callosum is at its heart an AI infrastructure company that makes AI systems work well on a mixture of chips. Rather than building AI systems to run on single types of chip technologies, our vision is that the AI systems of the future will be best served by a wide range, a plurality of different chip technologies with different trade-offs.

Around 2022 and 2023, the biggest trend in AI was scaling up single massive models. I describe this as the big, bad super-god model that would effectively take over every task society would ever need. What we’ve basically realised over the last few years is that this is just wrong. The world will not be best served by a single model.

If you think about the hardest problems in the real world that we actually need to solve, they have features that are not amenable to a single model. Every real-world problem — which is why our brains evolved heterogeneity — is inherently multi-turn, uncertain, and specialised in nature. The problems we solve won’t be done by a single model; they’ll be done by many different models of different sizes, different specialisations, built in many different countries.

You mean big players are talking their book?

Yeah, exactly. And what we’re seeing is that it’s just wrong. The world will not be best served by a single model.

So serving all those different use cases with a homogenous chip stack creates an enormous amount of inefficiency.

Exactly. If the models of the future will look like multi-agent systems of different models, then the compute stack underlying those workloads will also be heterogeneous. It’s not necessarily true that single chip architectures will serve all of those cases perfectly. You naturally get what I call a disaggregated inference space — the space of how we serve AI models will split, many new and more specialised chip technologies will emerge, and in that world you have trade-offs.

I want to back up a second. You’ve described a spectrum of heterogeneous computing — from datacentres on one hand to edge computing and on-device inference on the other. Can you paint a picture of what’s in between?

On that spectrum from the data centre all the way to edge devices — your phones and so on — things become increasingly heterogeneous. Data centres right now are overwhelmingly homogenous. They’re only heterogeneous for market reasons, like when you have to get a new order of chips in and might have some old ones lying around. They’re inherently built to be homogenous, particularly for training AI models.

It’s the same paradigm, just different generations of chips.

Exactly — same paradigm of GPU acceleration. As you go further towards the edge, things become inherently more heterogeneous because there are much more specialised use cases, and more energy constraints.

So the market evolves to have more heterogeneous computing towards the edge because it’s more energy-sensitive. In between, we’re still working it out. There are people betting that edge compute is the future of AI and we won’t need data centres — but that doesn’t seem to be the case right now. Most people want the best models, and they’re willing to wait for a response that comes from a data centre rather than run something on their edge device.

Our thesis is that it’s a one-way track — it will become inherently more heterogeneous across the entirety of that spectrum, for both technical and market reasons.

It’s funny — with my Claude subscription, if I’m using a frontier model for everything, there’s a kind of intellectual security in using the best intelligence available to me, even if I’m using it for incredibly basic things half the time. What are the drivers for people to not always just go for the most performant model?

You’re right that today people gravitate towards frontier models, for the simple reason that if capability is going up really quickly, you don’t want to be left behind. So people just want to make sure they’re running on frontier models, and they’re willing to pay a cost for that.

The question then becomes: when does single model capability reach diminishing returns? When do you realise that all the frontier models are basically the same, and start switching? We’re already seeing that in enterprise.

I’d say there are four big measures on the demand side: performance, cost, speed, and sovereignty. At the moment, people are highly sensitive to performance. But model switching velocity is actually quite high already — people will switch if they hit rate limits on one platform, for example.

The question I have is: for really hard problems — what enterprises, scientific labs, and robotics companies want to solve — when you reach a performance plateau, to what extent will you find that your problems are inherently heterogeneous? That you need multiple different models, some highly optimised for speed, some for cost? Most enterprise use cases I see are still performance sensitive, but cost sensitivity is coming.

If you have an API call to one of the frontier labs, you don’t really have sovereignty over the cost or the models. Those four measures are what people will switch on. At the moment people aren’t switching because of performance, but it’s just a matter of time before they hit diminishing returns.

There are founders who would say similar things about those four variables, and then answer: “so I want to build a new chip.” You buy the thesis about the brain, and about the enormous efficiencies in co-designing hardware and software — so why aren’t you doing that?

I’d say that’s what distinguishes us from most people in this space. The problems we need to solve in the real world won’t be solved by any single specialised chip, no matter how good it is.

If you really want to solve real world problems that are fast, capable, efficient, and sovereign, it won’t be done by a single chip technology. The story of chip evolution has been from general chips to specialised chips — if you build a new chip, it will inherently be specialised. That’s the only way to beat the current incumbents. And the question becomes: where in the space of specialisation should I build?

What we’re saying is that if heterogeneous computing is the better solution — and we’ve released work showing that’s the case — then you won’t build a single chip. You’ll work on how you orchestrate across different chips.

What does orchestration across different chips look like practically?

It means different models running on different chips, working collaboratively and communicating between each other to do productive work. It could be single models on single chips, multiple models on a single chip, or larger models spread across many different chips — and those chips could be within a single vendor or between vendors, co-located in a single data centre or operating across different clouds.

Really, orchestration is what I’d call a workflow problem — how you define your problem — and an agent problem — what models you utilise — and a hardware problem at the lower level. Crucially, every layer has to be aware of every other layer in its optimisation. The gains come from the synergy of those layers, not from separating them out.

There’s a deeper reason for this rooted in the foundations of computer science. Geoffrey Hinton described what he called the “immortal computer” — the idea that software and hardware should be inherently separated. You can build your chip without thinking about the software, and build your software without thinking about the chip. The entire computing and AI stack today is built this way — horizontal and independent.

That works well for certain applications. But when you want to solve real world problems with intelligence, you need to make those layers much blurrier. You need to co-evolve them in context of each other — building a vertically integrated stack that can do cache management in context of your workflow, and build kernels in context of the constraints of the problem. Everything that will ever want from AI — cost, performance, speed, solving the world’s hardest problems — has this feature. We estimate something like five to ten percent of AI problems in the future could be solved by a single model on a single chip. Practically everything you care about will require something more.

Walk me through the product experience. It’s akin to a two-sided marketplace — you have a variation of chips and models on the supply side, and customers with different use cases on the demand side. How do you break that down?

On the supply side, the question is: what chips are available today in 2026, and what will become available over the next five to twenty years?

I want to mention something called the hardware lottery, which is the idea that because building a chip costs so much money and takes so long, you build it and it’s a lottery whether it’ll actually be used. Nvidia won because they had a great use case in gaming — their chip was ready when AI took off. Because of the capital required and the timelines, this has actually been stagnating the diversity we could have on the supply side.

What we do at Callosum is everything we can to help new entrants in the inference chip market commercialise their hardware effectively. They’re entering a diverse market — not necessarily winner-takes-all — and they need to be able to understand their trade-offs in context of others. We work very closely with supply side customers who are building amazing inference chip technologies. We also work with networking companies building entirely new ways to connect chips together using light. So the supply side is becoming inherently more heterogeneous, and you need a company with a wider scope that can bring them to market.

What does “bringing them to market” mean?

Depending on their technical readiness, it ranges from running simulations all the way to testing on real-world workflows. Chip companies typically define their offering in terms of hardware metrics — clock speeds, FLOPs, energy efficiency. What we provide them is real-world AI-level metrics.

The reason Jensen Huang was so good is that he didn’t sell Nvidia chips on hardware metrics. He sold revenue — he solved real-world problems for people. What we’re aiming to do for these new inference chips is provide those metrics: here’s how your technology fits in terms of the longer-term market.

There’s been analysis recently about TSMC’s under-investment in the early 2020s — now a major factor in the chip supply crunch. Fundamentally it was TSMC shifting risk onto model companies, who then couldn’t move as quickly as they’d have liked. And over that period, the big AI companies made a lot of money, but not as much as they perhaps could have done had everyone else scaled up alongside them. Is there something here about how, in a different paradigm, there are a whole set of players who can make revenue in a way that is less dependent on those constraints?

Yeah, exactly. People have their core technologies and they need to de-risk themselves by offloading certain things onto someone else — that’s ultimately the story of hardware supply chains. ASML and TSMC, despite where they sit in the stack, don’t necessarily accrue all the value because they’ve offloaded certain risks.

What we’re saying is there’s a paradigm shift happening in how hardware will be sold. People are raising capital to build their technologies on the premise that it’s a winner-takes-all market — but it likely won’t be. And that’s actually not a bad thing. It means the ecosystem becomes more healthy.

So almost like you have new chip companies saying “there’s this massive Nvidia monopoly, we’re going to take down the giant” — and maybe each of them eats off a different bit of the business?

Yes, exactly. And there’s a lot more to it than that. When we usher in photonic networking, for example, the way we design the software to run these chips also changes dramatically. There are a huge number of interoperability questions that will arise.

And it’s not just bringing them to market — we also do work testing their chips on real-world workflows. What’s really valuable about our position is that we know exactly what people are running in the real world, and we know a huge amount about the priority of different chips entering that space. So we can go to our supply side partners and say: here’s what workloads are being run right now, here’s what people are paying for, here’s our estimation of where the needle needs to move economically to make new use cases possible.

There’s an analogy with the energy system that might be useful here. We’ve talked about how important energy constraints are to intelligence, but there’s also a whole shift happening in the energy system itself — from turning coal power stations on and off to meet fixed demand, to variable generation sources and flexible shiftable demand. That’s a very different model from what came before. I see an analogy between homogenous data centre architecture and big AI labs on one hand, and big centralised power stations on the other — versus a much more distributed system of intelligence.

Exactly. Why would you want something to be more static or more dynamic? It all depends on the types of problems you’re solving. If you have relatively static infrastructure doing a role that’s invariant to the problem — super predictable — then static is fine.

But if you’re running AI models to automate complex work, that’s a hugely dynamic process. Think about all the complex work you do day-to-day, let alone when we move into robotics. As a result, our infrastructure needs to be too.

There’ll be some things that are slow to update — the grid is not easy to update. But there are things we’re building right now at the software level that allow us to make data centres into living, breathing organisms.

I use the analogy of Geoffrey Hinton’s “immortal computers” — where hardware and software don’t talk to each other — versus “mortal computers,” where they’re inherently interlinked. The brain is at the mortal end, tightly coupled to its energy supply through blood flow. What we’re doing is slowly instantiating principles from mortal computers into the immortal, digital ones. What Callosum is doing is accelerating that journey — utilising mortal principles of computing in immortal digital computers. I think that’s going to be a hundred-year trend.

On that note — you’ve just raised a round. What does the next phase of this business look like?

We raised $10.25 million in a pre-seed led by Plural, who have been amazing and are looking to back the most ambitious technological unlocks possible.

Right now, we’re doing everything we can in software — all the way from the AI systems layer down to the compile level — making heterogeneity work as well as possible: faster, better, stronger, and as easy as possible for new entrants.

Our north-star ambition is to redefine not only how AI systems will serve people — through heterogeneous, multi-agentic systems, in robotics, in scientific simulations — but to redefine the infrastructure needed to enable that properly. Over the next phase of our company, we want to not only provide that software, but to really reimagine how the compute stack will look from first principles. Will data centres become more distributed? Will they become inherently more heterogeneous? I think so. We have a unique opportunity where what we need compute to do has massively changed, and as a result we don’t need to just use the cloud model of compute infrastructure anymore.

To draw out two things from that: first, you’re turning homogenous data centres into heterogeneous ones. And second — speaking to the sovereignty point — in a world where countries feel dependent on streaming intelligence from abroad, if intelligence becomes an input to production the way energy is, and overnight becomes twenty percent more expensive because of some tariffs or policy change, that could cause an enormous economic shock. Is there a version of what you’re describing where we can rewire the economy around intelligence without having those choke points in quite the same way?

Yes. Where we are today — particularly in the UK and Europe — is that we are hyper-reliant on the hyperscalers for the majority of our infrastructure throughout society. We’re starting behind, significantly.

But there is a possibility that being far behind becomes a saving grace. It’s actually quite useful to be behind if what others have been doing turns out to be wrong. If there’s a compute overhang — so much capital sunk into the old paradigm — there’s a future where we can redefine how we build these data centres from scratch.

And inherently, we don’t necessarily need the cloud and hyperscaler model at all. It’s very possible that many different chip companies build their own clouds, distributed across the world in different ways. Every country is building their own chip. Every cloud is building their own chip. Many companies are building their own chips. So long as there is investment in new chip technologies, there will always be trade-offs, and always a huge surface area for optimisation across different points. You know, there’s a deeper argument about how this fits into the increasing multipolar world we’re entering, and the race for sovereign capabilities — but we don’t need to simply copy others.

We’ve talked a lot about the enormous ambition here. What are your riskiest assumptions?

The whole thing falls down if my thesis is wrong — that heterogeneous systems of intelligence are just better than homogenous ones. We’ve done lots of work showing that homogenous systems today are just a special case of a wider heterogeneous system. It’s harder to build heterogeneous systems — we’re not building this company because it’s easy — but it is strictly better. Our conviction is that orders of magnitude of improvements in cost, speed, and performance for real-world problems will come from system-level architecture, not from improving a single model.

Then there are market risks. The first is whether there will genuinely be a diversity of compute available to us. One reason we have homogenous compute systems today is not because they were better — it’s because they were practical. We didn’t have different alternatives. We now do. When Jascha and I started this company and spoke to people, they thought we were too early. We’re certainly not early now — inference chips are coming to market and becoming performant.

The biggest risk is ultimately timing. There is a period — as there has been in many paradigm shifts — where two trajectories can emerge. The biggest risk is that we don’t capitalise on that opening. Which is why we’re operating so fast. The other risk is of course that our technical thesis is wrong and we will be ruled by an AI overlord after all — but I’m very, very confident that won’t be the case.

What is it that allows you to see this world in a way others haven’t?

If it’s true that all the gains of AI will come from the co-evolution of AI systems with their underlying hardware, then we have a problem and an opportunity. The problem is that the culture of people who work in this space is totally different on each side — hardware people and software people have entirely separate timescales, technical vocabularies, and focuses. That’s a problem. But it’s also a huge opportunity, because when you find talented people who can speak the language across the entire stack, they’re incredibly undervalued in the market.

The ideal in our company is that someone can describe how a legal case was solved in context of the clock cycles of a chip — meaning: what were the AI agents saying, what documents were they reading, and what were the contributions of the underlying hardware to enabling that? That kind of end-to-end thinking is what we’re going for. At the moment, companies in this space are very horizontal and separate — they stack on top of each other. We’re verticalising the entire stack.

Jascha and I came into this from a deeply multidisciplinary angle. We asked the foundational questions of how you build efficient computing systems of intelligence — and the logical conclusion is that it has to be integrated across the stack. When I speak to specialists across the whole stack, they realise what we’re saying is right. But they wouldn’t have got there themselves without the angle we came in on.

One of the things I find really exciting about being in London and Europe is the huge amount of undervalued young talent that hasn’t really been leveraged in this area. I took a lot of inspiration from DeepMind in the early days — Demis identified undervalued AI talent across Europe and concentrated it. We’re trying to do something similar with Callosum for hardware-software co-evolution. These are entirely new ways of engineering systems. There’s a lot ahead of us, but it’s incredibly exciting.

What does it mean in practice? Semiconductors have become a strategic chokepoint, but what does that actually mean when the Netherlands, which is home to ASML, the monopoly provider of the EUV machines that are critical to fabricating chips, cannot really even leverage this position. What does “sovereignty” look like when it’s interdependency all the way down?

Elsewhere, while the US AI capex buildout continues, is that the only strategy available or are there other approaches that suit Europe better, especially given our energy and compute constraints?

And then there’s state capacity, which is critical to industrial policy going well. What will it take for SovAI, the government’s new strategic AI venture fund, to succeed?

To explore these questions, I had a lot of fun catching up with my friend Lawrence Lundy-Bryan, who has just launched Europe’s first dedicated semiconductor fund, Cloudberry. He’s spent his career going deep on frontier technologies and has thought more than almost anyone I know about the interplay of AI progress, structural shifts in venture capital, and UK sovereignty. Lemme know what you think.

How to invest in AI sovereignty, with Lawrence Lundy-Bryan

• Lawrence Lundy-Bryan is GP at Cloudberry, Europe’s first dedicated semiconductor fund. He also writes the State of the Future substack, which researches frontier technology analysis.

You just launched Europe’s first semiconductor VC fund. Why on earth would you do that?

[00:01:37] Lawrence Lundy-Bryan: The truth is I’ve jumped on the bandwagon here. Like Veera and Rene, my partners, did all the hard work. there’s a VC bet that I’m making, and then there’s a macro bet.

The VC bet is, there’s only gonna be two types of VC funds in a decade:

the capital agglomerators, the AUM, the Andreessen’s $15bn four, five funds or whatever. Sequoia, Benchmark, those guys that just raise all the money.And then specialists,

I’m not sure exactly the extent to which I think deep tech was a sort of the first wave, the first attempt to move away from just SaaS to understanding infrastructure atoms and deeper, technologies.

But even then, that’s still quite broad.

Yeah.

[00:02:21] Lawrence Lundy-Bryan: I mean what is, deep tech? It’s a meaningless term. you see some funds that just invest in quantum computing. Or funds that just invest in photonics or, nuclear fusion, I think the world, will be made up of lots of those micro funds, that just know one technology and market really deeply.

Number one, because you can actually deeply understand something to the same level as a founder.

Slightly lower than a founder.

We’ll come back to that.

[00:02:45] Lawrence Lundy-Bryan: You at least can pretend.

And so you understand something, but then also you can diligence something really well with your network.

So you can actually make good investments. And second, you can “add value” in a way that isn’t just air quotes because your LPs, or your former colleagues, or your previous investments are all part of that same ecosystem.

Yeah.

[00:03:06] Lawrence Lundy-Bryan: So you really can say, once we’ve invested, we’ll introduce you to, in our case, Global Foundries to help you tape out your chip.

Or in other cases, a hyperscaler CTO as a purchaser. my view is on the VC side, specialism is the place to be. And on the macro side, semiconductors is massively over, sorry, massively under invested in.

Over invested?

[00:03:27] Lawrence Lundy-Bryan: Was over... I mean NVIDIA’s price, potentially, but that’s irrelevant.

just from a VC perspective, it’s like a huge gap in semiconductors. And I think over the next decade we’ll start to see, a way in which we haven’t seen in the past 20 years, chips be made much more specialist.

broadly speaking, there’s only two types of chips that exist CPUs, for a very long time, GPUs, and really there’s very few other types of chips that exist, at the margins. the bet over the next decade and beyond is that we’re gonna have an explosion of new types of chips. that creates good investment opportunities.

So I definitely wanna come back to the smiling curve that you talked about — in terms of the agglomerators versus the small specialist funds — because I think it’s relevant to wider, like industrial strategy and how you think about the partnerships that you might want in government and between sort of capital allocators. But before that, let’s just go deeper on the semiconductors thesis. Like, you’ve gone way deeper into this than so many people. If you are a lay person, you’ve seen like NVIDIA explode over the last few years and you go, oh, I guess like GPUs are important. Now you know what GPUs are. You previously just had your CPU in your laptop, that’s it.

What’s missing from that story? People may know Intel and NVIDIA, but there’s so much more of the story here? Just talk us through the story of semiconductors, like where we’ve been, how we got here, and then why you think that this is interesting going forwards.

[00:04:53] Lawrence Lundy-Bryan: The truth is Silicon Valley, semiconductors is one of the first, examples of venture capital. It’s surprising to me that it’s flown under the radar, because I think the truth is everybody was looking at software and the internet. most people that you speak to, apart from veterans, really grew up in the internet and software age.

So most mental models around financing, around innovation, around what a foundry even looks like are based around the social network and Facebook and that sort of stuff.

It’s not so much, it’s not been important. It’s just been under the media radar or the average person’s radar.

But it’s always been fundamental. I mean smartphones, right? ARM, people might not have heard of Qualcomm, but people have probably heard of ARM and how you design chips. Then there’s this $500 billion industry, maybe larger now, probably a trillion by 2030 It’s this huge industry that is global, one of the first globalised industries, hugely interconnected, but flies under the radar, right?That makes it a good investment opportunity, firstly. But the reason to say that is because it never really mattered to the average person what Facebook ran on. No one needed to know that Facebook ran on, whatever it was. And like how Google was able to serve trillions of queries. No one really knew or cared. And I think it’s reached into public consciousness for two reasons.

One is, COVID and the automotive example, people might have heard of that. It’s oh, we can’t make cars because, oh, there’s this one component that we can’t get access to. So it made it into public, consciousness. And then people realized how global the supply chain was, how political it was because of Taiwan. And I think there was the book Chip Wars.

Yeah, the Chris Miller book..

[00:06:37] Lawrence Lundy-Bryan: Yeah, exactly. So people started to read that. I think people realised it’s a geopolitical thing, number one.And then number two, before the NVIDIA runup, this data centre build out and people starting to realise that AI needs a lot of computing power. When you say, why does it need so much computing power? Then you say, GPU, and then you say, what are these compute chips?

The reason why I think it’s really interesting now is we’re still really at the beginning. Regardless of if you think it’s a bubble or not. We have 5, 7, 8 year build outs of data centres that are being bought and paid for that involve buying chips. this isn’t slowing down or stopping if you believe, in, the importance or the value of ai. And I do. I think all of it needs to run on computer chips. I think it becomes one of the most important components, most important technologies for any company or country in the next decade.

And play that forward. What are the bottlenecks that you foresee existing chips and kind of chip architectures not being sufficient for?It’s funny. First principles is never the best way to think about how things are adopted.

It’s mainly path dependency, but the truth is, we are using the same thing that ran the SEGA Dreamcast to run large language models. It’s just these GPUs, right? It’s remarkable, frankly. And that’s not quite true.

What’s, wrong with that?

[00:08:05] Lawrence Lundy-Bryan: Yeah, exactly. Well, path dependency, and it’s the same with lithium ion batteries. We can come up with a quote unquote “better chemistry”.

But you can never beat the cost and the economies of scale. So as an engineer, you can always point holes at something and say, I can make that better because of whatever reasons. But it is, an observation to say we are running these huge models on ultimately something that was designed for something else.

it’s not quite true because they’re not really GPUs now, they have little accelerators that accelerate certain parts of the operation — like matrix multiplications really fast. so they are bespoke to the tasks of running AI training and AI inference.

But there’s two things that have changed that mean that we probably will start to see the world that I described, I could call it heterogeneous computing: the idea that, we’re gonna have lots and lots of specialised things out in the world.And that’s for two reasons. One is that we just can’t make enough GPUs, like we had actually can’t make enough and

Taiwan just not working hard enough.

[00:09:10] Lawrence Lundy-Bryan: the Taiwanese don’t work hard enough, clearly.

what’s funny about that supply chain, it’s not really the GPU — I wrote about this a while ago — it was never, GPUs aren’t really the problem we can make enough GPUs. It’s the memory.

Memory.

[00:09:22] Lawrence Lundy-Bryan: it’s actually the high bandwidth memory that you need. which are only made by three companies globally.

it’s funny when you look at bottlenecks. You always find a good example is, as I say, it’s one of the most globalised, industries in the world.

If you ever think that we can’t make enough of something, and you look all the way down the supply chain, it probably, it wasn’t really even, we couldn’t make enough memory, it was actually that there was only one facility TSMC (Taiwanese Semiconductor Manufacturing Company) owned that does advanced packaging where you ship these things to and they actually put this thing together.

So that fab didn’t have the right, it wasn’t large enough. So they’d have to build another fab to be able to... so the capacity is always growing and shrinking and, if you’re TSMC you don’t just want to ramp up because you’ve got demand now, because what about demand in two or three years?

So the point there is that it’s this very orchestrated machine.

But you, ask, why can’t we just carry on with what we got?

Number one: demand is outstripping supply and the type of things that we want to do we need specialist chips. So the, reason that I think we’ll live in a world of heterogeneous compute is because everyone’s chasing server chips — bigger data centres, power them with nuclear power plants, and just stick these really powerful chips with loads of memory in it. But that’s not what the world will look like in 10 years time. The so-called edge is going to be a really critical part of this.

Can you break that down? When you say that’s not what the world is gonna look like in 10 years time... talk us through this contrast.

[00:10:49] Lawrence Lundy-Bryan: Why.

Yeah, why.

[00:10:51] Lawrence Lundy-Bryan: Well because, if you want a really fast response, whether it be in an autonomous car or a drone, high frequency trading or something in an industrial plant, you can’t wait for the response to go all the way to the data centre and come back.

So number one, latency. If you want extremely low latency, you’ll want as much as possible to happen on the device. Whether that’s a robot a car a smartphone or whatever. Right now we don’t have the applications that require super low latency, but we will.

Privacy being another reason as to why I think, we’ll, have a lot of local devices running models so we don’t have to send all the data back to the cloud unencrypted, to wherever datacentre could be on any jurisdiction. So we want things, private. We want things local. And cost.

one of the really interesting factors is that...what’s the build out? A hundred billion, they’ll probably exceed that. That was last year or this coming year. We’re talking $500 billion build out for just Stargate. These numbers are extortionate, and my own personal opinion is that the value will justify that capex spend. there is still value, but we will need to offload a lot of that performance, a lot of those applications, to the user and to local devices.

So what I mean is it’s cheaper to run a local device on a smartphone, just from an infrastructure cost, and also the user pays for it. The user’s paying for the electricity bill. The user’s paying for the phone. So there’ll be a point from a cost of serving customers’ perspective, where soon we will start to see as much as possible, be offloaded locally.

And it won’t just be data centre and edge. It’ll be how do you, what needs to run at the edge, what needs to run at the data centre? And all of the different performance trade offs, as I mentioned, latency, performance, privacy, cost.

That’ll be traded off by the application. So you might see autonomous cars will be an edge. Whereas, scientific simulations will all be data centre. But at the moment, everything’s data centre.

And you, are you interested in investing across that spectrum or are you focused, like when you describe many of the features that you’ve just described, feel like the sorts of things that the UK and Europe and middle-sized countries with energy constraints and a whole load of other kind of structural issues.

They might be particularly interested in as a way to be able to capture value in this, new stack. And it feels like maybe we’re gonna be able to fix our grid infrastructure, fix our planning system,

[00:13:26] Lawrence Lundy-Bryan: I’m sure we will.

…and build out loads of, data centres.But also whether or not we do that we could also be doing quite a lot more on, on device inference.

And that is probably true across Europe or…there’s this company, Emerald AI, which is doing like compute workload orchestration so that, you’re doing model training runs, at the right time based on grid requirements that’s the sort of thing that feels like it would make sense in a sort of like energy constrained sclerotic state like we have today.

And so are you, we haven’t just talked about yet, like you’re focused on Europe, like you’re not investing globally, you’re focused on Europe. What are the companies that you think will come out of Europe? How much do they relate to these factors that you’re talking about?

[00:14:11] Lawrence Lundy-Bryan: I think that we don’t have a, okay. There’s an investment perspective, then there’s Europe and UK perspective, I would say

Yeah. Obviously, you’re bullish on Europe in public, and then in private…

[00:14:19] Lawrence Lundy-Bryan: No, because we have to do something. We might as well try it. I think thatwe can’t play the CapEx game. We, haven’t grown in decades.

So there’s there, okay. I’m a vc, so you play the probability I can get 20 shots on, goal, so I’m wrong, 17 times, that’s fine.

So I’m used to being wrong. that’s fine. But also I think in, probability. And thinking in, in, in bets 50% of all AI inference,

take place at the edge. I, don’t know, but in that world,

We have a shot at the, UK

and Europe.

is it 20, is it 30%?

It probably depends on, the timeline.

but point being, we can’t spend as much as the US the Gulf, and China.

we know that. So let’s just deal with the world as it is. And, take the bet on edge. Now, am I a hundred percent certain? Of course not.

but I can say for sure that, more AI inference will take place at the edge tomorrow than it did today. And certainly that will increase to some ceiling. So yes, I think your instinct, in the question is correct. We haven’t lost the game. And of course it’s in the US’s interest and OpenAI and others’ interest to, to just go all in on data centre. Because they can, it’s easy, for them

to raise money. But let’s not play that game then. And, we have,

Do you not think there’s at least a minimum viable domestic capacity that’s required?

[00:15:48] Lawrence Lundy-Bryan: it depends on your perspective of geopolitics. The uncertainty of geopolitics over the next two to five years.

You’ve written about, say, what if Nvidia or Broadcom or whoever else are banned from exporting to us due to some wider tariff dispute? I think actually I was reading earlier and you literally wrote about what if the US wants to invade Greenland and it starts using these economic tariffs and here we are in this week where this is exactly the question we’re facing. Maybe not specifically with those companies, but the playbook is in play.

[00:16:15] Lawrence Lundy-Bryan: Oh, yeah. I, yes.

Like, if the long term thesis is that AI becomes a critical input to production in the way that energy is, or labour is today, and you are basically streaming that from abroad, and suddenly overnight your input costs become like 20% more expensive.

[00:16:35] Lawrence Lundy-Bryan: It depends on the extent to which you think, deglobalisation is a structural trend or a fad. a fad is, will this, will the deglobalisation, trend continue for the next 20 years? put a probability on that, and everyone will put a slightly different probability.

Yes. I’ve said I think that it’s close to impossible to truly have a sovereign stack. So sovereign technology

I think that’s coming from a semiconductor perspective. that’s really telling, because as I mentioned earlier, this is complex orchestration of thousands of component suppliers.

So the first order you go, let’s have our own AI chip and let’s have our own data centres. sure. But, okay,

Isn’t that a little bit binary? I definitely get your point about the limits of being able to leverage these choke points or whatever.

ASML is like a massive choke point in the semiconductor supply chain, but the Netherlands can’t leverage it because they don’t control their defence or asml are themselves dependent on suppliers based in the us. So it’s not clear that they have like escalation dominance.

But equally Netherlands does have a stake in like the semiconductor supply chain in a way that we do with ARM, although maybe it’s harder to leverage. It’s not quite so physical. But like, there is surely, sovereignty is a spectrum, not a binary. Like we don’t have to go full autarky.

[00:17:54] Lawrence Lundy-Bryan: Brexit. we don’t have to go all in.

No. I’m not saying we should be self-sufficient. I’d like to be totally clear.

[00:17:58] Lawrence Lundy-Bryan: Yeah. Okay. Fair point. Alright. Good pushback. I think, I think, yeah. My point would be you want some, strategic,

assets. and we have to think, what are our strategic assets?

What are we good at? And how do we, grow those industries,

to be even more strategic?

I would say two things. What we are not good at as a state, and I think what Europe

is, I don’t know if this is true of the political class more broadly, but

you need to say, you need to think what will be strategic in five years.

Not what strategic now. Yeah. And, we can get back to that because I think

that’s important

and beyond that it’s, do we have the stomach, both in terms of capital, but also in, in attention to, to think about this over the long term. Which is what China’s particularly good at.

When it comes to quantum, when it comes to,

semiconductors and other things, which is to say, you need a very clear industrial strategy. Isn’t this next three years, we need to make sure we’ve got a 20 year roadmap. And if you really want to do this and build out capacity, it’s gonna cost money for 20 years.

you can’t just suddenly get FOMO’d into doing industrial strategies everyone else is.

And when it comes to the AI stack, as I say, if we were serious about it and Europe was serious about it, we are not doing anywhere near enough thinking about what that means. It is not a AI chip and it is not a AI interconnect.

this, is half of the battle was my joke. I was in the substack, who’s your neon guy? Where you getting your neon from? And they’re like, yeah. and it’s,

There’s interdependency all the way down.

[00:19:27] Lawrence Lundy-Bryan: Interdependent. Yeah. Turtles all the way down. Exactly. so let’s the, world as it is, so to your point, what are we good at?

What could we be good at? What could the choke points be?

or where we have the equivalent of an ASML, and this is where being a vc,

you have to look not what, as I say what the world is, today

and think three, five years.

What markets, what applications, what, what customer demand will exist that doesn’t exist now, that’s likely to grow. And in which case, what is the UK or Europe currently producing, but at low volumes.

it would require government support to grow to the point at which maybe you would find external capital.

This is a really hard pitch to make and I can make it really specific: compound semiconductors, which I’ve written about. Compound semiconductors is basically, instead of using silicon, which is what all of our chips are broadly made from,

use other types of materials. So things like silicon carbide, which is a different material

or gallium nitride,

Each of these materials has a different property, which makes it useful.

they are better at running hot, at higher frequencies. you can’t stick a silicon chip next to a battery.

it just gets too hot. so you use other materials, silicon carbide or GaN think about, electric cars being a really good example.

What about the chips inside? You don’t think China? Were thinking about that 10 years ago. yeah,

and so that’s just like example.

but we do have very good,

compound semiconductor capacity in uk.

We’re very, particularly good at a Cardiff, down in Cardiff we have a great ecosystem. We have

some companies that are leading, in this space. the reason I tie this back to the, government,

focus is because I’ve looked at lots of compound semiconductor companies,

and with my VC hat on I look at pretty small markets, pretty small growth rates, limited number of buyers, a couple of ev in the uk. And I think, that, doesn’t seem like this developer tool company around the corner that’s growing a hundred percent year over year.

it’s not the best place to put my money.

that is a place the government should be thinking.

So do you, I wanna tease this out a bit more, you’ve just launched a semiconductor fund, where do you see for the companies that you’re speaking to or the ones that you’ve already invested in, who are the bigger customers that are buying their products?

There’s existing demand in the semiconductor ecosystem today. And you think there are new companies that can be founded to serve that existing demand, but better. Or are you also looking at new applications, new use cases, growing parts of the market where there’ll be new customers coming to the fore, enabled by some kind of new chip design, whatever it is. Like how much are you at a high level with your thesis, do you have a rough, weighting towards one of those two?

[00:22:12] Lawrence Lundy-Bryan: Space is a good example.

Interesting.

[00:22:14] Lawrence Lundy-Bryan: Put chips into space. We do, but they, have to be radiation hardened because space has radiation, so that’s pretty tricky for silicon chips to deal with.

we can stick,

more chips into space. That’s not to talk about this idea of

So

This is specifically for for the compute of space operations rather than…I’m also thinking about quite a lot of people in Silicon Valley now being like, we need to actually make chips in space?

[00:22:37] Lawrence Lundy-Bryan: yeah, that’s an interest and we have a good company Space Forge that is trying to do this there are some benefits in terms of gravity, in terms of. Solar power, but I know that aside

Fewer planning restrictions in space…

[00:22:48] Lawrence Lundy-Bryan: Exactly. that’s interesting. I’m not sure if the economics actually makes sense in just doing it on, the ground, but,

but certainly we have these satellites that are very underpowered in terms of what they can do. And also if you think about the round trip — I mentioned latency — the round trip to a data centre.

The roundtrip up to space is costly. You don’t want, you don’t wanna send, much up there. So there’s some really interesting, designs and thinking about, if we wanted to, put a more powerful computer up in space. how would you do it first? how would you design it?

some of this is called compute stick,

the compute in the sensor itself. you can actually do some logic or operations within the sensing data. imagine this, you take a, a huge,

a couple of a terabyte worth of data, right? And instead of send, right now, you’d send all of that back for processing, and then you’d send

the results.

But if you could do some processing at the edge or on the on board, and it turns out the only important part of that is the bottom left, whatever it might be, and you just send that back, you save yourself fortune in terms of, cost. So that’s just like one example of we don’t really have specific specialised chips for space, but we could, another really good example is, and I think this is gonna be a much larger part of the fund, is photonics:

computing with light. We already move pretty much all of our data around the earth over fiber optic, So we already use light to move data. as we get better, at building little lasers to make the light, and as we get better at building little modulators to modulate the intensity and photo detectors to know if the light is on or off.

All of these things, again, smaller and smaller, to the point in which,

relatively recently we’re able to stick with these on a chip or photonic integrated chip. this is all very immature compared to silicon. which churns these things out in the trillions.

it’s still hard to do.

It’s still quite complex, but we are getting to the point now where we can put all of these components

on a chip,

on a different material. It’s generally not silicon.

what can we do if we can shrink, these lasers and things to smaller and smaller sizes?

A whole bunch of things. if you’re wearing an Apple watch You’ve got an image sensor on the back of that.

blood and sensing various other biomarkers. If you have a more powerful,

photonic integrated chip, you could sense more biomarkers.

any sort of wearable device will be seeing more and more powerful chips.

Yeah.

[00:25:12] Lawrence Lundy-Bryan: So they’re just two examples and you can think about autonomous cars or robots as other places for which,

we’re probably not gonna stick GPUs in the next five or 10 years. That’s just a couple of new application areas That we can see whole new chip companies.

Interesting. I want to unpack a little bit more about, the fund is a $30 million fund. And you talked about making kind of 20 bets, give or take. Talk us through the fund sizing. You were talking a little bit before about some of the massive deals in this space. I’m interested in…if I’m someone that wants the UK to win in semiconductors. I think funds that invest in UK companies in semiconductors are a good thing. Why have you rightsized your fund to make that work?

[00:25:56] Lawrence Lundy-Bryan: We haven’t have, we, this is the truth. Maybe that’s the question. It’s not for lack of desire, ambition or, willing. It’s an MVP to prove that this is, important. just like any startup, I see this as a pre-seed, to prove, and what validation points do we have in 24 months

That prove, you should give us more money to do this bigger and better, and to scale. That’s the best way to think about this. it’s not right sized.

it’s too small. a good data point recently, is

There’s a company, in Europe raising a 200 million seed.

Etched, which is public information now, which is an AI chip company raised a 500 million series A, I guess.

And some of those numbers are signal. “We are serious.” But also some of those signals are we’re taking on these big chip companies. we need to have a lot of capital both for hiring, but also like actually going to a fab to make chips costs a fortune.

there’s an interesting thesis about how do we make chips cheaper to make Which is like a how do we get away from the, fab cost 10 billion. why doesn’t the fab cost a hundred million? It’s an interesting series of, challenges there, but broadly, the world as it is now, you want some tape outs? Might need a hundred million, like for like advanced nodes, which is like the most, if you wanted very small chips for smartphones or for AI, chips.

The point is, these things cost a lot. So what can we do with 30 million?

We have to be very early. And be, and because we’re specialist, we can go earlier than the average fund. Because you’d like to think we roughly understand the markets we’re operating in.

the second reason why I think we could be more capital efficient, I dunno if the right number is, we could be a £50m fund or a £100m, I don’t know. But we do have strategic investors that we can introduce to our portfolio where we can give them services that wouldn’t be able to get elsewhere.

So one, one of those is Global Foundries, the third largest, fab facility in the world. We also have a, Taiwanese,photonic, company called Radiant Optoelectronics.

We can introduce our portfolio to these companies and their surrounding ecosystem to do things faster and cheaper.So the hope would be: without us, you’d have to raise £5 million, with us you raise £2.5m. We could be more capital efficient, but I’m not here to say that’s anywhere near enough especially for these frontier AI chips. Now we’ve spent quite a lot of the interview talking about one important, but small part of the broader semiconductor space.

We could be talking about wifi chips, or innovation in sensing, hyperspectral imaging for drone warfare. We can talk about loads of other things where you don’t need a hundred million. you are talking much more sensible, deep techy type numbers. We still need millions, but you don’t need hundreds of millions. If we were only investing in AI chips. Yeah, you probably one a $200 million fund to off to start, but we’re not, and there’s loads of other spaces that are small, will grow.

And you can invest £1m.

I think it’s interesting ‘cause you maybe to go back to the point you started with explaining the fund, explaining the specialist kind of focus and structure. There’s clearly been this massive shift underway over the last few years.

You have, as you call them, the agglomerators kind of hoovering up evermore and more capital and actually also like ever increasing returns to scale because they can invest in the product, they can add more value, et cetera, et cetera. And then at the other end, the sort of specialists and maybe the messy middle eventually gets washed out.

We’ll see. But when you think about that context, if that’s one trend and then another trend that you’ve definitely thought about quite a lot is the increasingly close relationship between governments and particularlylarger capital allocators.

It’s interesting if you’re looking at this from the perspective, not just as like the GP in the fund, but from like a government perspective of

how do we secure sovereignty in AI? How do we invest well, how do we understand how the market is shaping around us, restructuring around us? And then like, where is the best place to intervene?

And so in the UK, the government has been proposing, is, launching SovAI, this new, this new strategic venture fund for AI, basically. And it’s not just investing in AI chips, it’s, gonna have a wider remit, 500 million, a mix of both capital and in kind support.

based on the reflections that you thought about there with Cloudberry, how do you think about applying those lessons to the government’s side of things as well? What does it, you talked about

with, industrial strategy, you’ve gotta be thinking five years ahead.

And Made in China 2025. It’s this long running strategy and it’s not impossible to do industrial strategy. I think there’s a lot of people who reject it on the merits, but it, requires the capability to do it well. And if you have the capability you can be more ambitious.

And this is a, SovAI is an exercise in capability building. What do you think is required to make that work?

[00:30:47] Lawrence Lundy-Bryan: I think that the single biggest challenge is that any way of thinking about industrial strategy, and us talking about it here, is top down.It’s define the specific things that everybody agrees are important.

and then allocate money to that. And that is sort of if you’re talking about manufacturing, agriculture. we want our own domestic food supply Because it’s relatively slow moving, space and industry. So this idea of top down, ‘this is important’, ‘this is important’ with some degree of certainty. Go ahead. I think you cannot do industrial strategy in 2026 by saying This is a hundred percent important, and this is a hundred percent important because we don’t know.

So the key, insight is that government and civil service are inherently not good at thinking about high levels of uncertainty and making bets because, That’s ultimately what you have to do good industrial strategy in this, in the age of ai, but more, more specifically the age in which things are moving faster because of software on the internet.So the first thing you’d have to do if you were with soft AI or you wanted to do anything is to, understand that you have high degrees of uncertainty. You don’t know all the answers when you set out.

how do you address that? you have two things. One is velocity and one is adaptability.

If you don’t know what the answers are and everything moves fast, you have to move fast. And the second thing is, for SovAI for anyone else is adaptability. That is to say, if you take bets, you will be wrong. So it’s not about the taking the bet, it’s about the process. It’s about understanding what did we get wrong in the process, that we could improve. But still aware that the bet might be wrong.

So I’ll give a good example. Quantum computing.

If you’re gonna have a discussion about strategic autonomy, you’re gonna have a discussion about what’s gonna be important in five, 10 years. The obvious candidate will be, let’s invest in quantum computing because that’s the next thing. if you own your own quantum computer… Autonomy dot profit, the dots are doing a lot of heavy lifting. Maybe.

It’s a bet. But the point is, we don’t know. and, more than that, it’s which type of quantum computer Yeah. Are we talking about trapped ions, superconducting? There’s no answer.

I remember — I’m a fan of industrial strategy, done well I should say — but I remember having conversations with policy people a few years ago saying oh, we need to invest in they would pick up themes. But no, you can only invest in assets. You can only invest in companies. And that is the sort of top down, “we need to invest in themes”.

[00:33:27] Lawrence Lundy-Bryan: Mea Culpa. That’s what I was bad at.

Interesting.

[00:33:29] Lawrence Lundy-Bryan: So that’s the learning, right? For me, over however many years I’ve been investing in a decade or so, my title was always Head of Research. If you do enough research, you could be super smart and you could pick the right themes before anyone else, and ... profit.

And, okay, there’s a part of that. Part of it is understanding, what a quantum computer is. And okay, how do you trap these ions or whatever, what is super conducted logic? You might understand it fine, but I, think a good learning, was with a nuclear fusion company that we were looking at.

And maybe the hubristic approach is, I’m gonna go and speak to 20 people in nuclear fusion, and I’m gonna understand, is this the right bet?

It’s not to say all 20 gave a different answer, but all 20 gave a different answer.

And so venture is a great example of where you go, okay, we’re happy with, lots of degrees of uncertainty. So we’re happy to take the bet regardless. But I think a top down industrial strategy and coming up with a theme is the idea that you can somehow predict the, future you bet on these themes and you win. My learning, having done research and done all those themes is, that’s probably the wrong way to do it.

Actually, really good founders will sniff out opportunities. If you are in touch with the ecosystem enough. Ideas, themes will bubble up or companies will bubble up. And if you are, bottom up, which is why the BBB maybe you touched on ‘cause you’ve really underutilized resource because if anyone has access to the entire ecosystem bubbling up, you could understand on the one hand, okay, we think quantum, on the other hand, let’s just speak to everybody in ecosystem.

The BBB is the UK’s largest LP I think it’s invested in nearly a fifth of funds in the UK. And therefore it gets costly report from every one of those funds with it’s not perfect information, but like it is sitting on like the best data set of the early stage venture ecosystem.

[00:35:29] Lawrence Lundy-Bryan: Can we, in the, can can’t we connect that to Claude code like now yesterday. And then we’d all get the answer.

Exactly. I literally used to fill in the quarterly reports, right? And I’d be like, this is how the company’s doing. These are the commercial bottlenecks. These are things that they’re doing well. Here are some policy and regulatory things that could be relevant to them. And we have that information and then we don’t do anything with it. Part of this is about capability, but can I just ask you, I just wanna go down a rabbit hole a little bit on what you said before about: you can be looking at something like fusion, you can speak to 20 different experts, they each say something different.

[00:35:55] Lawrence Lundy-Bryan: Yeah.

And I really, relate to this point. There’s a thing I’ve been trying to articulate, which I’m gonna roughly call “domain elasticity”.

[00:36:04] Lawrence Lundy-Bryan: Sure, you are

If you look at founders who have basically been incredibly successful in one domain. How do they switch from that domain to another domain and still be successful? Elon Musk is a good example.

[00:36:15] Lawrence Lundy-Bryan: There’s very few that do, is the answer. I think.

Well, yeah. Okay. that’s, that’s interesting. But like the, if the job of a VC is to find the people who are capable of doing it, you look at the Boom Supersonic, CEO. He goes from Groupon to founding a supersonic flight company. it’s interesting. What is it about that person?

It’s not necessarily that like while he was at Groupon he was actually a supersonic like physics PhD. But he just had some other skillset.

[00:36:39] Lawrence Lundy-Bryan: Yeah.

And then the flip like the, just to go and couple that with, as a VC, you are speaking to founders across a whole range of different domains every day, who in theory should probably know more about what they’re building than you will, you might have the breadth, but each one of them will have more, specificity.

[00:36:55] Lawrence Lundy-Bryan: Yeah.

And so you have to, there’s always gonna be some gap in knowledge of how do I build conviction in this person to go and build this thing? And you more than most people have gone deep onto the research to be like, I’m gonna get as far as close as I can to understanding the products and technical hypotheses that this person is coming to.

[00:37:13] Lawrence Lundy-Bryan: Yeah.

But not every VC will have that approach. They’ll have different approaches or they’ll go later stage and try and wait they’ll have it derisked in other ways.

Part of what I’m trying to do with these podcasts, particularly when we have more founders on, is to try and help policy makers and other people in and around this sphere understand why some companies are credible and legitimate. And basically I think startups are always playing on hard mode. they don’t have the brand recognition, they don’t have the reputation, et cetera. quite a lot of other people, very reasonably will go, I know nothing about what you’re building, so I dunno why I should believe you when you tell me that, the nuclear fusion company you’re building is actually real.

[00:37:49] Lawrence Lundy-Bryan: yeah.

I’ve been in meetings with very senior foreign policy people who would love us to fix our energy policy. But then whenever anyone mentions an SMR, they’re like, can I see one please? I’ve never seen one. They just keep being promised to me…blah, blah, blah…

[00:38:01] Lawrence Lundy-Bryan: fascinating,

So it’s really hard for them to overcome this sense of why would I believe that the constraints which you’re telling me are gonna change will ever be broken? So all that’s to say when you are thinking about new classes of problems, new founders who are like some, I dunno, step beyond where you are on something. Like how are you building conviction in someone and understanding them to be credible and legitimate in a way that like maybe other people can learn from too.

[00:38:29] Lawrence Lundy-Bryan: Yeah. I did the project probably five years ago called State of the Future, which is still the name of the newsletter.

The conceit behind that, and this was before, LLMs, which rendered it pointless. we looked at 150 technologies, which actually retrospectively the wrong way to look at, innovation.

You speak to an academic about how innovation happens, but the truth is like there’s a WhatsApp group and someone shared a term sheet and said, who’s in? I’m not sure that matches how we think this is done.

We should be looking at problem statements.

But regardless, whatever, looked at the technology.

So everything from brain computing interfaces to, chips to these little Lego blocks of how you make chips to, mRNA vaccines. So like really, broad.

And the idea was to, try and identify a couple of things within each. So I would look at the technical maturity of some of these things.

Was there a market catalyst? Was there something changing in the market? Whether it would be technical? on the, supply side or on the demand side, like consumer behavior change. clean meat would be another example of where the consumer behaviour changed.

Look at how different the technology is to what already exists. So novelty and then impact. how impactful could this technology be? Is this a $1 billion market? Is it a $10bn? Is it a $100bn? Is it a trillion?

[00:39:49] Andrew Bennett: The founder is necessarily almost, if it’s a venture backable company, they’re coming with a contrarian hypothesis.

You speak to all the existing experts and they go, that, that won’t work.

[00:39:58] Lawrence Lundy-Bryan: Yes, okay..You are, I’m not, an expert, but I can tell you something about something you don’t know, which might be relevant.

This is all around how you commercialize an innovation.

Yeah.

[00:40:06] Lawrence Lundy-Bryan: It’s not so much about the specific innovation and some of the market challenges, consolidation of the market, or how many buyers are there, they’re all pretty standard things.

But, so you can tell them, so you can help them understand how they sit within the broader context, which I think is really useful as I’ve looked at lots of technologies, but I think the learning really to how you can sniff test is that most of the fund returners, most of the outliers say pretty outlandish things that if you are asking experts, the experts will say it’s probably not plausible. So this is, the exact game. And which is why VCs can often look stupid ‘cause you back stupid, potentially stupid things. So number one,

All the experts told you it wouldn’t work and it didn’t work.

[00:40:52] Lawrence Lundy-Bryan: Sure. And it didn’t work. Great. Yeah. of course not. Exactly. So, I think, there’s, yeah.

What’s the like Nat Friedman thing? Pessimists are right, optimists to make money?

[00:41:02] Lawrence Lundy-Bryan: Yeah, exactly. And, I think that’s a good one. How we do that is, is just I like to think it’s the trade.

If you’ve been doing it for 10 years, speaking to, 10 founders, 10 founders a week.

You’ve got a really large data set for which just right now, some intuitive view of is someone bullshitting, right?

Which is a reasonable, you can get yourself to some level of this person is bullshitting.And then below that, really it’s will this work?

It depends when you’re investing. At our stage, ‘will this work? Maybe.’ ‘Is the market big enough if it works, yes or no? If yes, okay.’ And then you, proceed down that route. And this is what I mean about taking bets. If you’re doing it early enough, you are almost certainly gonna find one or two people that say this will never work.

If you are used to risk mitigation, if you are used to not wasting taxpayers’ monies, if you are used to betting on certain things, then you can always find a reason not to make an investment.

And you can always seem smart because, I spoke with someone at X big company that you’ve heard of, and they said it will never work.

Yeah.

[00:42:04] Lawrence Lundy-Bryan: So I can feel good now. That’s the challenge. And I often find that’s the challenge in hiring people as well, like hiring people who are prepared to stick their neck out a little bit. It’s really high degrees of uncertainty and being very comfortable with being wrong a lot.

they’re not then, that’s not how we are taught in British education system, I would argue. Being wrong is hard.

And you also made this point before about if you are, protecting taxpayers’ money, like you, I think when you were writing about SovAI, you were saying there is always a reason for ask to ask for more data, more review, more like process. And that might go against the velocity that’s required for that particular institution to succeed.

[00:42:43] Lawrence Lundy-Bryan: I know I’m pretty radical about this, but also like I totally understand the incentives, because there are no incentives To be fast and be wrong.

That these incentives don’t exist and they barely exist in venture capital firms. The truth is some of this is just structural. Fund one, you’re investing in weird things for whatever reason. You have some alpha

[00:43:01] Andrew Bennett: You’ve got something to prove, yeah.

[00:43:02] Lawrence Lundy-Bryan: And then by fund two or three you’ve put processes in place, a bureaucracy to avoid failures. It’s, very much, a structural problem in all institutions. Interesting. So it’s no surprise in my view that, that this happens in institutions that have been around for centuries.

[00:43:17] Andrew Bennett: How did the three of you at Cloudberry make decisions?

[00:43:20] Lawrence Lundy-Bryan: I’ve got a, few different learnings from different funds, which I’m trying to take with me here.And many and I think there’s lots of way to make money, frankly. You could be consensual, you could be lone wolf. The key is having really different, like coming from different places.

By that I mean, if we are thinking about what our fund looks like, I’m the one that’s gonna, I dunno if this is right or it’d be interesting. I’m not sure if this is right, but I think I can put myself in the mind of other VCs.

[00:43:52] Andrew Bennett: Invest in things that will get markups ‘cause other people like them.

[00:43:55] Lawrence Lundy-Bryan: Maybe.

[00:43:56] Andrew Bennett: Yeah.

[00:43:57] Lawrence Lundy-Bryan: but it’s one way to play the game, I think.

[00:43:58] Andrew Bennett: I’m just quoting your substack,

[00:43:59] Lawrence Lundy-Bryan: Quoting my own substack quote me back at me.Which I think could be helpful. You might end up, converging on the same things. Yeah.

I think, the, Rene and Veera are better at understanding what other semi customers might want, semiconductor customers, what will a customer want?

[00:44:12] Andrew Bennett: Is that their background?

[00:44:13] Lawrence Lundy-Bryan: What will a CVC... Exactly, yeah. So that’s their background. What will a corporate venture capital firm be interested in within the ecosystem? Operationally, Rene has built a company and sold a company. So he understands lots of things that I don’t. I think

what I do at pre-seed is I invest X amount of money to get to Y milestone. That’s the key thing I’m doing. identify what is X milestone. What does it take to get there? And once you’ve hit X milestone, who will invest to get you to the next milestone?

if you can reduce. the complexity to that.

You can then think much more clearly about, okay,

two years, 3, 4, 5 years down the line, the uncertainty continues to go up. What can I really think about?

[00:44:49] Andrew Bennett: Do you think that relay race effectively as you’re describing it, like I invest so that they, the company will get to the next stage and then someone else will come in and help them get to the next stage.

Do you think that relay race breaks down when you have these, structural shifts going on between the smiling curve of the specialists and the agglomerators, it’s

[00:45:11] Andrew Bennett: are the agglomerates just gonna, are they gonna come in at the seed and then back them from there?

Is that how you

[00:45:15] Lawrence Lundy-Bryan: see the No. ‘cause if, think about if you think about, institutional bureaucracy

I think then, why not wait? Why not wait to [Series] A, we can de-risk it at [Series] A and then we can give them 50 million so it’s, I think, actually a better way of thinking about it: venture capital, investing in truly, unusual things, new markets is venture capital and anything, maybe post B is private equity growth capital — it’s different. And I think we have to think about that as probably bifurcating in a way that it hasn’t before.

[00:45:47] Andrew Bennett: Interesting.

[00:45:48] Lawrence Lundy-Bryan: And if it does bifurcate, then could they come further down and do seed?

Yes, but it doesn’t materially move the dial for them in terms of what’s 1 million in a 15 billion funds

[00:45:58] Andrew Bennett: Also just like as a founder you’re just like a tiny part of their portfolio. Yeah.

[00:46:02] Lawrence Lundy-Bryan: they’re not gonna be on what side responding to your messages at midnight or whatever.

So I think there’s, the specialists will always be there, but I do think if that theory of, structural change is right, the smiley curve. you really have to know the rest of the capital stack. the relay races may be a really good analogy because you really need to know who you are handing it off to.

And I think my previous view through investing in a theme,

[00:46:28] Lawrence Lundy-Bryan: of venture capital theory is in the theme. And you get the timing right and the market emerges. And of course, follow on investors will arrive because the markets arrive. in the perfect world, you’ve invested in the right theme at the right time.

And by the time they hit [Series] A, they’ve got £1m, £1.5m, £2m revenue and it hits. That’s the perfect world. And in most cases, that’s never quite right. So you can de-risk it by speaking to the AUM aggregators, the later stage investors: getting a sense of what’s interesting to them, what will move the dial for them? What are they interested in? And trying to match it up.

I’m reluctant. to go all in on that as the game.

[00:47:09] Andrew Bennett: Yeah.

[00:47:09] Lawrence Lundy-Bryan: Because it’s the sales game. You hand it off to someone else, you get markups.

I don’t think that’s how you build a sustaining fund. If Cloudberry becomes fund five, fund six, I don’t think it’s because we did that well. I think it’s because we did back sort of unusual companies doing unusual things, outlier type people.

It wasn’t obvious who the next investor is. you have 20 bets and I don’t think you can take 20 bets on weird, funky, unusual people where you have no idea who’s gonna invest next. Because I think that’s probably too much risk to take.

[00:47:40] Andrew Bennett: Yeah. I think it’s pretty interesting that you’re doing this in Europe and that you’re also bullish on the opportunity to do this in Europe.

Can you just say a little bit more about where that bullishness comes from?

[00:47:53] Lawrence Lundy-Bryan: We have to do something, I dunno what to tell you.

honest, honestly.

because.

I’m terminally online. No, I don’t think it’s because I’m too online, but,

[00:48:04] Andrew Bennett: Europe already has a strong semis ecosystem and it can be even stronger.

I think there’s a very good case

[00:48:09] Lawrence Lundy-Bryan: I make the financial case. Yeah. that, I think

Europe has this huge, talent base. okay. number one, I’m, pretty radicalised to the fact that Europe needs to wake up and we are in an age of deglobalisation.

strategically, we’ve been reliant on the US for too long. Hopefully one day we can rely on the US again, but as it is today,

we need to, invest, in our own capabilities.

[00:48:31] Andrew Bennett: Would you call yourself a patriot?

[00:48:33] Lawrence Lundy-Bryan: Are you trying to think about how we can reclaim that word?

[00:48:35] Andrew Bennett: Yeah.

[00:48:36] Lawrence Lundy-Bryan: Yeah. Interesting.

I think there’s something interesting from a reclamation perspective how we can

[00:48:41] Andrew Bennett: Right.

[00:48:41] Lawrence Lundy-Bryan: Regain that from a, as liberals,

so yeah, for certain a patriot, which is why this. Europe, UK thing is interesting. Again, for me, now, first and foremost, I think of myself as a Brit.

Not European, but for sure our European partners and colleagues. And this is a Finnish fund.

for sure this is your European play, but from a UK perspective, I think we just sleeping for too long.

hopefully it’s not, terminal.

Hopefully we can wake ourselves up again. But I’m radicalised to the idea of we have unbelievable talent. R&D like number one, technical R&D talent. We have Eindhoven, we have Bristol, we have Southampton, Oxford, Cambridge,

Munich.

Lots and lots of really deep pools of talent.

what we’ve lacked,

and I think this is the bet I’m making, what we’ve lacked is enough ambitious founders that want to win globally. I don’t think that’s because they don’t exist. I think that’s because they were, I think their ambition, they were defanged.

I think their ambition has been reduced because they were too busy talking to EIS funds that were saying where’s your P&L? yeah, at preseed. So number one, I thinkthe founders were there, but they’ve been slowly battered and so they asked for 800 k because that’s all they think they can get.

I think, Cloudberry and others, Plural, namely, but there are plenty of others out there saying we can be more ambitious. Because we have ambitious investors too. So number one, I think the more we shout about the fact that think big and we’ll back you, I think that means we can, enable people to return to their natural ambition, right?

We don’t need to create more founders, they’re there. And number two, I think a lot of them went to the US quite rightly. That would be how I would de-risk my company two years ago and now they don’t. So they’re here unless give them the capital they need to build the businesses here.

[00:50:36] Andrew Bennett: I think there’s this concept in, evolutionary biology called niche construction.

a good example of this is a beaver building a dam.

it’s like there will be one organism that intervenes on their ecosystem and that intervention means that the ecosystem reshapes itself around them. And there’s something in, what you’re doing, Plural, and like a few others where there is a sort of like unreasonable belief and ambition relative to the, status quo that is creating value because there are founders that no longer have to go down these, slower paths and they can believe in themselves and go, down the ambitious path.

[00:51:13] Andrew Bennett: And there’s just something interesting there about

[00:51:16] Andrew Bennett: part of what we’re doing here is figuring out how to solve problems through company building, through capital allocation, and through policy and government,

understanding actually, like the craft of building a venture company is interesting in and of itself. A venture fund is, its, Cloudberry is itself a company.the journey that you’ve gone on to build it and the intervention that it can have in the ecosystem is interesting.

[00:51:40] Lawrence Lundy-Bryan: The thing that makes me think of is peculiarly a European and British thing, but maybe this is why people, head out to the, US which is, it’s not necessarily the tall poppy syndrome, but there’s definitely something in, our reflexive culture, which...

“i wanna build a trillion dollar company.” People will laugh at that.

[00:52:01] Andrew Bennett: Yeah.

[00:52:02] Lawrence Lundy-Bryan: There’s some instinctive, skepticism in that somehow the and I think the, US do it, well. I think, I’m not here to change culture because one person doesn’t change it, but there is something in a small group of people, for which I think you are, collecting, pulling together small group of people that do shout it from the rooftops that say we don’t have to just accept declinism anymore.

[00:52:25] Andrew Bennett: Yeah.

[00:52:25] Lawrence Lundy-Bryan: We can build a new city in Cambridge. Why not? and actually think we’re gonna build a trillion dollar company. how are you gonna get all

the team? How are you gonna raise the money? yeah, we’ll figure it out. Like this sort of bullishness.

I thought that I would be more bullish in my twenties, and then slowly you become more conservative and less bullish. But for some reason I’ve come out the other end and I’m that’s why are we can’t just keep accepting this. there’s many more people like me.

Maybe my age or not, who are thinking like, why are we just accepting this declinism? We need to shake people out of this is the way it has to be. And a good example is all the stories, what did we get for GDP growth? 0.1%.

And it was written up like ‘good day for Rachel Reeves’. Good day. Good day. 0.1%. But there’s no context.

[00:53:07] Andrew Bennett: yeah.

[00:53:08] Lawrence Lundy-Bryan: Okay. If someone said we want to aim for seven, then 7% is ridiculous. let’s say, 3%, let’s say 3% or 2.5%. crazy. Ask Dario at Anthropic, or let’s say we wanna aim for 3%.

Yeah.And that is the target they’d get, you’d get laughed at. Yeah. But I think there’s that sort of level of ambition, not just from politicians. I’m sure politicians would like to say that, but they would

[00:53:31] Andrew Bennett: Yeah.

[00:53:31] Lawrence Lundy-Bryan: Get killed for doing so. I think you can do the same on the venture and ventures and yeah.

It’s a bit easier to say we’re gonna build a trillion dollar companies, wanna do it in the UK and we’re gonna invest in them. And we’re gonna help them be ambitious to do that.

[00:53:43] Andrew Bennett: I think, what you’re speaking to there is a really interesting cultural difference if you have a financial stake in being optimistic and experience with companies who are like, no one knows who they are, they have no reputation, and then over five years they become global winners.

You believe that things can change quickly. Whereas I think if you often don’t have that and governments have a lot of other things to, to think about and care about as well, but it’s really hard to believe that the future can be radically different

from the present day

there’s a culture that you’re inculcated in that is, transferable is important for actually raising ambition.

there isn’t enough of it.

[00:54:18] Lawrence Lundy-Bryan: Yeah. Like how do we change a thought about this from a, political perspective,

which I’m, treading dangerously outside my ream of competency, which is my job as a VC

[00:54:28] Andrew Bennett: Of course.

[00:54:28] Lawrence Lundy-Bryan: So I will continue. Of course. what, could I do before the next election? in the next two years. That the average person would feel like things are changing.

one idea is, everybody sees autonomous cars on their streets. And then we say, the taxi drivers are gonna revolt Sure. But things can change. There’s an autonomous car on the street. That’s cool.

[00:54:47] Andrew Bennett: Yeah.

[00:54:48] Lawrence Lundy-Bryan: That’s interesting. That didn’t exist.

Yeah. That’s a new technology, a new enable. It’s gonna bring lots of problems. It’ll run over cats. It’ll be on the Daily Mail headline. And everyone’s gonna say, but

[00:54:55] Andrew Bennett: they’re also safer than most cars.

exactly. The logic, doesn’t, we’ll not win that argument. It does on this podcast.

[00:55:02] Lawrence Lundy-Bryan: Okay. Logic wins on this podcast. It’s a safe space.

what could you do to say the future will be brighter?

[00:55:07] Andrew Bennett: Yeah.

[00:55:07] Lawrence Lundy-Bryan: How do we go back to the atomic age, the futurama, like the future. I think that’s the missing link. And I can’t do it from a semi fund, like a semiconductor fund is a part of it, but

[00:55:17] Andrew Bennett: Maybe you make train wifi possible.

[00:55:19] Lawrence Lundy-Bryan: Train wifi is a starting point. Then nuclear

[00:55:21] Andrew Bennett: fusion. Visible. Tangible, yeah. Nuclear fusion and train wifi. Let’s have both of those by 2029.

[00:55:26] Lawrence Lundy-Bryan: Fix the potholes. Yeah. And then, and, train wifi and then fusion on the other end. But these things should be

[00:55:31] Andrew Bennett: visible, tangible things.

[00:55:32] Lawrence Lundy-Bryan: Yeah. Visible. Exactly. The autonomous cars is the only one I could think of that could actually

[00:55:36] Andrew Bennett: happen. there’s tons of others. Clean meat in supermarkets,BV loss, drones that fly beyond visual line of sight do logistic and then that’s, licensed.

Yeah. Yeah. Drone deliveries is a good one. there’s a, whole chunk of things like AI medical devices that, that currently, like again, like there’s a whole lot stuff across our regulatory state that is like interesting, super important, these are markets that matter new technologies and companies can improve our lives in such a visible, tangible, visceral way.

[00:56:02] Lawrence Lundy-Bryan: Do you think

[00:56:03] Andrew Bennett: that drives a wedge between that and just like the social media giants that people have an aversion to

[00:56:08] Lawrence Lundy-Bryan: Yeah.

But you’re touching on it there, and AI.

People have, this is the question, and this is the pushback when I floated this to autonomous cars, people, the public won’t like that.

the extent here is the leadership take, like leading people into the bright future, the bright light of technology. things will be better. I think if you were to do that, many of those things, the drones, people complain about the noise, they complain about the risks.

I think it would be a negative view of it. Not necessarily a positive. So here I am and we are, as technologists saying, this would be so cool. I think there’ll be an immediate pushback on this isn’t cool for all these reasons.

[00:56:46] Andrew Bennett: Yeah. And it’s hard to know how to. Medical devices might be the one, get cancer diagnoses down by a hundred days or whatever. I think there probably are ways to make it much more positive, but the instant reaction would be a negative one.If we stay in this model of nothing ever changes and we never have any of the visible signs of progress that then give people a stake in, like things can get better.

Yeah. Then, we’ll end up in this, we’ll stay in this terminal valley nuclear fusion. Fix the train wifi.

[00:57:16] Lawrence Lundy-Bryan: I’ve set up a, silicon semiconductor fund, which I think is an important an element to it, but because a lot of what you’ve described, is relying on semiconductors. Yeah.

Yeah. So if we wanna make people’s lives better, we want economic growth. Yeah. I think, yes. So also I think nuclear fusion could be part of it, but that’s a longer term horizon. The core engine, of economic growth over the next decade will be how we turn energy into intelligence.

Again, I mean to summarise the whole thesis behind this is we’ll have energy and that’s a problem, but then we need to turn it to intelligence. Both whether it’s in the data centre or in cars or whatever, but turning into intelligence, that’s a silicon problem.

That’s actually how we make the chips.

But broadly speaking, how do we turn it into intelligence? And, that is the defining question of this decade for me and for the UK and for policymakers.

[00:58:01] Andrew Bennett: Good luck!

[00:58:02] Lawrence Lundy-Bryan: I’ll do my best.

[00:58:03] Andrew Bennett: Thanks so much for coming on, Lawrence.

[00:58:04] Lawrence Lundy-Bryan: Thanks.

[00:58:06] Andrew Bennett: That was Lawrence Lundy-Bryan, GP at Cloudberry VC, Europe’s first dedicated semiconductor fund. And frankly, for the UK to win, we’re gonna need so much more of his energy. Over the course of this year, we’ll be speaking to more founders, investors, policy makers, writers, artists, about how we can raise our collective ambition for who we are, where we’re going, how we get there. So please share your feedback, small or large. Help make this as good as possible. Like, subscribe, share, help me toil in the content mines. And as always, thank you to the Centre for British Progress for supporting Sovereign Albion, to Julia Willemyns, David Lawrence, and Alys Key for editorial support. And to Podcast House for production support. Until next time.

That might sound a little grand, but it’s when I’ve taken these ideas seriously, and stopped caveating the ambition, that I’ve had much deeper conversations and built much deeper friendships and relationships to learn from — especially with founders, artists, policymakers and many others.

These conversations tend to skew a bit more Sovereign — i.e. focusing on hard power, critical industries, state capacity and AI — but frequently stray into the Albion: the aesthetics and cultures that run in parallel to progress. In effect: who we are, where we’re going and how we get there.

So I’m going to record some of these conversations. Maybe you’re like me: you enjoy listening to Dwarkesh, but you’re left wondering what AI continuing to scale means for us. What is our strategy, in Britain, as it does? Maybe you’d like to see us building more infrastructure and homes across the country, but also think protecting the mysticism and enchantment of nature is critical for restoring collective national purpose. Maybe you believe that startups are the delivery units of progress and you’re frustrated by generic, reflexive ‘anti-tech’ sentiment, but you look around and recognise that we’ve made it easier to build slop than sovereignty, industry and state capacity.

Sovereign Albion looks for better choices. Progress and preservation. Based and woke. Rational and mystical. Metal and moss. Means and meaning.

Episode 1 is with John Fingleton and Mustafa Latif-Aramesh, who were part of the UK’s brilliant recent Taskforce on Nuclear Regulation. When it launched, the report quickly became a rallying point for the many people who refuse to accept that Britain is at some terminal value, both economically and culturally. Nuclear could play an enormous role in enabling safe, clean, abundant energy — solving the energy trilemma of cost, climate and capacity! — but we have made it slow and expensive to build. Britain was the first country in the world to have civil nuclear energy, but today we haven’t built a new nuclear power station for 30 years. I want to understand how we got here, and what needs to come next.

Through these conversations, I hope to straddle both the macro and the micro.

I’d like to go down rabbit holes of cultural and technology romanticism, because I think going harder on culture matters in order to:

1. Raise our ambition

2. Find conspirators & build relationships beyond value exchange, and

3. Coordinate at scale, ie. at the level of a nation

But I also want to understand the turpentine of ambition: unpacking the craft of company-building and policy-making — how big things actually get done — so that many more can go down the paths carved out by others. I’ve done this before (see this interview with Octopus CEO Greg Jackson about building an energy giant), and have many ideas for how to go bigger.

I’m also just excited to start imperfectly and to improve at what Tamara Winter calls ‘deploying your taste’. This is definitely an experiment and there are many reasons why it might not work or scale. But if it does, I hope to build an audience and leverage that distribution to give power to the people, ideas and machines rebuilding our national spirit and restoring national agency. I have zero interest in epiphenomenal ~discourse~, but I do care about building a real culture machine that’s upstream of action. So please share your feedback, small or large, so that together we can make this as good as possible.

Finally, thanks to the Centre for British Progress for supporting this experiment, to British AI company Faculty, for making and open sourcing such a beautiful font, and to everyone who has been generous to have great conversations with me this year. Here’s to many more.

Thanks to Julia Garayo-Willemyns, David Lawrence and Alys Key for their editorial support with this episode, and to Ben Mills and The Subthread for production support. Intro news footage source, ©Sellafield Limited.

Subscribe now

How to Regulate British Nuclear

• John Fingleton is a leading economist and regulator. He was the CEO of the predecessor to the Competition and Markets Authority and was on the board of UK Research and Innovation, and now runs Fingleton.

• Mustafa Latif-Aramesh is a leading infrastructure planning lawyer at TLT LLP and Parliamentary Agent who works across nuclear and other nationally significant planning and infrastructure projects.

Britain was the first country in the world to have civil nuclear, to connect nuclear power to the grid at Calder Hall. But today, we haven’t finished a new nuclear power station in 30 years. What changed? How did we get here?

[00:01:12] John Fingleton: The context in which they were built was different. They weren’t as safe as the ones we’re building now, first of all, very clearly. And the, we’ve, a huge decommissioning legacy coming from that.

Secondly, I think in the post-war construction period and, rebuilding Britain, I think it was easier to get things done faster in that period. You were able to build other infrastructure, housing and so on at the same time. I think we’re now in an environment where, people care much more about health and safety, about the environmental depletion.

Quite a lot of that happened over that period as well. some of that stuff’s not as easy to do as it was then.

[00:01:49] Mustafa Latif-Aramesh: I think the only thing to add to that is, is that, is to underline the point about environmental, targets that we’ve set ourselves. The desire and the goal to reach net zero by 2050 and a kind of enhanced need for low carbon energy generation is something that has changed significantly over the last two, three decades.

Just to double a click on that, there’s a lot of people in this space who would say, we can meet our energy needs purely with renewables and nuclear is expensive and they worry about it for lots of other reasons. And we can come back to how much of that expense is self-imposed. But can you just give us a top line summary of like why, we should still care about nuclear today?

[00:02:34] John Fingleton: The Taskforce’s job was not to make the case for nuclear. The Taskforce job was to say if the government wants to build nuclear, how can we do it more efficiently, more effectively? But of course in the course of that, we had to think about, the relative benefits of nuclear and non-nuclear. The cost of solar has gone down hugely over the last 20 years.

However, we don’t get enough sun in the UK. It’s lovely and sunny in the background now, but you look at the amount of sun period in Spain and other countries, we don’t get enough sun. So solar will always have limitations in the UK relative to other countries. We may be able to pipe solar from Morocco or somewhere in the future, but at the moment that’s expensive.

Wind has not gone down in cost. And we do have more wind, but wind is more correlated geographically. So the intermittency problems are more serious. You can argue that if you build lots of batteries, then you can deal with the intermittency through batteries, but you then end with up with a quite high capital cost, a huge amount of land usage, which has its own environmental, challenges.

And then questions about grid stability and the ability to scale that up at the level like if energy demand doubles in the next 20 years, which it could do, the ability to scale that up, at the level you need. So I think all of that, I think that explains why so many people are advocating nuclear, but I think somebody who’s critical of nuclear could make also quite compelling points as to why at the current cost level nuclear is not particularly attractive either.

So maybe let’s just dive in on that. I was really struck the first time I read about the comparison of nuclear construction costs in South Korea versus In Britain it’s six times more expensive. It’s twice as slow to build. Can you just, before we dive into the Taskforce recommendations, can you just give us a breakdown of like how we have such a large gap in the UK versus South Korea, both in terms of expense and time. What are the key factors at play there?

[00:04:27] Mustafa Latif-Aramesh: The other important thing to note about the report is that it is not just about civil nuclear. So when we’re talking about why is there a need to deal with some of the issues that we’ve identified, it also relates to defence. And so this isn’t just about environmental sustainability and trying to get low carbon energy generation, it’s about national security, it’s about energy security generally.

And even if you were in a position where, you think there are cheaper forms of energy generation that doesn’t forego the need to deal with some of the issues that we’ve identified from a national security perspective On the drivers for, the differences between say South Korea and the UK.

I actually think you almost want to do this in stages. So there’s the difference between South Korea and the UK, as you said, six times the, cost in the UK compared to South Korea, but it’s also places like France, Finland, where we’re two or three times more expensive than very comparable European nations.

And our report, I think, identifies three key issues in the UK, which are probably a substantial part of the explanation for the increased costs. The first is risk aversion, and, we’re very clear in the report not to single out any particular actor, any type of actor. We think it’s a systems problem, but there is risk aversion that flows through in terms of how even developers, regulators, and government reacts to a nuclear project.

The second key, driver is the lack of incentives that exist in the system. We have not just a ratcheting effect of the standards that, are applied, but we also have a. almost a desire to put process over outcomes. And it’s that last point, about process over outcomes, that is, I think, a substantial part of, the key differences in cost.

For those who don’t have the full kind of international context, is that particularly unique or particularly strong in, in the UK?

[00:06:25] Mustafa Latif-Aramesh: Just to give very specific examples, we quote in the report, a study of all EU member states in how they implement the habitats regulations. And what that, evaluation found is that we have one of the strictest and most, stringent applications of the habitats regulations, compared to other European countries.

it does not take as long to produce a, environmental impact assessment, a habitats regulations assessment in many parts of Europe as it does in the UK.

[00:06:57] John Fingleton: I think the other thing about. South Korea is that they have built the same, reactor multiple times in what’s called a fleet approach.

And there’s been a lot of, I think, political consistency, clarity and leadership around we’re going to do this and we’re gonna do it over time. I think one of the lessons from nuclear and from infrastructure projects in this country generally is the sort of stop start nature of it, which does increase the cost.

So in addition to potentially different, environmental or other health and safety standards are difficult to compare across countries, but you do have the, that sort of, you we would be better off with infrastructure if we had a lot of clarity and a lot of drive about what we’re doing and doing it quickly and doing it in a standardized way.

There’s something about the South Korea model that I think is, also important to that, which is the role of the state as a key financer and deliverer of, these fleet infrastructure projects. Obviously the state is still involved in the UK to some degree. It bears lots of the kind of, financial risk, but is not necessarily involved in quite the same way in terms of delivery. How important is the role of the state in fleet delivery? do we, can we deliver fleets at scale privately?

[00:08:10] Mustafa Latif-Aramesh: I think there are two aspects to that. The first is it’s more about the state setting a framework that will allow a fleet approach to be adopted. Under our current rules, regulations, practices, we have almost a, an instinct to treat each application afresh and have very bespoke consideration of each particular project, our planning policies, our environmental regulations, even the way that we implement the application of particular safety standards.

We’re treating each thing anew. so there is a role in the state in setting a framework that says. Actually, we really believe in fleets. And just to John’s point about start, stop, the example that I’ve heard Julia Pike from Sizewell C mention quite a few times is, because we hadn’t built a nuclear reactor for so long in this country, EDF had to fund their own welders academy.

and, we’re at risk if, of substantial works that sizeable don’t start that we’ll have to reinvent the wheel again. On the second aspect of the question, which is more related to the South Korea dimension, you are right that the, the head of state that the Prime Minister’s office in South Korea takes a very hands-on role in terms of nuclear regulation and nuclear delivery.

I think the issues that we have in this country aren’t only solvable by asking for kind of complete state delivery and complete state control over nuclear development. The comparisons really are other European countries, the US and I think one of the aspects that we, we have to be quite open about is what is, what are the constraints that the UK has that South Korea does not have.

And there is an element of public finances. What is the effect of additional investment? This government has invested quite substantially in SMRs with the announcement of the Rolls Royce project in Wylfa. And so it’s not a case of zero or all out. And so it’s just finding that balance, I think.

I want to understand the unit economics of nuclear construction today and then what that would look like, going forwards. And so I think, the Hinkley point C Contract for Difference (CfD) is £127 per MWh. Sizeable C is financed differently under a Regulated Asset Base model that guarantees investors a return to fund new nuclear. Can you give us a breakdown of, where the main cost drivers are in, in a major nuclear project today and what that would look like, going forwards under, under these implemented recommendations?

[00:10:45] John Fingleton: I think that’s quite difficult to do.

We looked really at the regulatory aspects, but we didn’t get into all of the other engineering costs and so on. So I think we don’t, we haven’t in the report tried to do that. What we have done, as Mustafa has indicated, we try to look at, reasonable comparable countries like France, Finland, Canada, the United States, and if you look at the, their cost base, they shouldn’t have a different cost base than we do in terms of regulatory standards and so on.

People have a high standard of living in those countries, similar systems, democratic government. And in that environment, they’re still able to do it substantially cheaper than we are. And I think we make the point in the report or the announcement around it that if you take the French or the Finnish costs, we could bring that price in the country of difference down to £80 or £100 pounds.

Rather than, the £137, whatever number you said it’s at, the moment. So there’s a substantial benefit there. But I wouldn’t stop at that. I think we should be looking at all those elements of costs. But some of them are quite difficult, and, post 9/11, we develop requirements to put concrete shields on nuclear power stations so that airplanes can crash into them.

That’s expensive. So some of these things are just naturally quite expensive. Whether you will always need that with every type of reactor is another question. And some of the, a good deal of the report isn’t just focused on Hinkley and Sizewell, but, or even the SMRs, but thinking about future technologies that are currently, they’re more than PowerPoint presentations, but they’re not a lot more than PowerPoint presentations.

Yeah. But actually they’re, they, point to quite exciting technological developments coming out of, universities and, industry that could enable nuclear to play a much broader and pervasive role in, in generating energy and heat in more environments.

What sort of, can you give us a sense of what sort of things you have in mind now?

[00:12:42] John Fingleton: We refer to maritime. So the idea is we currently power the merchant shipping fleet and cruise liners and so on with, fossil fuels. But you could imagine a world in which we developed a safe, a small, safe reactor that could do that and decarbonize the shipping fleet.

You could imagine building small nuclear plants on barges, which gets rid of a lot of the geological costs that we have at the moment, but also might enable you to relocate them if you need a power in different places. I think Russia has done three barges now with nuclear plants on them.

And then I think, thinking about advanced, modular reactors that serve towns or areas of towns and do combine even power something on a much smaller scale. And when you, I think I read that there are 127 small modular reactor designs in the world at the moment, and then you go into advanced modular reactors, there’s a lot of stuff there and a lot of it will be nonsense, but some of it will be potentially very powerful. So we, what we’ve tried to do is to future proof the regulatory regime so that as those come along, our system may be better able to deal with it.

Maybe we can spend a bit of time on the specific recommendations. I just want to hear from maybe each of you, which you think the most important is of the 47 recommendations that are in the report, what’s your favourite child?

[00:14:04] Mustafa Latif-Aramesh: Yeah, it’s a difficult choice if there are 47 children. But, if I’m allowed, I’ll give probably four and I’ll be brief. One of them is the government defining the tolerability of risk. And just for context, the tolerability of risk, at the moment, is a document that is established by, initially the government, now HSE (Health & Safety Executive), which defines what society considers to be a tolerable level of risk.

And the real issue with that, in a system which has some risk aversion and a slight ratcheting effect, is that you will always be driven down to going beyond what is strictly necessary to ensure a safe, nuclear development. And what the recommendation does, is to say, government should take control of this.

It should explicitly define what is a definition of risk that society thinks is worth the development of new nuclear so that we can be much, clearer about what parameters we’re using. The second is the idea for a commission. So we call it the Nuclear Regulatory Commission. And this is effectively a body that will have the powers of all the other regulators.

And it’s intended to be both used as a kind of final arbiter and escalation process to determine quite knotty issues, but also can be used to directly decide things like nuclear site license applications and planning applications. And the thinking behind that is making sure that there is some element of a goal, a duty, we call it a secondary duty, of delivering new nuclear safely and quickly. And I think that has the potential to really change how these decisions are made.

Can I just come in on that? So you previously talked about, you don’t want to point fingers at anyone in particular. There’s a sort of wider systemic problem. That particular commission is designed to accelerate the delivery of nuclear, recognising that there is there is a systemic problem that just slows things down. Can you just give us a bit of a sense of, without necessarily pointing fingers, like what are the constraints that, or what’s stopping the existing regulatory state in nuclear actually getting new nuclear built quickly?

[00:16:18] John Fingleton: On the specifics of the permissioning, you have a number of problems. One is that you often have junior people inside a regulator who will take a view that’s very conservative. The company won’t want to challenge that because they might get a reputation within the regulator of being less safe.

And so they’ll then build something that’s quite goldplated. It is systemic because on the other side, they may not have strong financial incentives anyway. And so therefore the sort of cost benefit to them of challenging it is also, maybe not optimal.

Can you just say more about their financial incentives?

[00:16:54] John Fingleton: Well if they’re, in a lot of cases, so for example, in all, almost all of defence, not all of it, but almost all of the defence, the suppliers are on cost plus contracts, so they just pass cost onto the taxpayer. But actually the arrangements for Hinkley and Sizewell. It’s, there, there is a substantial part of the marginal cost or the additional cost that’s borne by the taxpayer or the consumer at the end of the day.

And it’s very difficult to have or to simulate competition in this environment, pure competition in these environments. But if it’s a supermarket that’s doing something that they’re gonna be very keen to do it at the lowest possible cost. And they’re gonna challenge the regulations up to the boundary and they’ll say, where is the boundary?

And so you get very precise lines, through the process of regulators engaging with competitive businesses. When you don’t have competitive businesses, you don’t necessarily get that. So part of the problem is the culture within the industry, which is to goldplate things and not challenge things back.

We’ve proposed a number of things in the report around escalation, around internal challenge function within the regulators to try and address some of those issues. And the idea behind the commission is that you can take these issues to the commission and get a quick answer, and in a way that’s quite transparent.

That would create precedent for other people and so on. But what we want is to get closer to the regulatory boundary rather than the current situation where everybody is like way under it.

And would that also, and I do wanna come back to the other two [Mustafa], but, I’m just interested in this interplay of, the regulatory incentives that we create and then also the sort of what that means in terms of how expensive these projects end up being. And so if, we’re effectively almost paying twice by both delaying the delivery of projects and then actually being on the hook for the extra costs that, that are introduced as part of that delay, if the recommendations were speed up delivery, do you have views on how we fix the financing side as well?

[00:18:46] John Fingleton: First, if we speed up delivery, we massively reduce the cost because a huge amount of the cost is the interest on the capital and a big part of the cost. But a lot of this is actually about taking cost out. So if you require something to be done that is vastly out of proportion to the benefit that’s being achieved from it, and you take that out, that’s really important. And, we talk about the fish deterrent, but there are many other examples and there’s examples of nuclear safety as well, where, a company is asked to do something, it’s gonna cost £94 million, it’ll save one life every 50 years, and the company doesn’t maybe, in this particular example, the company said, we don’t agree with that. We’re gonna challenge it. They challenge it. The regulator said, yeah, that’s fine. But they might not have challenged it. And in other circumstances they don’t challenge it.

The difficulty is, Andrew, it’s like many mole hills make a mountain. There isn’t, there’s not one that has necessarily cost £4 billion. It’s there’s one that cost £94 million here, there’s ones that cost £50 million here, £20 million here. I remember when I finished my PhD, my supervisor said my thesis was too long, 20% too long.

And I said, should I drop a chapter? And he said, no, every sentence is 20% too long. So it’s a bit like every individual component of it is overdone, rather than there being just one thing you can strip out. And so when Mustafa talks about the tolerability of risk that affects everything.

It affects the working conditions in the plant. It affects a whole lot of the operational, the operation of the existing plants, the construction of the new plants. And so bringing proportionality into all of these individual micro decisions will mean savings across a whole range of small areas that add up to a large number.

[00:20:34] Mustafa Latif-Aramesh: Just to give two very specific examples of that. They range from very small things which have a substantial cost to very significant things which have an even greater cost. An example of the former is, someone will, as part of the design development process, say the particular screwdriver that we’re going to use is X millimeters.

Turns out nobody in the world makes screwdrivers of x millimeters. So you end up getting a factory to bespoke manufacture these screwdrivers so that it fits the design that someone just, you know, at some point, said is the thing that we’re going to use. An example of a more kind of pervasive issue is in the decommissioning context.

So the report outlines billions of pounds of savings that could be made through changes to the regulatory regime. A really good example of that is the Winfrith site, which, you know, somewhere in the south of England where they are in the process of doing something that is actively good for the environment by decommissioning the site, making sure that things are disposed of, responsibly.

But the current, kind of framework forces them rather than to keep some of the rubble on site to potentially take it offsite by using 10,000 HGVs. Trucks going on the road network. That is not good for the environment, that is not good for the overall delivery of the decommissioning objective of this government and virtually every other government.

So you’ve got these very, what seem like small decisions on what screw are you using all the way up to actually we might need to use 10,000 vehicles, when really you could just keep it on site. There is a good segue onto the third one I was going to mention. So there are some eye catching recommendations on, the environmental side, but one of the ones that, I don’t think has gotten as much attention is a requirement to report where a mitigation measure is more than £0.5 million. And I think this actually relates to a number of things that John has mentioned. So we have many case studies which highlight the costs of some of these measures, but there’s no requirement to report exactly how much they cost.

So the Fish Protection System costing tens of millions, hundreds of millions of pounds was not something that was widely known. And so the recommendation is, be upfront, say when something is going to cost more than £0.5 million, so that at least the decision makers, and actually the developers, you know what you’re getting yourself into.

And you’re being open about it because a lot of these costs are very difficult to find. So that’s one of the kind of really good recommendations that I hope is implemented and people start to use that as a reason. “Oh gosh, are we actually spending £1 million, £2 million, £5 million on this?”

That doesn’t seem right. And then the last one I was gonna mention is, actually also slightly related to the overall kind of framework we have for the delivery of nuclear.

As it currently stands, we have something called Semi-Urban Population Density Criteria, which prevents nuclear reactors being located near urban population centers as the name suggests.

The methodology that is used to determine whether that test is complied with or not,

1. is from the 1960s

2. is based on an old AGR

3. does not take into account topography

4. does not take into account enhanced safety features or decay rates.

It is quite frankly a really unreliable test and it is being used to determine where we put our reactors. That’s really bad. Not just because it’s a bad methodology, but also some of our existing sites which have very welcoming local communities to nuclear are at risk of not complying with this criteria.

And so the recommendation is that this really needs to be revised so that we can have a system that enables new nuclear reactors in places without affecting safety. We have regimes in place to protect people.

So just to make that clear, there are current places in the country where we could not build new nuclear potentially, that currently have nuclear or are already welcoming?

[00:24:30] John Fingleton: So Heysham is a good example where we’ve allowed housing to be built close to the nuclear power station. So it’s created this asymmetry that now you wouldn’t be allowed to build a nuclear power station under the Semi-Urban Population Density. Even though the new one would be safer than the one that’s been being replaced. So that’s the sort of perverse outcome you get from that.

How quickly do you think the recommendations that you set out could be implemented and how quickly could they be showing an impact?

[00:25:00] Mustafa Latif-Aramesh: On the day of the publication of the report, the Prime Minister also published a strategic steer to the industry. And that says that the government is committed to the quote, “complete implementation of the report by the end of 2027”. That date isn’t picked out of thin air in the report.

In each recommendation, we have a target date for each of the recommendations. The final date that we had was December 2027. And, now that, the Prime Minister has clearly committed to making sure that there is an implementation of the recommendations by then. If you’re asking specifics on how long does it take, it’s fundamentally driven by, can something be solved through guidance? Can something be solved through secondary legislation, or do you need primary legislation? And it’s the primary legislation that is driving the December 2027 date.

From that December 2027 date, can you give us a sense of how this would change the culture and behaviors in the industry?

[00:25:57] John Fingleton: I, think from right now, I think it is, there’s enough a debate about this, that within existing construction at Sizewell and other places, I think there’s probably greater openness for our ability for companies to challenge back and ask questions about stuff. So I would like to think that they, that actually from today, you have a change in some of the debate about where the boundaries should be on some of these issues. Number one.

Number two, on the tolerability of risk that Mustafa mentioned, which we’ve said the government should do urgently. Something like that could have a good, a big effect on the cost of existing operations and new build very, very quickly.

Thirdly, the next project we’re going to start is the, the Rolls-Royce SMRs at Anglesey in Wylfa in Wales.

That’s a really good test case of, can we do this at a substantial lower cost than we’ve done it before? So I would say that, obviously there’s a lead time into doing that. There’s various processes to go through, but we should be trying to do that really quickly. for example, the commission idea requires primary legislation to give effect to it.

But what we’ve suggested is in the meantime, the regulators get together and develop a lead regulator model that begins to implement what that would look like in practice without the statutory provisions. And so that’s something I would expect to be applied with immediate effect from, from now. So quite a lot of this can happen. And even if you know that, something requires a statute, the fact that it is going to happen begins to gear things up towards doing that.

The other point I would make, and it’s a more general point about supply side reform, that people always say, and it could be the expansion of Heathrow or the Oxford Cambridge corridor or whatever else, “Oh, it’s gonna be years before it’s built.”

But actually we know as with HS2, that the moment there’s a commitment to doing something, the property values begin to change. Investment begins to flow in because if you’re a company and you’re thinking of investing in the UK, you’re going to build a plant, you’re going have to employ people, train people, etc.

And that’s gonna take many years. What you want to know is there’s a commitment to having the infrastructure that you need to be delivered by that time. So actually it crowds another investment earlier on than you might otherwise think. So I think it’s a mistake to think with supply side reform that the benefits only come when the project is built.

I want to, go inside the Taskforce process a little bit. If, you’ll go with me. The Taskforce was effectively a six month process. What does the sort of chronology of that process look like?

[00:28:29] Mustafa Latif-Aramesh: So I think it’s worth saying the very kind of high level summary of what we did, and then we can go into some of the details of it.

So we were established and immediately we, fairly immediately, we put out a call for evidence. And the call for evidence was relatively high level saying, asking people, what do you think works well? What do you think doesn’t work well? And we had consultation responses. We then considered those in tandem with.

Various site visits, various discussions with, regulators, government departments, environmental groups. And on the back of that, we published an interim report. That interim report started to drill down into very specific questions. Based on a initial understanding of what we thought the issues were and what the potential solutions might be.

We certainly hadn’t reached a kind of firm view at that point. We wanted to make sure that we were doing a, carrying out a process that was start high level, get more specific. We then had an encouraging amount of responses to the interim report that was published in August. And at that point we started to get down into a level of detail, which was about what are the specific solutions, what are the specific delivery vehicles for dealing with specific issues?

And we had a a series of workshops. Not just with government departments, again with regulators, environmental groups, developers, new build, decommissioning. And we, essentially just started floating ideas, asking provocative questions when we think, when we thought people weren’t going far enough to see exactly what the limit was.

And that’s what led to the final report. There is a brief few weeks between where we landed and the publication of the final report, which was based on testing the recommendations with, again, various government departments and people just to see.

And regulators.

And regulators just to see exactly where

Tell us about that, the Taskforce. Obviously it’s an advisory body, it doesn’t have formal powers. To see implementation, you need both political will, but also for departments and regulators and all of the bodies that you’ve just spoken to, to be the delivery arms of this going forwards.

How do you think about when you’re doing this taskforce, how do you think about allocating your time between getting to the ideal right answer and then also working with other agencies, organizations, bodies in the state to understand what their constraints are, what they’re gonna do, getting them bought in, and being the sort of, vanguard of this going forwards?

[00:31:05] John Fingleton: We had a, civil service team about 10 people, and they did a huge amount of work in the background talking to other government departments about things and then surfacing to us areas where we needed to get involved with those departments on specific issues like, I don’t know how many meetings we had about the Semi-Urban Population Density Criteria, because I think we had some meetings and then we thought, let’s just scrap it. And then we were saying, scrapping it is, could create even more uncertainty for developers in the short term. So let’s leave it there, but create an alternative pathway, which is what we came up with.

So actually, quite a lot of those refinements benefited from interaction with people in government departments who said to us, have you thought about the implication of doing something that dramatic? And we thought, actually, maybe we’ll refine this. So they did get refined as a result of this process.

But I think essentially, I would say, it took probably seven months in total. There was probably two months of initial scoping of issues, visiting sites, finding out stuff. Then there was two months of more in depth discussion with people, and then there was two months of finalizing recommendations with some gaps in between, because some of these things take a little bit of time.

I think we, we told people that what they told us was not going to be attributed, so that was important as well because I think, again, goes back to this point about outcome over, over process. You could have run a process where you took evidence from everybody and reported back what everybody said and so on.

And I think some people who might disagree with us think, oh, you should have published everything. Everybody said, actually, we need to get something done quickly and efficiently here. And I think we wanted to demonstrate with the process that nuclear itself might be done in this way, that you actually just get on and do things and work out, like what’s the answer?

And try not to agonise too much about, every, every single detail supporting that.

There were five of you on the membership of the Taskforce. How do you think about splitting up and allocating responsibilities between you? Who does what? Who holds the pen?

[00:33:04] John Fingleton: I guess that happened, maybe not quite top down, but I think I did ask Taskforce members to take the lead on particular things. Each person led on particular chapters and particular themes and, when we had the workshops, different Taskforce members chaired the meetings.

My general style is very distributed. I don’t need to, if I’m at a meeting, I don’t need to chair every meeting. So Mustafa chaired lots of meetings that I was at and other people, and that just spread the load. So actually, I would say as chair, I, I probably did, much less than a chair might normally do because we spread it out across the Taskforce members very effectively.

And I think it was also, it was interesting when we had our very first meeting. Everybody on the Taskforce was up for quite radical change and really bought into the idea of that. But, Mark Bassett had done a lot of work internationally, really understood all of the international aspects and led on the international.

Andrew Sherry understands culture very well, but also has deep expertise on defence nuclear. And then Sue Ion has been in this sector for a very long time, has a very distinguished record. Also, had been on the board of AWE, the Atomic Weapons Establishment, so knew defence as well as civil, and is one of people credited with getting Tony Blair to change his mind about nuclear 20 years ago.

She had deep knowledge across the entire sector and then Mustafa has a great deal of experience in, in planning and environment in other areas. And then I understood regulation, but knew nothing about nuclear. So there was quite a nice debate between us because we all brought very different expertise, but we were able to, I think, respect each other’s expertise.

How does a Taskforce membership body like that get pulled together in first instance?

[00:34:53] John Fingleton: Oh, that was, to be perfectly honest, that was perhaps one of the more frustrating aspects. Because I was asked to do this in early February, but then it didn’t get announced until May, until mid…well, my appointment was announced in mid-May, but the rest of the Taskforce members wasn’t announced until late May. I was early April. Then the rest came in mid-May. So although the Taskforce formally began on around the 10th of April. Actually, the other Taskforce members didn’t get appointed until, I think like the 28th of May or something. Yeah.

So I interpreted the six months is when they got appointed because we needed that. We had begun the, I’d begun the consultation process before the others got appointed. But that, I would say if the government was doing this again, just getting on and getting everything up and running slightly faster would be, could have been done, could have gone us there a month earlier.

But, it worked well once it got appointed and started.

If you were starting over, what else would you do differently?

[00:35:50] John Fingleton: I would have liked to have had more, more economics in it. You asked the question about cost. More economics and more finance, and we didn’t get that work stream up and going until rather late in the project for complicated, actually probably quite boring bureaucratic reasons, but it would’ve been nice to have had a constant economic work stream throughout it.

I think it would’ve given us more numbers to base things on. We held off putting them into a lot of the numbers into the reports simply because I think if you’re going to put numbers into a report, you need to be sure about them.

But I think that’s some work that’s work that the treasury and others could do now, particularly to demonstrate to the OBR, the Office of Budget Responsibility, what the productivity benefits might be, what the cost savings on decommissioning might be. We’ve given some rough indications.

I would’ve liked to have been able to calibrate what the extra funding into nature conservation restoration projects would’ve been, because I think it’ll be substantially greater than what it is now.

But it’s quite difficult to do that in a short timeframe, and it would’ve required an extra layer of consultation. So that would probably the thing I would’ve probably liked to have done more.

[00:37:05] Mustafa Latif-Aramesh: I’m going to be slightly diplomatic and answer your question by saying what I think went really well. And

That’s so boring!

Mustafa Latif-Aramesh: there, there is a, point in it, which I think is helpful for the underlying principle of your question, which is one of the things that is now in the public domain is a letter that is, is sent by the ONR, but it has a number of co-signatories across the nuclear chairs.

So we’re talking Nuclear Decommissioning Authority, the Environment Agency, EDF. That letter says quite clearly, we welcome this report. We are keen on the implementation of this report to help drive the industry forward, having effectively the nuclear establishment, and the key regulators saying that they endorse a report that is quite radical on a number of different measures, I think is a significant achievement because it shows how much desire there is to change things.

And so for future government reviews and reports, I think if you want to get in a space where the Prime Minister is saying we accept the recommendations, or the steer is saying we’re committed to complete implementation by the end of 27, you cannot get there unless you have done the legwork.

And I think there’s always room for more, legwork in building up that support. So if we had an additional month, we could have got even more. But where we have landed is a position where the ONR is, saying very clearly, we welcome this.

[00:38:40] John Fingleton: Although if we had another month, I worry, I do worry that, a lot of people will began to, because, this is, this requires a good deal of effort and ingenuity and actually challenge around preconceived ideas to implement.

And I worry, I think there was a right moment in time just before the Budget to get this out there because I worry that people always want to pare away things. And if we had another month, more might have got, we might have got more pushback. I was quite keen to get it out there and to get political commitment to it, because my preferred approach to do that is to say we’re doing that from a position of strong political leadership.

Is the deep state real?

[00:39:27] John Fingleton: No, I mean in the, in a letter I wrote at the start of the report, I do make the point that it’s an awful lot easier in our system for people to say no.

And that very few people have the ability or the incentive to say yes. And there’s a lot of, there’s a lot of risk aversion.

Parliamentary committees perform an important role, but they do tend to punish people who’ve taken calculated risk-based decisions. So you get a lot of risk adversity in decision making. So there’s, that’s one problem.

The second problem is that every government department has its own priorities. And I think that getting big reform that requires, and this touches on 5, 6, 7 departments of government at the same time to put effort into prioritising the same thing.

That is not something that’s easily done, and only the Treasury and the Prime Minister can really push that and they have to do it together. So having the Chancellor in the Budget and the Prime Minister in the strategic steer and his speech on Monday, have them come out so clearly, positively, in favor of doing this, I think does send a really clear signal to those government departments and to others. So I don’t, no, I don’t think it’s about the deep state. I think it is. And I think if you talk to civil servants leading in these departments, they feel they need strong particular leadership in order to give them the ability to move forward.

So you get this chicken and egg thing because the politicians will say, oh, the civil servants aren’t delivering this as fast as I want, but the civil servant, the politicians are not covering my ass on the risk. And so you want to get that positive, virtuous circle where you have political appetite for change.

Political risk appetite and political backing. And then I think the civil servants are prepared to deliver that at pace.

For what is worth, I’m not sure in my mind that the deep state is actually some like grand hidden conspiracy. It’s more sort of emergent property of splitting the difference before the decision is actually really taken, and this kind of compounding over time. Risk aversion. Everyone’s wanting to find the compromise rather than necessarily like just starting from the high ambition thing and then following that through. And if everyone is doing that at scale across many different departments, it’s very hard to retain the sort of like original high ambition, stretching level of ambition if sort of every layer of sort of process someone is going, as you said, like just shaving it off, shaving it off, shaving it off. And they, often are doing it in my sense from for what they believe to be very well intentioned kind of reasons. But the cumulative effect is that you, end up in a very different place from where you started, I think.

[00:42:03] Mustafa Latif-Aramesh: It’s a bit strange to be quoting John when he’s sat next to me. But in the forward that he mentioned, there is a line that says “you will be encouraged to consider consult and dilute. I encourage you to resist.”

And I think that’s what you’re getting at. There is, I certainly don’t believe in some grand conspiracy to try to water things down, but there is a process that everything goes through. And so as John said, the reason for being quite keen on the very full-throated endorsement from the Prime Minister, and the Chancellor was to start from the position of this is about the detail of implementation now. The principle has been accepted. So now we are working towards the detailed solution. We’re not going to undo everything. And I think, and anticipate that’s how most of the people that we’ve engaged with will, will see those see those endorsements.

Can you just say something from the joint experience you’ve had with this Taskforce, it feels like there’s a sort of emerging taskforce theory of everything where if the government really cares about getting something done, it will create a taskforce. On vaccines, AI safety, nuclear regulation. How scalable is this model to every other sort of class of problem that, that we have in government?

[00:43:19] John Fingleton: I think there are other ways of doing this. I sit on a advisory group that advises the Australian Treasurer so the equivalent, their Chancellor. And it’s a very different mechanism, but it’s actually produced quite a lot of substantial regulatory reform in Australia.

And what they do is the Australian Treasury brings us papers and proposals. And there’s seven of us. And we sit there and it’s been going for two years now, and we give them advice on it. But actually once something then has gone through that process, it has a lot more credibility within the system and then it goes out for public consultation and so on.

So there are other ways of doing it. I think the key is to implementation. So we model the implementation recommendations on something called the Haddon-Cave report, which is an investigation into, an airplane crash in Afghanistan about 15, 18 years ago that showed up substantial failings in the safety regime for, for aircraft.

And what Haddon-Cave did in his review was he actually, his implementation group was established before he reported, but the implementation group tracked the recommendations in a very precise way and it all got done. And so we were trying to model it on that. So as Mustafa said earlier, the, the, they, each recommendation has an owner and has a date.

And we’ve said there should be an independent implementation group that sort of holds the government to account on that. And part of that is about doing that.

The second observation I would make, and I think this happens in the private sector as well as in the public sector, there is a big difference between asking somebody whether, it’s in the business, the general counsel and others, “what can we do about X?” and saying “I want to do this, tell me how to do it.” And I think very often you get a different answer.

One of the consistent themes throughout your work has been that the sort of broken regulatory state can encourage rent seeking, it can encourage entire industries or people to go into sort of unproductive activity. Can you just illustrate a little bit about how that works?

[00:45:24] John Fingleton: I’m sitting beside an excellent lawyer and many of my friends and my colleagues are lawyers. But there’s a wonderful paper, in the American Economic Review in 1993, Murphy, Shleifer, Vishny, which basically looks at economic growth rates and regressed it against the ratio of engineers to lawyers in the economy and finds that the more, the higher the ratio of engineers to lawyers, the better the growth rate.

And you look at countries like China and others that are run by engineers, and you look at countries like Ireland and the UK that are often run by lawyers, and we have a different approach to doing these things. That’s slightly tongue in cheek. But I, do think one of the important measures of, being able to do economic growth is actually a focus on building things and getting things done rather than the focus on having debates about building things and getting things done, is really, important. And in this country we rely too much on legal process to sort out problems rather than trying to agree things ex ante so a lot of the issues we’ve seen in this sector about delaying, delaying big projects because of debates about issues that are not actually central to the project, is really, important.

And that’s not to say that we don’t need excellent lawyers to sort out these things. But I just see we, we end up often with just way too much litigation. Way too much rent seeking. Way too much argumentation about dividing the pie rather than argumentation about how to increase the pie.

[00:46:50] Mustafa Latif-Aramesh: I think just to give a specific example of that, and then I’ll deal with the abolished lawyers point, but the specific example of that is something called regulatory justification. This is an approval that you have to obtain, in the very early stages of a nuclear project where you ask DEFRA if the benefits of your technology outweigh the disbenefits to human health.

Now immediately alarm bells should start ringing because that question is dealt with in planning. We have a whole site licensing process that is focused on safety. We have dose limits. So this additional approval is something that is highly duplicative of other aspects of our process. The government, probably three weeks before we published our report, issued a notification that it had awarded a tender which cost £8.7 million for I think six or seven consultants to help it administer the regulatory justification process. That is obscene, because this is not a process that is required.

If the question is, does nuclear technology, the benefits of it outweigh the harm to human health? The answer has already been given by government. So the fact we are awarding contracts of millions of pounds is an example of, I think a process in which, this kind of behavior that John is talking about is allowed to flourish. From my personal perspective, I think lawyers should be less involved in these kinds of processes because they are fundamentally planning judgments. They are safety judgments. They are not that exist in a legal framework. They shouldn’t be legal decisions. But, if I was speaking to other lawyers, the argument I present is an economy in which we have greater amounts of development means greater amounts of work.

So even though you are less involved, there is more of a, a portfolio of projects to be involved in. John’s 48th recommendation is abolishing lawyers. But, there is a serious point that there is scope.

[00:48:52] John Fingleton: One of the, one of the issues we deal with in the report is the Aarhus convention, but we list the Judicial Reviews that have happened in on, the two big nuclear plants.

And by and large, they’ve all failed. They’ve all been found, almost a lot of them to have not been either without merit or not having great merit. They’ve all been designed to delay and slow down something. And to, to my mind, there’s been just huge wasted activity, both in terms of the cost of stopping everything in the supply chain, but also just the cost of going to court.

And you sort of think there must be a better way of doing that. And so a lot of the gist of the report is how can we try to get to the answer we’re going to get to in a faster, more efficient way? And so I would think, for example, with Sizewell C, there’d been a number of, of cases that have delayed or increased the cost of the plant.

But in actual fact, if you’re living nearby, I can see, if I lived near Sizewell C, I would be frustrated by the cost of the construction, the trucks on the roads, everything else. But actually a lot of these challenges just drag that out. So instead of getting on with it, doing it quickly,

You get on with it and do it slowly. And it’s like ripping off a plaster really slowly.

So if you live nearby, it just means you have the construction for longer, it’s down, it’s back, etc.. And that’s often may even be not in the interest of the people bringing some of these cases that, that, they end up harming themselves or damaging their own environmental, environment locally more than any benefit they get for anybody else.

I think there’s actually a really important point here around, we’re in the worst of all worlds of sort of weak state capacity and zombified private operators who can’t move quickly, don’t have freedom to move at pace. And at a time when you have a sort of, there’s this crisis of liberalism and on the one hand you have people saying, all corporate actors are bad.

On the other hand, you have a, version of ‘everything-is-fineism’. There’s clearly such an enormous gap in the middle for a critique that drives a wedge between rent seekers who kind of poison the well for everyone else and the productive capitalists who are, potentially founding companies and building way out, out of malaise.

And it feels like this is an agenda that is is core to that. If anything we’ve almost made it easier to build slop and sovereignty and industry because it’s so hard to build these projects.

[00:51:19] Mustafa Latif-Aramesh: There, there is a line in the report, and it’s in a section called Our Philosophy, and it says there is enough blame to go around.

And I, again, just to emphasise the point, we’re not trying to pick out any particular actor in this entire framework and say, you are responsible for this. What we’re saying is there’s a systems problem. And I think that’s the gist of, or the thrust of what you are saying, that actually that there are a number of different components.

The risk aversion aspect is not just something that is directed towards elements of the state. It is something that is deeply embedded in our current duty holders, the developers of nuclear development, of our existing reactors and future build. The other aspect to this, I think is it is worth noting how thin some of the initial resistance is.

So as an example, I mentioned regulatory justification. One of the, immediate responses to that is actually there’s an international convention that says we need to do justification and then optimisation. And then you’re thinking, okay, that, that sort of makes sense. How does everyone else in the whole world interpret this?

Turns out we are an outlier. We have, the ability to follow rules in a way, that is completely different from everyone else and say that this is the correct interpretation. So whether it’s that convention or another one, I think there is a serious issue with us giving weight to quite thin, thin resistance.

And part of the report is saying that’s not just one person saying that in a room. It is all of us acting in a way that is highly risk averse. And I think that is a much more fundamental issue than anything else.

[00:52:56] John Fingleton: I think it’s incredibly important that people have access to justice, that regulators are held to account that there is, there are appeals mechanisms that can do that.

I think it’s when you see that the appeals that are brought generally don’t succeed, you sort of think, maybe the system isn’t working as well as it should. I think in other areas of regulation, I see judicial reviews that are actually correcting quite important errors that regulators have made.

I also see judicial reviews in other areas where, actually, the consultation process hasn’t been well run. And so it hasn’t taken account of the effects on people. But I think it’s when you’ve had an excellent consultation process being well run and the regulators follow the process, and then people bring judicial reviews simply because they don’t like the outcome.

And you can understand that there’s a negative impact on them. I also think as a society, government could be better at just trying to compensate some of the losers on some of these things.

I, I, first got into regulation 30 years ago on recommending deregulation of taxis in Ireland.

But actually I built into it a model of compensation, and I think

How did that work?

It basically said that, every taxi driver should be given a second medallion and allowed to sell it. And the medallions, the scarcity of the medallions was what was holding up the price and restricting supply.

And so it doubled it, it doubled the number of taxis overnight and brought down the price of the medallions. But, it meant there was an element of compensation in there.

But, if I look at, other areas of reform, so completely outside of this, but you look at, for example, something like stamp duty reform.

The government finds it very difficult to move from stamp duty to some sort of land value or property tax. There’s a very nice paper by the Centre for British Progress, which I’m on the advisory board of, that actually goes into the transitioning of this, how you would do this in a way that minimises the impact on the losers and gets you to a different equilibrium.

And I think just generally in public policy, thinking about how you move forward in a way that that avoids a large class of people who lose out from a reform in the short term, even if the beneficiaries are far greater, we need to get better at doing that. So for example, in the context of energy in France, they give people who live near power facilities cheaper power.

Reflects the disruption.

We do a bit of that. So for example, there’s been quite a lot of work done by EDF and Sizewell on, with the local communities, and if there’s a power station or a nuclear station, but any big facility, even a wind farm or whatever being built nearby, you’re going to have disruption to local people.

And it’s, I don’t think it’s enough just to say, “oh, grin and bear it”, always. Sometimes having some compensation mechanism can massively improve the efficiency of the delivery.

[00:55:54] Mustafa Latif-Aramesh: I think just on that point to, to give a, another kind of case study of that, there is a famous Supreme Court case in this country, which relates to a wind project in the Forest of Dean. And in that case, the developer was trying to do the right thing by saying we will give a local community package. They were giving essentially some discount on local bills, in exchange for the kind of concerns that John’s highlighted. The Supreme Court ruled that is unlawful.

And so one of our recommendations is actually about, we should be legalising fully, to the fullest extent possible, community benefits so that you can actually compensate people for some of the losses that they’re feeling. The current rules say the payment of that kind of community benefit cannot be considered in planning decisions. That seems wrong. That’s a disincentive to the kind of compensatory win-win solutions that we want to see that will reduce the overall amount of objection to new development.

I also think the fact that case had to go all the way to the Supreme Court. It’s quite telling only for the developer who’s trying to do the right thing to be told you are doing something unlawful.

I think there’s a wider point here around, there’s some brilliant writing by Aled Maclean-Jones on the different role that nuclear plays in the imaginary in France and in Britain, and how it is a sort of real source of civic pride in France in a way that maybe we haven’t quite been able to yet build it in, in, in Britain.

So much of this is about, detailed regulatory recommendations, that’s absolutely understandable, but really what’s upstream of that is like building the sort of culture of consent for nuclear and a sense of civic pride that definitely exists in other countries.

And one of the reasons I’m so excited by this report is it is the plan for building way more nuclear over the course of the next few decades than we’ve ever done before and doing so safely, doing so cheaply, doing so at scale.

And I just really appreciate you guys taking the time to, to share more about how this has worked and hopefully we’ll see it implemented in full going forwards.

[00:57:54] John Fingleton: Thank you very much, Andrew.

[00:57:55] Mustafa Latif-Aramesh: Thank you.

Good news: dunking on Europe is becoming low status and a growing number of founders and investors are building incredible companies and capabilities to secure Europe’s future.

Bad news: while founder energy is critical to saving Europe, only a handful of people are bringing this intensity to fixing European policy.

The lesson from EU Inc, a necessary but narrow proposal that’s still been hard to get done, is that progress on any of the bigger issues — capital markets, AI, defence, energy, internal trade, labour market etc. — will require a much bigger effort. It’s hard to win these arguments at the best of times, let alone when they require countries to pool their sovereignty.

This means changing tack. Too often, too many VCs and founders complain about Europe’s ‘fragmentation’. But Europe’s starting point isn’t a fragmented, single entity so much as a fragile, partial and incremental union between independent countries with their own histories, priorities, cultures and languages. Nor is Europe just the EU: Ukraine, Britain, Switzerland and the EEA countries are all part of this mission.

This has two implications:

1. Europe isn’t really one market, but many: the Draghi report showed internal barriers in Europe were akin to a 45% tariff on manufacturing and 110% on services. Europe has the talent and increasingly the capital to build global winners but, at least for some categories, it just doesn’t have the markets.

2. Any big ticket item critical to Europe’s future that requires integration and agglomeration is not going to happen naturally. It’s going to be hard won. If European optimists want to stop losing, we need to do more than simply celebrate the Draghi report and move on. Vision without execution is hallucination.

Instead we need a much deeper, longer term commitment to win the argument on Europe’s strategic interests, win power and deliver.

That means not only getting out beyond groupthink tech conferences and hacking Brussels, but building bottom-up ambition across the continent, i.e. the only way to secure genuine and durable consent rather than just temporary, top-down imposition.

European VCs should be leading from the front. The most overlooked lesson from a16z’s American Dynamism movement is that a fund is both a financial and a cultural institution, and these forces reinforce each other. a16z had a cultural, political and investment thesis which coalesced into a fund. They created a flywheel, translating private information into a cultural movement with real founder fit, investing in the resulting companies, and all the while bending the world in the direction they cared about. In effect, they have combined technology, capital, policy and culture to build a full stack power machine:

This crystallised when they recently hired Erik Torenberg and Alex Danco, with the latter capturing the thesis perfectly: “Power transfer technology…is the business of VC…to give founders more legitimacy.” Read the full post, which is a deeper theory of power and progress — by building and leveraging distribution to give founders the power they need to build a better future — than I’ve heard from any other investor or policymaker. For all the granular regulatory work that a16z do, their ability to use media to confer power on their portcos, building awareness and legitimacy with customers and policymakers equally, may be the most valuable ‘policy’ work they do.

No one in Europe is building anything even remotely comparable, because European tech lacks a theory of power.

You don’t have to agree with everything a16z believes to respect or learn from the underlying strategy. The constructive lesson is to radically change the upstream cultural environment, raising ambition and creating the space and consent downstream for specific changes. Progress happens at the speed of culture, because culture is how you coordinate at scale. So let’s speed up culture! One reason that a16z are so good at this is because they’ve just got so many reps in by now.

Building this in Europe matters particularly because the necessary ambition definitely isn’t going to come from the existing group of stale trade bodies or righteous campaigners that make up Europe’s policy apparatus today. You don’t protect European democracy by endlessly describing and restating our values (e.g. EuroStack). You protect European democracy by acquiring leverage and winning. Only then do you get the chance to advance your values. Until we fix this Europe will stay in a holding pattern: often virtuous, rarely victorious.

So we need a different approach. For a start, there is just an enormous undersupply of media and culture to inspire a) aggressive founder ambition in the national interest and b) policy progress both in Brussels and across the continent, produced in different languages and tailored to countries’ local contexts.

Where is the founder-led, European bible of audacity, the cri de coeur, akin to The Technological Republic? Where is the high-variance grants programme for tech and economic progress writers, to kickstart the posting-to-policy pipeline across Europe? Where is the talent pipeline for political candidates? Where are the events — public and private — that get European tech and policy leaders organised?1

Europe can, and must, win. For decades, we’ve outsourced our security to the US, energy to Russia and industry to China. Now all three look vulnerable. But for too long, a myth took hold that Europe can’t compete, which is neither true nor helpful. Meek belief breeds weak results. Leave that to the handful of sycophants who choose transatlantic herd approval over European agency.

Maybe not so long ago Europe didn’t have a generation of founders and investors with major exits who could anchor and fund this new European power machine. But that’s no longer the case. US tech has now funded American Dynamism, The Technological Republic, a $120m Abundance and Growth philanthropic fund, many more $millions for super PACs, think tanks and other grant programmes, and now even a new film studio about American exceptionalism. Why can’t we build something similar — even if at a smaller scale initially — to seed/scale more efforts like these and build a cultural platform to flood the attention ecosystem?

The vibe shift in European tech — led especially by a new generation of audacious, hard tech founders and investors — is overdue and welcome. But there is so much more to be done. It’s time for the European counter-elite to lead from the front. If that’s you, let’s talk.

So while AI scenarios are indeterminate, the directional choice facing the UK is clear. We can either:

1. Rapidly build our stake in the emerging AI value chain, enable economic growth, protect and widen our tax base, and retain geopolitical leverage and national agency, OR

2. Stand by while British labour is automated by foreign capital, lose control over the cost of and access terms for this newly critical economic input, and put our tax base, welfare system and social contract under even greater pressure

Either by default or by design, we are currently opting for (2).

Sovereignty is often confused or conflated with self-sufficiency, protectionism, or owning the entire AI stack. The goal is not to domestically produce everything: this would lead to low-quality, uncompetitive capabilities. But as automation becomes attractive and AI becomes a critical economic input, imported access to models may be throttled, weaponised or taxed — leaving the UK exposed to supply shocks, higher input costs and rules we have no say over. Without domestic capacity and control, labour automation will amount to offshoring and we could be left streaming our economy from abroad.

We’ve seen a version of this before, as software ate the world and everyone discovered the implications of deprecation. When Microsoft closed its ebook store, we learnt a new phrase: “the books will stop working”. Or as Alex Danco wrote in Everything is Amazing, But Nothing is Ours:

Worlds of scarcity are made out of things. Worlds of abundance are made out of dependencies. That’s the software playbook: find a system made of costly, redundant objects; and rearrange it into a fast, frictionless system made of logical dependencies. The delta in performance is irresistible, and dependencies are a compelling building block: they seem like just a piece of logic, with no cost and no friction. But they absolutely have a cost: the cost is complexity, outsourced agency, and brittleness. The cost of ownership is up front and visible; the cost of access is back-dated and hidden.

National sovereignty is about having the power to durably control our destiny. As a friend said: “we can rely on global markets for the supply of almost every good or service we might ever need. Digital brains might be one of the few exceptions.”

So we must build capabilities that limit our dependencies and generate strategic leverage — by controlling particular niches or chokepoints — in order to secure access to all required capabilities. This requires major reforms to energy, planning, and regulatory bottlenecks to deployment. And we must also urgently reposition the state to secure the social contract, by anticipating and preparing early for the coming changes to our tax base, welfare state, education system and defence capability.

Subscribe now

Context

1. AI is quickly becoming a critical input to the economy: not only are technical benchmarks being surpassed daily, but the top AI companies are generating revenues approximately 4x times faster (FT) than the previous generation of leading software companies. This represents a large and accelerating shift in value capture which is likely to continue, as model improvement shows no sign of slowing.

2. The UK has very little stake in the new AI value chain: For decades, the UK’s economic model has relied on outsourcing core inputs and focusing on value-adding services. Now, we’re experiencing nearly 20 years of wage stagnation just as AI threatens to automate our remaining strength in services. Yet even with GoogleDeepMind in London, value is largely realised and taxed in the US. We have very little stake in the emerging AI value chain: energy generation, semiconductors, networking, datacentres, models, drones, robotics and applications.

3. Datacentre capacity is a key litmus test: As AI automates cognitive labour, domestic datacentre capacity is critical. The UK will want at least some AI inference (i.e. deployment), if not all, to run locally within its borders – e.g. due to military or regulatory sensitivities, or where low latency is particularly important (e.g. trading). Yet compute demand already massively outstrips supply, so middle countries with limited domestic capability — like the UK — are at real risk of missing out as leading nations secure their own supply first. Relying on other countries brings risks too: the US has already shown its willingness to use export controls for semiconductors, and tariffs more broadly, to put economic pressure on adversaries and allies alike. Given that you cannot ‘stockpile’ intelligence in the way you can for energy, if the UK wants control over its own economic production — a core feature of any sovereign state — then it must build domestic supply to avoid losing control over both the costs of, and access to, this newly critical economic input.1
   
   The UK does have a large and growing base of datacentres built for cloud computing, some of which will be useful for AI.2 But a good test of AI seriousness is to look at the pipeline of supercomputing facilities which are particularly designed for AI training and/or inference. On that measure, only a handful of small facilities are live or planned in the UK, which is way behind not just the US, but comparable nations like France:

Where the UK needs to increase ambition

The UK is not yet taking the risk of losing this race — and the resulting risk of economic and geopolitical irrelevance — sufficiently seriously. It can do so by changing course in 3 ways:

1. Fixing energy and planning to accelerate datacentre build-out:

AI datacentres require lots of electricity: in just 3 years, AI datacentres are likely to draw ~50GW of continuous power in the US alone, roughly 1.5x the average electricity demand for the entire UK economy.

Yet the UK’s energy policy today is not focused on building abundant capacity or doing so quickly: instead, the goal is 95% grid decarbonisation by 2030 and a fully decarbonised grid by 2035. This is easier to achieve if you assume only a modest increase in electricity demand rather than significantly increasing capacity to enable entire new classes of always-on, energy-intensive compute.

To that end, NESO forecasts electricity demand growth of only 11% by 2030, mostly from EVs and heat pumps, while estimating that datacentre electricity demand will increase from 5 TWh today to 22 TWh by 2030. Yet, as Alex Chalmers writes, this is considerably less than what DSIT thinks is necessary: DSIT’s Compute Roadmap forecasts that the UK will need “at least 6GW of AI-capable data centre capacity by 2030” and that “demand could exceed this baseline significantly”. If these datacentres draw 6GW power continuously, this would amount to 52.6 TWh demand annually by 2030: more than double the 22 TWh that NESO is planning for.

Instead the UK should focus on energy maximalism — i.e. increasing electricity generation as much, and as quickly, as possible — in order for datacentres and other energy-intensive industries to be viable in the UK. This would also be a good way to distribute the benefits of AI: when electricity power becomes a bigger constraint than technical progress, countries that can supply power will capture significant economic value.

The other variable here is cost. Although cheaper electricity wouldn’t hurt, the major cost driver for large datacentre projects is the upfront capex, so developers actually tend to care more about speed and grid capacity (e.g. it currently takes approx. 15 years to get a grid connection) than price. To that end, nuclear power is probably the only way to massively increase electricity generation safely, cleanly, reliably and land-minimally (i.e. it’s good for land use efficiency and to reduce transmission costs). And it could also be much cheaper: while project mismanagement is rife, one of the main reasons that UK nuclear is 4.5x more expensive than in other countries like South Korea is due to disproportionate regulatory issues.

But electricity prices clearly do matter for wider sovereignty concerns: the UK has the world’s highest industrial electricity prices, averaging 4x US industrial electricity costs in 2023, as Jonno Evans writes. This gap plays a significant role in where companies invest, hire and build new projects, particularly affecting energy-intensive strategic industries like chemicals, steel, batteries and manufacturing.

Unfortunately, it looks like one of the main drivers of cost increases is our focus on grid decarbonisation, due to both increased wholesale prices of renewables and increased transmission capacity requirements (e.g. to transport electricity from generators in the North to users in the South). As someone who cares a lot about the climate, I am increasingly concerned that we are overly focused on decarbonising electricity, which represents only ~20% of total UK energy use, rather than prioritising cheap electrification of dirtier industries like transport and heating to reduce overall emissions. Maybe electricity is clean enough for now?

2. Clearing regulatory bottlenecks to enable AI training and diffusion:

As countries adopt AI, and reap the productivity gains at different rates, this will have an exponential impact on their economic power. So it’s important to start compounding early.

As such, clearing barriers to deployment is critical for the UK to accelerate economic growth and remain a relevant power with influence over AI’s future trajectory. In practice, there are 2 particularly important dimensions to this:

1. Copyright restrictions mean it is currently illegal to train a competitive AI model in the UK, yet unfortunately these restrictions do not even protect the UK creative industries, given that more liberal jurisdictions for model training exist elsewhere. Regrettably, the UK’s current policy is therefore a lose-lose. As Julia Willemyns has argued, the UK should therefore accelerate a Japan-style, commercial text and data mining exemption so that competitive models can be trained in the UK, in turn enabling us to support domestic value capture.

2. Elsewhere, regulatory bottlenecks hold back rapid, confident deployment of AI in health, transport, finance and many other sectors. We will need to unblock these bottlenecks via derogations or non-enforcement agreements, calling in regulatory decisions, or issuing new licences to firms where regulators are moving too slowly. So far, the Regulatory Innovation Office seems to lack the powers required.

3. Repositioning the state to secure the social contract

Britain is caught between a fractured national spirit and the onrushing upheaval of this new world. This is fertile ground for zero-sum, destructive forces who tear down more than they build. To avoid overseeing potential decline akin to the 70s and 80s, this government must get ahead and prepare.

While there are lots of uncertainties in AI, it is almost certain that the UK economy will undergo a major re-ordering much more quickly than most policymakers expect. This is likely to include significant labour market automation and major changes to the tax base. The government should prepare for this shift by:

• Planning for major shifts in the tax base: given the likelihood of increasing labour automation (starting with young people and recent graduates), payroll taxes like income tax and national insurance may shrink as a share of government revenues, while corporation taxes may grow in importance. Yet this also emphasises the importance of retaining UK HQs, so that value (and tax) is realised in the UK rather than shifting outside the country. Taxing less mobile sources of value, such as consumption and/or land, should also be considered.

• Welfare, university & healthcare reform: The welfare state is already creaking and the tax burden is high, due to demographic pressures, waste, and a failure across the state and NHS to adopt even internet-era technologies. The university system is already falling short for students, researchers and public finances, as Ben Reinhart set out in Unbundling the University. Now, as graduate jobs get automated, we face an overhang of universities failing to prepare young people for employment and saddling both them and the state with the expense. The only way to transform any of these public service at scale — to improve outcomes while reducing costs — is via bold new national institutions organised around modern technologies.

• Defence and industrial capacity: Strategic assets in the AI value chain can only be leveraged if you can defend them. For example, in trade disputes with the US, Taiwan and the Netherlands are unable, respectively, to leverage TSMC (which fabricates semiconductors) and ASML (which makes the ultraviolet lithography machines critical to the semiconductor fabrication process) despite them occupying monopoly positions in the supply chain, because neither country controls their own defence. It is therefore critical to fix procurement and rebuild the UK’s defence and industrial base, particularly at a time of heightened geopolitical tensions and unreliable security guarantees.

This piece is intended a) as a personal exercise to develop and untangle my thinking on AI, energy, regulation and UK sovereignty, and b) to write down an argument that I think is often implicit, but rarely championed publicly. There’s a good chance I’ve been imprecise in parts or missed some key detail somewhere, so if I’ve got something wrong please let me know.

Subscribe now

Where we are

Britain is in a bind. We are exposed to external shocks — geopolitical instability, energy volatility, technology outside our control — and internal strains that weigh us down: weak state capacity, high tax burdens, creaking infrastructure and poor value public services.

We are buffeted around with no redundancy in the system or counter-leverage to offset these pressures. We have little optionality. Policy is constrained by sclerosis and scarcity.

Now we face an AI inflection point that, for all its benefits, risks further instability: a new value chain we have little stake in is emerging — models, drones, robotics, networking, chips and energy — just as the vulnerability of our services-heavy economy is laid bare. Value may quickly accrue to AI companies domiciled and taxed elsewhere. The impact on our society and the prospects for government revenues is stark.

Where we need to get to

The correct response to volatility isn’t stability, but momentum. We can only regain control by restoring dynamism. We can only protect dynamism with control over our future.

That future is indeterminate. Security guarantees are unreliable, superintelligence is looming, bond markets are skittish. So we must build capabilities that give us optionality in a range of scenarios, by maximising opportunity within current constraints and pushing back against them.

The ingredients of comparative advantage — energy prices, frontier science, industrial capacity, talent networks, etc. — may be fixed in the short run, but are up for grabs in the long run.

Equally we need both tactical mechanisms to sidestep incumbent vetoes today and to build a culture that dissolves them long-term. Tackling parliamentary impotence, creating durable fiscal headroom, and competing globally on institutional agility are all critical to create the most scarce resource of all: room for manoeuvre.

Sovereignty must not be confused for self-sufficiency, protectionism, or trying to control the entire AI stack. It is about building leverage to durably control our destiny. National sovereignty underpins individual agency.

In today’s world, long-term strength requires hard power, economic abundance, and breaking the trade-off between state capacity and cost. So restoring national advantage requires ambitious supply-side reforms, building early stakes in new value chains that may upend the global economy, and fostering large firms that generate leverage either by occupying strategic chokepoints or by sheer virtue of their scale.

How we get there

Today, we lack the urgency and consensus to reposition the country quickly and get ahead of what’s coming. Britain is caught in a liminal space between a fractured national spirit and the onrushing upheaval of this new world. This is fertile ground for zero-sum, destructive forces who tear down more than they build.

Instead, those who can see the path ahead — not only for AI, energy, defence and state capacity, but also for collective purpose and national identity — have a duty to get organised.

Even if we can renew domestic productive power — reinvigorating economic growth and restocking the assets of statecraft — the risk of extraterritorial value transfer, from British labour to US capital, poses an existential societal threat.

This comes at a time when, already, the welfare state is creaking and the tax burden is high, due to demographic pressures, waste, and a failure to adopt even internet-era technologies, leaving the state overweight, indebted and underpowered. We will need new national institutions to transform the state at scale, while also acting as anchors of national purpose.

Every few decades, Britain faces a generational moment of self-reflection, when a ruling consensus breaks down and a new settlement is required. Today, to navigate the overlapping and indeterminate challenges of AI, energy, security and state capacity, the new order must focus on rebuilding optionality and national agency. In effect, a new Sovereign Consensus.

Subscribe now

PART ONE

Intro

The UK is facing economic decline, renewed security threats, and an AI inflection point that should prompt reflection about our priorities. Navigating this successfully will require decades of consistent, durable statecraft, investment in hard capabilities, and a steadfast commitment to deepening sovereignty and accelerating progress.

But to secure our material futures, I’m beginning to think we first need to revive something deeper.

In a secular, rationalist and increasingly atomistic society, we’ve lost many traditional sources of binding, higher purpose. Increasingly I think this means we lack not only individual, spiritual fulfillment, but also a durable, collective consensus for progress.

The old sources — tradition, religion, country — can speak to awe, the surreal, beauty, wonder, belonging, and place, but for many are too irrelevant, inconsistent, or contested to underpin collective purpose.

Many fellow champions of progress perhaps see the pursuit of science and tech progress as a source of purpose itself. But I think there’s something missing.

We need a new national story, a refounding. We need to craft new myths and institutions that not only inspire progress but also a deeper sense of place, connection, and meaning along the way.

Over the last couple of years, as well as caring more about our national trajectory as the stakes of decline become ever clearer, I’ve developed a deeper connection to this land, its lore and folk histories. It not only offers a source of higher meaning, but also a potential backdoor to a broad-based, intentional, liberal patriotism that goes beyond the sort of latent, patriotic revealed preference I suspect many feel today. This reassertion of national purpose and collective identity is essential to underpin a new, durable consensus for science, technology and economic progress.

These two strands — science and tech progress, and a deep connection to British land and lore — may initially feel disparate. But they are not only twinned by magic, alchemy, weirdness and variance, they can help us deepen our connection to this place and one another. With a wider sense of national purpose, we can create a more durable cultural consensus to navigate choppy waters ahead.

The need for a new story

This government has championed a ‘decade of national renewal’. I would love to see this promise live up to all of the spiritual and cultural implications it entails, rather than being reducible solely to policy reforms and infrastructure projects (necessary though they are).

National renewal should be about building a durable coalition for progress that’s bonded by purpose, spirit and belonging, not just transactional, contingent, short-term incentive alignment.

The moment has come for a refounding, not only due to the decline of higher purpose, but also because we’re caught in the liminal space between a fractured national spirit and the onrushing upheaval of a new geopolitical, post-AI order.

In Revolt of the Public, Martin Gurri writes of how the sclerotic public institutions of today's developed economies are a legacy of the industrial age. In this era, states derived their authority from a monopoly on access to information, administering a sort of 'political Taylorism' of top-down bureaucracy. Now, a decentralised tsunami of information and the attendant transferal of power has not only disrupted old institutions, but impeded new ones being built. As Gurri writes, the "mirror [in which] we see ourselves...has fallen and shattered, and the public now inhabits the broken pieces on the floor." The task now is not just tearing down legacy structures, but building something new, substantial and enduring.

I’ve never been a huge royalist, but when the Queen died and that thread of history had been severed, I found a new appreciation for the pull of tradition and the need for durable institutions that outlive and steward us through time. If the internet and information abundance enabled the latest wave of institutional and national introspection, then even extremely cautious expectations about AI timelines should demand an even bigger reflection.

Alexander Karp and Nicholas Zamiska’s new book, The Technological Republic: Hard Power, Soft Belief, and the Future of the West, is maybe the best and grandest attempt I’ve seen to diagnose the coming technological upheaval and synthesise science and technology progress with a distinctly national project.

I managed to see a copy pre-launch, and whatever partisan leanings you expect, I urge you to give it a read. It eviscerates the spurious complacency of many technologists’ failure to work on and take responsibility for projects of major importance; argues that “the market is a powerful engine of destruction, creative and otherwise, but it often fails to deliver what is most needed at the right time”; and laments how techno-utopianism has devolved “into a narrow and thin utilitarian approach…[where] the vital yet messy questions of what constitutes a good life, which collective endeavors [sic] society should pursue, and what a shared and national identity can make possible have been set aside as the anachronisms of another age”. I suspect people with very different politics to Alexander Karp will find much to agree with here.

More than progress

There are sparks of this spirit in the UK. TBI’s A New National Purpose series long ago made a similar argument. Entrepreneur First, with a different theory of change, has made entrepreneurship high status — a cultural achievement that’s arguably more important than financial returns.

But we need much more of this energy and institutional experimentation, and it cannot only be based on the cold hard logic of production.

If the progress crowd — the boosters, techno-optimists, e/acc, uk/acc, anglofuturist and progress studies types — genuinely care about widening our coalition, then not only do the downstream benefits need to be widely tangible, but the spirit and aesthetics of these movements also need to inspire broader, durable support.

So far, despite rallying a solid base, I think these projects are either failing spiritually or aesthetically. Too often they reduce what matters to pure materialism, offering little higher purpose, while the metallic textures of these futures are often coarse and soulless. Run your finger over the contours and you get brushed steel, dimpled concrete and muted glass, not squishy moss or brittle bark. There is little enchantment, minimal poetry.

Meaning and aesthetics are a way of both attracting and repelling. Not everything can or should be about everything, and often it’s good to repel some people (e.g. free-riders, doomsters, NIMBYs) while attracting others. But especially if you’re explicitly building an aesthetic/cultural project, with rallying visions and in-group identifiers, and you’re only really drawing a fairly narrow (and often v male) group of people, then something is going wrong. It’s not just bad vibes, it means your project won’t scale to underpin the durable, wide consensus necessary to survive contact with reality’s messy tradeoffs. I think this is where a lot of tech/progress-adjacent projects are falling short. (There is also, way too often, an offputtingly fratty culture: ‘are you serious & based, or are you a soft NPC?’)

Another cluster are more focused more on care, community and dignity. They argue that the first group are often silent on much of this — which has more truth than many care to admit. But too often this group has little to say about generating economic prosperity. At best, they’re focused on principle but not power. At worst, they free ride on the invention and execution of others. ‘Tech CEOs need an ethics class’ is just ignorant; ‘tech for good’ is too naff to inspire widely. But softerfutures.xyz or cute/acc? There’s something there.

Effective Altruism is one of the more ambitious attempts to synthesise these strands into a secular higher purpose. But utilitarian welfare-maximising is just so…mid. EA’s equalisation of all lives, everywhere, is laudable in theory, but in practice it shaves off the textures of place, community and personality. It makes life a commodity, a series of fungible units, denying identity and our own humanity. This same philosophy encourages value exchange over relationships, brevity over lyricism. Do not leave the conversational door ajar, just put it in an email! We’re sanding down our minds with microscopic tl;drs, until we’re perfectly, exquisitely smooth-brained.

You cannot build a world on axioms alone. You need texture. Divorcing us from this richness isn’t how you build a movement that inspires.

We need better choices: progress and preservation, based and woke, rational and mystical, metal and moss.

What’s holding us back?

I think a significant amount of the tension comes from tech, policy, and London having these accelerating, centripetal forces that incentivise people to acquire as much of the central currency as possible — money, power, approval — without stepping back and equally advocating for everything else that matters for the good life.

I only recently read CS Lewis’s Inner Ring but I’d strongly recommend it on this point, alongside Sebastian Garren’s excellent critique about CS Lewis’s pessimism on the capacity for technology and applied science to provoke wonder.

Or, if tl;dr, Garry Tan’s take works too:

I mostly ‘touch grass’ by actually touching grass. But over the last ~18 months, I’ve got more interested in ancient sources of meaning — if anything due to the absence of anything equivalent in secular society. Others seem to feel that something is lacking too. Religion is growing not only in US tech and wider society, but also in the UK.

I still struggle with much about religion, but congregation, sermons, rituals, fables, mysticism, guidance, reverence, contemplation and higher purpose? Sign me up. Maybe what I call beauty, awe, wonder & transcendence, others just call God?

And there’s always David Foster Wallace (my emphasis):

You get to consciously decide what has meaning and what doesn’t. You get to decide what to worship.

Because here’s something else that’s weird but true: in the day-to-day trenches of adult life, there is actually no such thing as atheism. There is no such thing as not worshipping. Everybody worships. The only choice we get is what to worship.And the compelling reason for maybe choosing some sort of god or spiritual-type thing to worship–be it JC or Allah, be it YHWH or the Wiccan Mother Goddess, or the Four Noble Truths, or some inviolable set of ethical principles–is that pretty much anything else you worship will eat you alive.If you worship money and things, if they are where you tap real meaning in life, then you will never have enough, never feel you have enough. It’s the truth. Worship your body and beauty and sexual allure and you will always feel ugly. And when time and age start showing, you will die a million deaths before they finally grieve you. On one level, we all know this stuff already. It’s been codified as myths, proverbs, clichés, epigrams, parables; the skeleton of every great story. The whole trick is keeping the truth up front in daily consciousness.

(I realised this the more I thought about traditions like marriage. For a long time I took a fairly postmodern, critical approach, deconstructing the symbols and institutions and rendering it all meaningless. But at some point I decided you can choose to imbue it with meaning. You can choose to believe in the power of symbols, tradition and institutions.)

Where I have always found a reliable well of spiritual refuelling is in nature and art. I have come to love London in many ways, but I will never not find the detachment from nature, the land and the seasons that a modern, urbanised, agglomerative, office-block services-based economy entails deeply depressing. Just over a year ago, I went to an immersive exhibition designed to reconnect you to nature. While the works were beautiful, nothing felt more isolating and disconnecting than strapping on a VR headset to transport yourself to a dense forest world.

I like my work, but I do envy the sense of stewardship and connection felt by those who have worked the land. I wonder if a connection to nature and folk/spiritual heritage could be the backdoor to collective national purpose — and thereby a durable consensus for progress — that we’re missing.

PART TWO

So, gear change. Should we all be a bit more…pagan?

In December 2023, I got one-shotted by this:

I think being open to magic and mysticism is a pretty good way to offset the cold logic of production and retain/protect life’s poetry. More importantly, it’s a choice to suspend disbelief. Someone once asked if I believe ‘truth seeking (≈ rationalism) and purpose seeking (religion, woo, lore) are compatible?’ That’s exactly what I want to synthesise. Holding ‘we made rocks think’ in your mind, just long enough to get the epiphenomenal enchantment from it, is actually about agency. Let yourself hallucinate a little, detach from reality and leverage the different logic of latent space, where rationalism doesn’t apply and proof doesn’t matter.

Science and secularism are all well and good, but we must not lose the myth and magic. Again, I’m not (yet?) religious, but I think the Church reformists have it right on this:

“the Church has been encouraged to [replace] the glorious other-worldly mysticism of saints and angels…Jesus has been transformed from the Lord of Heaven and Earth, and the Judge of All, to my best mate, with all the gravitas of a crisps advert…Christianity needs to get just a little bit more weird and badass. More Caravaggio, less pastel. The Church exists to address the great mysteries of life — death, life, forgiveness, fear, passion, hope.”

We can find a secular form of this transcendent weirdness in ritual, poetry, folklore and the surreal, while still celebrating engineering and rationalism. But so far, no one has woven these features into a project of national purpose and progress. So let’s try.

Sovereign Albion synthesises all of these strands. It’s a world where science and tech progress coexists with a deep connection to the lore and the land and we encourage new rituals to restore our sense of higher purpose. It’s a place where rocks think and science is magic. It’s somewhere you can find meaning and purpose beyond religion, in secular third spaces, silicon wafers and the natural world: dancefloors, saunas, mountain summits, hot springs, pubs, bothies, libraries, cairns, stone circles, stadiums, lab spaces, crashing tides, choir halls, chip fabs, and permanent exhibitions. It’s a Britain not only rooted in the empirical rhythms of the natural world, but perhaps also the supervenient nourishment and guidance anchored around the wheel of year…just without the degrowth-y associations that usually go with that.

I grew up in the North West, so for a long time I was extremely sceptical that the South could compete with the beauty of the Lake District, the spirit of the Welsh mountains or the drama of the Scottish Highlands. But one simple psyop I’ve used — it’s really not that deep — is just being more intentional about travelling around Britain and going deep into the lore and culture of the places you visit. If you like to listen to the music or read the literature of wherever you’re visiting when you travel, did you know you can do that for Britain too? I dare anyone to listen to Cornwall, My Home and not be roused and/or become a Cornish independence sympathiser. You can just try on new rituals (a lesson I learned after reclaiming German citizenship yet finding myself knowing none of the customs).

The moody Tors of Dartmoor, the mysticism of Cornwall, the ceremony of Lud’s Church, the idyll of deep England in the Cotswolds and beyond — all of these places have their own mythic charm. I genuinely believe Britain is one of the most beautiful places in the world, and not enough people know it. Yes, I’m biased. That’s the point.

If Britain is an immersive experience, then Albion is its open world game:

• it has main quests: progress & purpose

• it has sacred sites to recharge: not just traditional places of worship or even ancient ruins, but secular third spaces too

• it has game mechanics: seasonal cycles, resource management balancing progress with preservation, and multiple play styles (builder, guardian, etc.)

• it’s rich in background lore and wisdom

Yes, it’s speculative, but it’s hyperstitial: sometimes you have to throw out the future in front of you and then catch up to it. Founders and artists use this sort of venturi effect all the time. We can all learn from their ability to create something out of nothing and build the futures they want to live in. You too can just build worlds!

Maybe my own personal experience won’t generalise. But as my partner says, the resurgence of i) tradwife cottage core simplicity, ii) horoscope girles and modern spiritualism, and iii) BRAT/hedonism creates creates fairly fertile conditions for this :)

If so, maybe the land and nature can be sources of non-rivalrous patriotism. By focusing on something we can all connect to, we can escape much of the dominant zero-sum, contested criteria for national belonging and rebuild a collective foundation for progress.

Sovereign Albion is just a playful idea, a useful device to project onto that I hope to keep working on. But national renewal is a necessity.

Here are some practical things I’d love us to actually do, and which I think could help restore a wider, intentional sense of national purpose:

1. Fine-tune an LLM on Britain’s founding myths, fables and tales to generate sermon-style guidance for making sense of a chaotic world and integrate connection to nature with science and tech progress; I mocked up a websim prototype here

2. Massively expand National Parks and rewild major swathes of the UK to enable much better biodiversity; then set a presumption in favour of planning approval everywhere else

3. Establish the Right to Roam across England and Wales, to restore access even if not ownership to our lands

4. Restore Stonehenge to its original design and ensure people approach up the Avenue (the hill up to the front entrance), to build the anticipation and wonder as in the Turner painting above, rather than English Heritage just inexplicably dropping you off in a bus at the back

   • Also install mini, Stonehenge-inspired trilithons across the country, akin to mini Shinto Torii you see across Japan [N.B. We have a lot to learn from Shintoism! But this piece is already mashing too many ideas together.]

5. Build new media platforms & idea machines to tell stories by, for and about Sovereign Albion

6. Reshore sovereign capabilities and industrial supply chains where possible via automation

7. Move the New Year and align the clocks changing to the Spring Equinox, to heighten celebration of the changing seasons

8. Get the BBC and HBO to commission a massive new TV series retelling the Arthurian tales, co-directed by Robert Eggers and David Lowery

9. Build & support more third spaces and ‘secular cathedrals’ across the country, built with British materials: should include some new, dedicated formats for secular congregation, contemplation and focus, but also more bathhouses, saunas, dancefloors, bothies, libraries, cairns, stone circles, choir halls, etc.

10. Restore Britain’s ancient kingdoms as part of a new federal state, or at minimum rebrand local authorities to resurface their ancient roots, in lieu of your latest municipal conurbation’s pallid aesthetics

11. Make Jerusalem the English national anthem

12. Clean up Britain’s chalkstreams and waterways

13. Build new towns and cities using Create Streets design codes, so we build with beauty by default

14. Offer a Bildungsroman bundle to every 18-year-old (instead of pretending we’ll bring back national service), to include: a UK-wide interrail pass, the right to use a student loan for purposes besides higher education, a premium Ordnance Survey app sub, a calendar displaying both the wheel of the year and dates of major British invention, a leylines map, and free entry to any National Trust, English Heritage site / a tour of any factory or lab space for a year

15. Create new chivalric orders both for innovation (see this proposal) and stewardship of the British wild

16. Adopt ‘progress, land, spirit’ as our national motto

The Sovereign Albion Starter Kit:

Until the last couple of years, I hadn’t really come across the word ‘Albion’ except for a) football teams and b) the VC fund. Honestly, I’m probably misusing it (e.g. I very much intend to include Scotland, Wales and Northern Ireland in this imaginary). But I’ve since read/listened to a lot that talks about Albion, in particular Zakia Sewell’s My Albion documentary, which explored the songs and symbols of British folk culture and her own place (being of both Caribbean and British descent) within it. It’s really worth your time, and I’m eagerly anticipating Zakia’s book on the same topic.

One of many, many takeaways I had from My Albion is that there is new folk art and culture being created all the time. Unfortunately I couldn’t shoehorn all of it into this piece, but I want to include at least a handful of further recs, which I’ll try to add to/organise further when I get chance. I haven’t focused here on wider science/tech/progress recommendations, as they are well covered elsewhere. But do read the Karp book when it comes out. I’d love to hear your suggestions too.

Subscribe now

I also feel enormously grateful to several people in my life who have not only let me bang on about so much of this for nearly 2 years, but helped me think through how it all fits together.

So, alongside all the links above, the starter kit:

Land, folklore and Britain

• There’s a beautifully weird connection between Charlie Cooper (This Country, Myth Country) » Mortimer & Whitehouse Gone Fishing » Zakia Sewell’s DJ sets and documentaries » and everything Weird Walk publishes. Go deep in them all.

• The same is true of Wolf Hall » Mark Rylance » Mackenzie Crook » Detectorists » Johnny Flynn » and basically anything Robert Macfarlane touches but especially the Lost Words / Spells

• Nan Shepherd’s The Living Mountain

• Richard Powers’ The Overstory

• Amy Jeff’s Storyland, a new mythology of Britain weaving together diverse stories and legends including prehistoric stories, the Brute Chronicles and many Geoffrey of Monmouth tales for the modern reader

• Wild Service, which explores a radical environmental politics to inspire greater stewardship of, and connection with, nature; compiled by Nick Hayes who leads the Right to Roam campaign

• Oliver Smith’s On This Holy Island, which chronicles a modern pilgrimage across Britain visiting religious sites, standing stones and music festivals

• All of these cinematic mossy quests, but particularly The Green Knight

• Merlin Sheldrake’s interview on Rick Rubin’s Tetragrammaton

• Guy Shrubsole, The Lost Rainforests of Britain

• Every seasonal restaurant ever, Julius Roberts, the Mainly Breakfast calendar

• Taboo, the TV series with Tom Hardy that inexplicably refuses to deliver on a promised second series

• Wild Isles, the BBC Attenborough documentary about Britain

• Scout Studio’s The Cull, The Eagle with the Sunlit Eye, and The Pine Hunters

• Stormzy’s Glastonbury set

• The new chapels, aka spaces for focus and flow state located around the country: Ashore, Unyoked, Unplugged

• Cornwall, My Home, by The Fisherman’s Friends. See also Scout Studio’s Men Are Singing short, about a Cornish men’s choir

• Carry Them With Us, an incredible album by Scottish small pipes player Brìghde Chaimbeul (who also plays on Caroline Polachek’s Blood and Butter)

• George Orwell’s The Lion and the Unicorn, and David Runciman’s Past Present Future discussion of the hypocrisy of English nationalism

• Roy Hattersley’s In Search of England

• This thread exploring Britain’s founding myths

Meaning & worldbuilding

• Mat Dryhurst & Holly Herndon’s The Call, a remarkable exhibition exploring the new aesthetics, rituals and protocols we might want to nurture for the AI era

• Elijah’s Close the App, Make the Ting, and his critique of BBC 1Xtra

• Alastair Monty’s proposal for the TxP Progress Prize to use non-invasive neuromodulation for ‘a weekly gathering that’s somewhere between a church service, Glastonbury, meditation, and an excursion with your friends’

  • See also Prophetic AI, a US startup working on non-invasive neurotech to induce lucid dreams and expand consciousness / the human experience

• Byung-Chul Han’s Psychopolitics and The Disappearance of Rituals

• Simon Critchley, On Mysticism: The Experience of Ecstasy, and his interview on Vox’s How To Feel Alive

• The Search Engine podcast in general, but particularly:

  • What does it feel like to believe in God?

  • How do you sit quietly in the middle of a storm?

  • Is there a sane way to use the internet?

  • How do we survive the media apocalypse?

  • Why didn’t Chris and Dan get into Berghain Part 1 | Part 2

  • What if ayahuasca made you stop podcasting?

• Softer Futures » Ida Lissner » Mat Collishaw » Ithell Colquhoun

• Sarah Drinkwater, The best companies build worlds

• Compound VC’s Time Since Launch clock

• FT: In praise of amateurism — a pianist’s story

• FT: Skimming, scanning, scrolling — the age of deep reading is over

• 1961 NYT Review of Arthur Koestler’s The Lotus and the Robot, which discusses the spiritual shortcomings of the West’s science and technology revolution and the search for enchantment from the East

Why are some folk a bit jittery?

• Not only are the weirdos and misfits that built ARIA, AISI and 10DS now out, but they haven’t been replaced at all.

• Tech has been moved out of the centre, with no residual capability in No10.

• Compute infrastructure is in limbo. Labour has plenty to say about tech in government & public services, but less to say about AI, startups, strategic advantage, and the broader innovation economy.

• The Home Office quickly slapped down any talk of lower visa costs, and wasn’t reined in.

What does the bull case look like?

• The new government’s focus on service and stability is a huge contrast to the last’s.

• We’re only a few weeks in. Mission Boards have just been set up, and announcements are due on the AI Action Plan and Regulatory Innovation Office soon. The new government deserves time to set out its thinking.

• GDS & CDDO’s influence had waned in the Cabinet Office anyway. Their move with i.AI is not so much a weakening of the centre, but a promotion for DSIT: its role today is now akin to a tech-native Cabinet Office, with a cross-government mandate to match.

• Labour has made some unequivocally top tier appointments — Sir Patrick Vallance, Matt Clifford, Varun Chandra, Emily Middleton — so they are accessing the necessary expertise. True, these aren’t like-for-like replacements, but the capability has been recreated & augmented in the aggregate.

• The temptation to centralise power was really a response to a failure mode of bygone department-led delivery. There is a path to progress where these appointees, and DSIT more broadly, are empowered to lead.

What stops us fully believing this theory is that it requires dedicated political and intellectual sponsorship for science and tech from No 10, possibly with specialists in DSIT, to back up appointees, advisers and officials when necessary.2 It’s clear that Lord Vallance’s de facto advisory role makes him more than just a Junior DSIT Minister. But there’s little dedicated empowerment direct from the centre. For the sake of 2-3 headcount, why not?

Without it, promising appointments will at best end up fighting against the grain – for attention, funding, and impact – and at worst be cut adrift from a centre that looks like it fundamentally lacks interest or instinct.

This is particularly important in the AI era, which requires every department to take responsibility. There’s a risk that Labour gets stuck catching up to the last decade of technology, focusing only on DSIT. But tech is not just a sector or stakeholder: it’s both a necessary force and an urgent opportunity, requiring every Minister to rewire their departments to accelerate wider economic & societal progress.

In absence of this steer, momentum could wane. Other Ministers will (and may already) be captured by the swirl of structural incentives, media crises and other well-intentioned but ultimately distracting initiatives that combine to form ‘the Whitehall machine’. AI quickly becomes a ‘nice to have’, or is discredited as simply ‘hype’, to avoid making changes too quickly. Already, turf wars have broken out over manifesto commitments. With a difficult Spending Review on the horizon, friction will build further. In the past, No10 and HMT clearly signalled that DSIT would win these battles by default. That backing isn’t yet obvious today.

Leadership matters. 

If you believe that exponential scientific and technology progress will define our future, then you need to be much more strategic about building a stake in that future as early as possible. It will affect the economy, jobs and our politics, whether we’re prepared for it or not. NVIDIA alone shows how technology shifts quickly lead to economic and geopolitical reordering: similarly, unfocused countries can quickly sink, while others rise.

Inertia isn’t an option. 

AI holds the potential to create a powerful cycle of growth. It could also be a force that replaces jobs, fosters inequality and leads to a potent populist backlash. Even AI-sceptics implicitly admit that its long-term impact depends on how we develop and deploy it. If our ambitions are modest – to use AI simply to improve efficiencies and reduce costs – the benefits will be modest, too. 

Yet if we use AI to rethink everything from health care and education to food production and energy generation, then the potential of AI is to reallocate and relocate human economic activity in a way that wasn’t physically possible even five years ago. The economic and societal gains would be significant.

Golden ages are not accidents. They can be missed, slowed down or accelerated by political choices. There are many possible futures, and we can shape and choose the ones we prefer. Do we want to grow our stake and say in the future, or be price and technology takers forever?

Time for new idea machines

Labour needs holding to account, to offer its own vision and political story for technological progress. It needs to make the leap from saying nice things about the importance of the tech sector, to really weaving it into a progressive narrative and spending political capital like it has for planning, infrastructure, energy and public service reform.

For those other areas, the government's job is a director, funder and project manager to unlock tangible value today. But those who are most excited about technology aren’t just motivated by efficiency gains or ‘digital transformation’, but by the power of emergent, experimental, curiosity-driven serendipity. This requires less marshalling and controlling, and more faith.

Take startups: they’re not just the most productive part of the economy and employers of the future, they’re the delivery unit of progress. They’re critical to Labour’s missions! They are hypotheses, speculations and blueprints for possible futures, just expressed in the form of a company. Today, they represent hazy, low-fi fragments of the future. But with optimism, agency and ambitious structural reforms, we can bring them into focus.

So far, Labour hasn’t quite developed its own instincts here. But that just means true believers need to win the argument. Those evangelising about tech and then complaining that Labour doesn’t ‘get it’ need to change tack. Nor can the vacuum be filled only with calls for regulation & stakeholderism, at the expense of material progress & prosperity. This is sales 101: first you need to understand demand, rather than simply pushing supply. Making things people want > making people want things.

So what does that demand look like?

Labour won the general election by promising change. And given the government’s adopted the core message that ‘Everything’s not awesome’, Labour will want to deliver change - to deliver progress - as quickly as possible. 

The last government was occasionally vibes-y and credulous, but it had a rigorous diagnosis about securing strategic advantage through science and technology. Instead, Labour’s default looks to be technocratic managerialism and stakeholder engagement. But that way lies reactive, low-conviction governance, not the proactive, strategic statecraft needed to capture this moment. This will not change the long run trajectory of the country. It will not deliver outlier economic growth.

How do we make the case for a different approach? We need to build a new world that not only champions science and tech progress, but one that connects to Labour’s priorities and values. Otherwise, you’re just shouting into the wind.

In effect, we need a new “idea machine":

a network of operators, thinkers, and funders, centered around an ideology, that’s designed to turn ideas into outcomes

As Asparouhova sets out, an idea machine:

is a self-sustaining organism that contains all the parts needed to turn ideas into outcomes:

• It starts with a distinct ideology, which becomes a memetic engine that drives the formation of a community

• The community’s members start generating ideas amongst themselves

• Eventually, they form an agenda, which articulates how the ideology will be brought into the world. (Communities need agendas to become idea machines; otherwise, they’re just a group of likeminded people, without a directed purpose.)

• The agenda is capitalized by one or several major funders, whose presence ensures that the community’s ideas can move from theory to practice – both in terms of financing, as well as lending operational skills to the effort. (Without funding, an idea machine is just that: an inert system that needs fuel to turn the crank and get it moving.)

This approach is about building a durable coalition that’s bonded by vision and values, not just transactional, ephemeral asks from trade bodies. 

To really get buy-in, it should attract, but also repel. 

A UK idea machine for technology-driven progress in the UK is likely focused much more on security, sovereignty and public value. It’s not scared by the words ‘industrial strategy’, but can focus these efforts on creative disruption instead of incumbent subsidy.

It connects signals from the frontier to the policy problems of today, putting people and problems first, technology second. It takes the opportunity of science and technology as central to future-proofing our economy and society, and to building resilience in an increasingly competitive and volatile global environment.

This machine should also inspire insiders to want to spend time in this world. Too many startup, science and tech policy discussions can be super boring, focused only on economic or transactional issues. While important, they’re rarely captivating. But we can choose to see the magic. 

It’s time to fire up a new idea machine. Who’s for rocket emoji securonomics?

Subscribe now