In short
EUVC Podcast Episode Notes
Episode Title
E670 | This Week in European Tech with Dan, Mads & Lomax
Overview In this episode of the EUVC podcast, co-hosts Dan Bowyer, Mads Jensen from SuperSeed, and Lomax Ward from Outsized Ventures engage in a holiday-home special discussion that tackles various impactful topics in European tech, geopolitics, and AI.
Key Points Covered
- DeepMind's National Identity Debate
- Discussion on whether DeepMind should be classified as a UK business despite being owned by Google.
- The implications of globalization on national identity in tech companies.
- AI for Science and Robotics (04:36)
- The emergence of an "AI scientist" era, where AI is used for conducting scientific experiments efficiently.
- National Curriculum Gemini (06:50)
- Concepts of using AI to create personalized learning experiences for every child based on the UK's national curriculum.
- SpaceX's IPO Speculation (09:39)
- Speculation on a potential IPO in 2026 and what investors are actually investing in.
- Discussion on the expected revenue from SpaceX's business models, especially Starlink.
- The Trillion-Dollar Imagination Gap (14:00)
- Commentary on the challenges Europe faces in the space race compared to the US.
- Portugal's Economic Success (19:43)
- Portugal was named "Economy of the Year" by The Economist for its impressive economic growth and stability.
- Europe's Position in Global Trade (22:58)
- Analysis of how Europe is caught between China's trade surplus and the US's waning focus on European alliances.
- Defense Spending Realities (32:07)
- Insights into the current state of defense budgets and the practical challenges startups face when strategizing in this sector.
- AI Corner Updates (34:25)
- Conversations about market sentiments around AI, including potential bubbles, innovations from companies like Mistral, and Meta's shift to closed-source models.
Detailed Discussions
- Sovereignty and AI
- The conversation highlighted how sovereignty now plays a critical role in startup investments, especially in defense and critical industries.
- DeepMind and the UK Government Partnership
- The partnership aims to leverage AI in public services, particularly in educational reform.
- SpaceX's Valuation
- The discussion outlined what constitutes the valuation of SpaceX, focusing on Starlink's profitability and future prospects.
- Portugal's Economic Growth
- The co-hosts reflected on the turnaround of Southern European economies, contrasting them with previous economic crises.
- Geopolitical Dynamics
- The hosts analyzed the shifting geopolitics with China’s dominance in exports and the US's changing priorities, particularly in relation to European countries.
- Defence Spending Insights
- They discussed the hype surrounding increased defense budgets and the actual pace at which funding is being allocated to startups.
Conclusion The episode concluded with a sense of cautious optimism about the future of European tech, emphasizing the importance of clear communication and narrative ownership in capitalizing on opportunities within the industry.
Additional Notes
- Call to Action: EUVC is offering a free bespoke communication strategy session to help teams in Europe sharpen their narratives.
- Disclaimer: The discussions in this episode do not constitute investment advice.
---
These notes encapsulate the major themes and discussions from the podcast, providing insights into the current landscape of European venture capital and technology.
Written by AI. May contain mistakes. Listen to the episode to check what was said.
Transcript
Automatic transcript. May contain errors.0:00Welcome to Upside, where we dig behind the headlines of the news affecting European venture. Today, Mads, Lomax and myself, and we are talking about SpaceX mega rays. But what are you really buying? Europe's euro success, the north-south polarity has flipped. Now the south is rich and successful in the north. Not so much. We're going to dig into that. China's$1 trillion trade surplus. What does that mean for us? Unpacking US defense spend, having a look at that. And more chips, please, in AI Corner.
0:43Distribution decides who wins. And in today's market, the teams who communicate clearly, the ones who own their narrative, are the ones who get ahead. At EUVC, we've spent more than half of a decade helping Europe's best founders and investors shape stories that open doors to capital, customers, partners, and even talent. To help more European teams communicate with impact, we're giving away one free bespoke communication strategy session led by one of our sharpest comms experts. Get ready to sharpen your narrative and strengthen your distribution. Scroll down to find the link and apply now. One lucky team will be selected.
1:25This show is not investment advice and the hosts of this episode may be invested in the funds and companies featured. But to start, we're going to look at DeepMind and the Department of Science, Innovation and Technology in the UK. Question for you chaps, would you classify DeepMind as a UK business or not? Well, it's owned by Google, so no. I mean, it's a simple answer. It's a UK origin business, the UK founders. That's where I was squinting. I was squinting, looking at it going, UK signing up with DeepMind. No, it's a subsidiary of Alphabet. No. No, fair, fair, fair. So what does it mean that a business is UK?
2:01If a German company makes cars with steel from a plant in Poland, are there German cars? Well, nothing is from anywhere, is it? I mean, I always look at the iPhone as an example. That was globalization, which, you know, is now dying. It's not completely dead yet. Your point is well made. I think you can make, I think the truth is you can make an argument either way. depending on what case you're trying to make. I think if the argument you're making is there's incredible talent. No, and also like look at imagine, you know, some of these public listed companies that we think of as European, like, I don't know, GSK or something.
2:40But if you actually look at the shareholder register, you might have a large majority, almost a majority of like US endowments, pension funds, like money managers. So actually none of the cap tables are really ever truly sovereign. And that, I think, is the beauty of the era we grew up in, which was, as you say, the era of globalization, where we all grew closer and this stuff started to matter a lot less. And we've now been rudely awoken to a new world where all the sovereignty stuff matters a lot more, much to my disappointment. But that is where we are. And that's the reality we have to face and act accordingly.
3:15with. It's very, very important these days, especially if you're investing in critical industries. If you take this back to the startup level, pre-seed, seed, series A, you do actually need to think more carefully now about sovereignty of your cap table. So obviously, if you're building in defense, you need to think about this. But if you're building actually in many like critical industries now, I mean, in the UK, they have this, I can't remember the name of the law, the law in the UK came in about four years ago that needs you to, if you're in one of certain You need listed industries, you need government consent for the investment, large investment to happen or the M &A to happen.
3:50So definitely is a thing for our time, of our time. I just don't know where the lines are. Anyway, I'm going to classify DeepMind as a UK business. If you remove the fact that it's a wholly owned subsidiary of Google, then yeah, it's very, very British. If you remove everything that's fundamentally true, then Dan, you can have your facts as you please. Well, anyway, let's dig into what they've done. So DeepMind, Department of Service Innovation and Technology in the UK, they have got together. They've, as of today, the 11th of December, they have announced a new partnership. Now, it feels to me like it's an extension of budgets that have already been allocated, but DeepMind has been named as the part of the provision.
4:32I don't know if that's true, but that's my sense. I couldn't find much in the press release around the money and the budgets. but it feels like we have the budgets approved, but DeepMind is going to be the partner. And what it means is new research lab in the UK, expanding partnership. So we're going to look at the AI Security Institute and look at what's going on there. Lomax, maybe you can kick off with a bit more of a kind of a scene set for what's going on with DeepMind and DSIT. Well, I think this is interesting, maybe ignoring the government bit for now and public services, AI and public services, more just there's a whole thesis that has been evolving and people have been executing against the AI scientist effectively is a lot of science is done by trial and error.
5:15Well, if that historically was done manually by scientists, it was very slow and by definition constrained by how fast the scientists could run experiments. Now with AI, with both software and robotics, so you have the software which actually in silico can design and run the experiment and then the robotics aspect, which can actually run with whatever materials you're working with. this whole thesis has been i mean look at ai for drug discovery which has had a lot of money gone into it in the last decade with mixed results but i think we're seeing better and better results now this is now happening across materials we see you know cusp ai physics x etc so this is just another example of this which is we can use ai and robotics to run mass experiments at mass scale to unlock new materials.
6:05In this case, I think they're particularly interested in superconductive materials, which can carry electricity with zero resistance, but there are other things they'll be looking at. I think this is hugely exciting. It normally requires a lot of money to do this properly, which is why sort of seed stage startups, I think have been constrained in the past by capital. So it's a good place to do it. It's good that it's happening in the UK. And the final point I'll make on this, why I think it's good that's happening in the UK as a seed investor, is that what often happens is the really smart people who are working on these projects sometimes spin out and create their own companies.
6:39I think GDM is the phrase that is used in our industry. The Google DeepMind X employees who spin out and create their own companies. We've seen that with Latent Labs. It's just one example. So I think that's exciting for me. It's great it happens in the UK because then that talent kind of spits out and creates new companies. But there is a government angle here as well, Mads, which maybe you had a look at. Well, I just, one particular thing I want to pick up on, and we sort of, we're still in this sort of third anniversary of the Jack Gpt release. And one of the things we talked about back then is this incredible potential of, for the first time in history, being able to provide a tutor for every child.
7:17I think it's the most transformative thing we can imagine in education. And you know, we're three years in and we know these things take time. It is, you know, the government is a bit of a supertanker. It can be hard for them to adopt, get things done. We know that many schools still struggle with how much should we use chat GPT and other LLMs in education. Is it just something that students use to answer their homework? Can you really use it to learn as a tutor, as a mentor? What we're seeing now here as part of this in partnership is that there's going to be a focus on creating an education focused version of Gemini grounded in the national curriculum, which is a wonderful idea and hopefully a signal that the UK government and UK education is going to take the use of LLMs even more seriously in education because I just think it is transformative on a level almost unparalleled anywhere else.
8:09They've also got the more wide public services mandate I think to create efficiencies behind the scenes so it'd be interesting to see how this all connects and how this all rolls out. Demise is obviously the closest you get to a super scientist rock star. I mean, he's just incredible. And kind of knowing that he is in there, he's focused on this. And yes, you guys kind of said a bit top of the call. Is this a UK firm? Is this a US firm? The lab is in London. It's led by a Brit who we know really cares about the success of the United Kingdom. And that's great to know. So yeah, I'm definitely a big fan of this.
8:48Yeah, the AI teacher thing, by the way, I think is fascinating. And there are some really cool companies now in C stage, Series A stage, Series B stage that are coming through that are executing against that thesis. So I think it's hugely exciting. Don't forget the Transformer paper, you know, attention is all you need, was not done in DeepMind, right? It was done in another part of Google. Was it not? I didn't know. No, no, no. It was done in like the Google Brain research team in the US, no? And so actually maybe DeepMind team weren't working on Transformers, but in a way they kind of like as the sort of marquee AI research lab of Alphabet, they kind of missed the initial stage of Transformers.
9:24But the point is, is that it's great that there's still a lot of investment behind Google DeepMind. And it's exciting. And as I said, it's great. It's happening in the UK because you actually end up, what I get excited about is founders who cut their teeth there and come out and want to go and build big companies. Yeah, let's do it. Let's bring them on. Having been basically taught and trained with the resources of Google effectively. Talking of big companies, thank you for the segue. SpaceX. So they are doing a massive fundraise IPO. Lomax, set this one up for us because I think most people will know that there's big noise this week about SpaceX raising.
9:57But I guess my core questions are, what are investors really buying? And is space worth$1.5 trillion? So that's what's going on in the back of my head. But set this one up for us, Lomax. Well, I don't think it's confirmed. And basically the news has come out that SpaceX is strongly considering IPO-ing in 2026, so next year. Before, SpaceX has raised famous amounts of money in the private markets. Potentially, like many of these companies, why would you bother IPO-ing if you can keep raising money in the private markets? If you can potentially do secondaries to keep your employees getting liquidity so they see value, why would you bother incurring the costs and or going through the scrutiny?
10:35but they are they clearly feel that this is the right time as mads have said here before sometimes you run out quite frankly of money in the private markets and spacex still has a lot of money to to raise so they're thinking of raising 30 billion next year the valuation mooted is 1.5 trillion now i think the last valuation that spacex raised money at in primary was only about 150 200 200 billion they've done secondaries of five six seven hundred billion but this is clearly a massive step up what are investors getting dan well look i mean the golden goose within spacex is starlink which is a massive like monopolistic business growing fast and printing cash as far as i can tell now obviously spacex is private so not all the numbers you say that that they've only been processable last year or did they break even last year i don't think i don't think they broke even last year, but they've also got the launch business.
11:30So the numbers are generally conflated across and they're doing huge investments in Starship as well. If you look at analysts, like Starlink is projected in 2026 to do$16 billion of revenue and potentially around$10 billion of EBITDA. So it's turning into a real golden goose, that business. They have nearly 10 000 satellites in orbit eight million customers and growing fast and there was before this news always a possibility of them demerging and effectively you know starlink comes out of spacex and then does its own ipo but that that may be put to bed now what are you buying so you have that starlink business which is monopolistic printing money you have the launch business which is also largely now monopolistic right it's less profitable on a cash flow basis but it's still a very, very unique asset that they've built.
12:26And then you have this effectively exponential uncapped upside with SpaceX, which is the Starship program, which is these bigger ships that will do, you know, trips to the moon, to Mars, eventually deep space, which again, it's like, it's so far ahead of everybody else. They'll be the only people who can run this program. I think they get, you know, they're building this big factory now in Texas to create, I think, is it two Starships a week, maybe three a week or something? It's insane. Automated. So you end up with, what are you buying as an investor? You've got the hugely profitable and growing Starlink.
13:03You've got the launch business. And then you've got this uncapped upside with Starship. Now, 1.5 trillion, I don't know. I mean, if the business next year is going to do 20 billion top line, because Starlink is still about three quarters of revenue. I mean, that's a pretty funky revenue on, you know, it's nearly a hundred X revenue, a multiple of revenue, not of EBITDA, right. Or not of earnings. But then for the reasons I just said, it's Elon Musk, it's monopolistic. They have a huge headstart, you know, a bit like in a much smaller way, Revolu in Europe is just like absolutely tearing ahead.
13:40And it's just like winner takes all type position in its market. SpaceX is the same and clearly a much bigger, a faster growing market. Where did X and Grok and Tesla, are they all under the same banner now? Wasn't there a roll-up? No, not a roll-up, wrong kind of phrase, but wasn't there some kind of merge? Didn't Elon throw X into something else? Well, X and Grok, I think, are in the same holding. No, these are all very separate. One thing that was interesting is that I found that, I mean, maybe this is perhaps obvious, So SpaceX employs 13 ,000 people, which actually is not many. Tesla employs 125 ,000 people.
14:20Okay, so it's a big bit. Anyway, I think that's what you buy with SpaceX. Why is this relevant for Europe? Well, it's sort of, you know, you look at this and you think, well, we've missed the boat here. SpaceX has Starlink and launch. In Europe, it's more fragmented. You know, you have on the launch side, you have ISAR, who only did their first launch this year, which wasn't even successful. You have rocket factory Augsburg who need to change their name. And then you have on the satellite constellation side, you have UTELSAT who also need to change their name. It's horrible. But importantly, Europe is more fragmented.
14:56It's behind. It's capital constrained. And this is winner takes all. Yeah, we're going to talk more about Europe in a minute. Mads, where else would you take the SpaceX piece? Well, I've touched on a lot of the good pieces here. Clearly, if you look at the value of the core business, so the launch and the starlink connectivity, you can maybe get that to 300, 400 bill of value on generous multiples. So that's about a third of the price tag generously. And so what is the rest? And you're actually right. Starship is part of it. There's another thing, Lomax, I don't know if you touched on, which is the whole orbital data center thesis.
15:31You know, Elon Musk has built the largest data center of kind of AI factory on Earth, which is his Colossus. But the challenge we have is energy. And so there's this whole notion of, well, in space, you have solar intensity that's six times higher than on Earth. You have free cooling. If you think about the transmission of data, light travels faster in a vacuum than it does in cabling on Earth. And so is there kind of an orbital data center thesis you can buy into? And so this is actually quite interesting. You might say there's a bull case there around it, which is there's a ton of places on Earth where you are struggling to get the right level of compute today.
16:15Maritime, you have a lot of data coming out of oil fields and oil extraction offshore. You have defense, you have remote places where you don't have data centers nearby, but where if you can do AI inference in the sky and beam it up and down, that might be a great value proposition. And so smart money has been piling in. NVIDIA has been backing StarCloud. DOD has been having contracts. China has been launching compute constellations. And so if you think about Musk's position in all of this, well, he's already got the largest compute cluster on Earth over at X. He's got all the technology around batteries from Tesla.
16:50You've got the SpaceX launch economics, and you're well ahead of everybody else. Computing on the edge, you're probably still at least five to seven years away before the edge can catch up in a meaningful way. So there's potentially a window there. The criticism that some people are saying is, look, there's really zero revenue in this today. In terms of the compute, you can actually get to function in space. It's still quite small. The kind of power you need is massive. The amount of solar panels you'd need to create gigawatt, if that's the scale we have, is massive. I mean, you'd need more for one gigawatt orbital space center to them, the mass of the entire International Space Station.
17:34So these are potentially big projects, but it's interesting. And I think that's part of where the market is inferring some value to this composed SpaceX IPO. Which is, I think, by the way, I mean, a matter of time before we started talking about data centers in space. I mean, every time I open LinkedIn these days or Twitter, it's plastered all over. Largely in thanks to a British founder, if we want to take this back to Europe. So Philip at StarCloud, we talked about, Mads just talked about. who are now sending an H-100, no, into orbit, is British. He's doing a great job of beating the drum for this concept.
18:11Congrats to him. He's a Brit that also has mastered, found his voice and mastered narrative and storytelling, because I think he's doing a great job on that. Are you sure he's a Brit? Well, I know, I know. But interestingly, I do remember chatting to him, and he said he's basically building the company in Redmond in Washington, And he said, that's the only place in the world I could build this company. Yeah. Forget Europe. Forget anywhere else. Talent. The only place in the world I could build it. And then if you go back to the outsized venture thesis of backing European founders who often need to go to the US to build big businesses, you know, it's kind of bang on.
18:44Props to him. Props to my friends who invested in him. I think it's a great, it's a cool thing. And it's very, very necessary. Let's talk about your backyard, Lomax. So Portugal has won. Clap. Clap. Cheer. Rah, rah. Portugal has won. Economy of the year and the economist. So it's not all doom and gloom in Europe, although we are going to get a little bit gloomy in a second with some punches in the face from across the pond. But Lomax, tell us what's going on in your backyard. So The Economist every year run a little competition called Economy of the Year, where they screen on five key macroeconomic indicators, which is inflation, deviation from inflation targets, GDP growth, employment and stock market performance.
19:23It's a competition of the 36 wealthiest countries in the world. And last year, Spain won it. And this year, 2025, Portugal came on top. 2.4 % GDP growth, inflation control at 2.3%, stock market up 20%. The stock market, mind you, is only 100 billion in market cap. So it's relatively small. It's great. Look, I mean, they only use these five indicators. So for example, in Portugal, if you look at the average monthly wage net is about 1100 euros, which is like a third of what it is in the UK. So it depends what data you actually use. However, on these metrics, Portugal comes on top. I actually thought Poland might be nearer the top because Poland's actually GDP growth at the moment is nearly 4%, 3.8%.
20:07But I think they do badly on inflation and employment. The UK is woefully near the bottom. The US is in the middle of the pack. And both didn't do so well based on inflation and employment, actually, with the things dragging them down. And growth. I mean, the UK has basically no growth, right? Well, it's nice that the Southern European countries flip the script, right? This was not the case five, ten years ago. Well, I mean, do you remember the pigs? Portugal, Ireland, Greece, and Spain, the pigs in 2011, 2012, when they were, yeah. I mean, it's a big turnaround. The Germans were going to kick the Greeks out of the Euro and all this kind of stuff.
20:47I think things have changed. Yeah, I remember it well. So let's go into Europe. So it has been a mixed week, as it always is. There's been a bit of euro bashing. Trump has been savaging the EU, apart from obviously his autocratic buddies. China's flooding the zone. They've got a trillion dollar trade surplus of goods, which is probably going to grow in 26. So there's going to be a whole more raft of Chinese products coming through at a low cost. EU's answer of made in Europe, this kind of mandated made in Europe directive, or I don't know if it's going to be law directive. Well, there's been a massive backlash and I think it's going to be voted on the 16th of December.
21:25It's going to go back, which was effectively a motion to encourage slash force European companies to buy European. I remember when I was a kid made in Britain and the Buy British campaigns in the 70s and 80s. But the challenge you had was that the British products weren't very good and everyone could see Japanese cars, German manufacturing products from all over the world. So it didn't really work that well. But I kind of like the idea of a bi-European or a bi-British, if it comes to pass. So we've got China, we've got Europe being punched in the face, and we've got the US, obviously Trump turning his back on us.
22:03Mads, where would you start with this? I think let's start right where you are. The US are talking about this new national security and safety strategy, where they're saying Western Europe is just no longer a top priority for the US. They favor sort of the Central and Eastern European countries where they feel there's more kindred sentiment and migration and bureaucracy, etc. You know, the Trump administration is very keen to do a deal with, kind of to end the war in Ukraine and get, as they say, Russia back into the fold and do lots of trade. I think lots of people in and around the White House are eyeing opportunity for lots of new deals and more business to be done there.
22:38and kind of these pesky Western Europeans are standing in the way talking about, you know, outmoded things like freedom and democracy and, you know, not attacking your neighbors and all that stuff. And obviously, the Trump administration is not having any truck with it. At the same time, you've got China flooding the zone. We've seen for the first time a trade surplus of more than $1 trillion. And that's just in 11 months of the year. We've just never seen that before. and the US tariffs, they seem to not really be working because although exports on a direct basis might be down, they're up via other channels.
23:14And so overall, Chinese exports are just going from strength to strength and China keeps increasing the global export share. And it turns out when you spend decades trying to perfect the things you do and becoming the world's best at something, well, guess what? Lots of people want to buy your stuff. China's GDP on a nominal FX basis is still smaller than the United States, But on a purchasing power basis, it is actually larger. It's the largest economy in the world. And that means that if you look at the output, if you look at the things the country makes, the Chinese economy makes more than any other economy in the world.
23:51And so that's, of course, remarkable. It's also remarkable that there is such a big gap. On purchasing power parity, China's GDP is 35 trillion. But on the nominal FX basis, it's only 18 to 19 trillion. So that's a 35 % gap to the US. And the fact is that because you have such a big swing between the two, well, really people are saying this is because the Chinese currency is undervalued. And so the question is what would happen if that swung into place? Well, firstly, China would then have on paper the largest economy in the world, but it would also make their exports less competitive and make it easier for others to compete.
24:30So there's definitely contention there. And why can China do what they do? Well, it's because they're not, again, having this thing that's called democracy, right? They're not lumbered with that. We have been producing great things in Europe, but the challenge is the surplus we are creating kind of by the democratic institutions. We are now, people, voters have co-opted those democratic institutions to funnel so much of the surplus into the welfare state, into welfare, into things that are nice but don't necessarily fuel future growth. And that means that every year we just sink a little bit more into debt without our economy getting better, whereas the Chinese are not struggling with those pressures.
25:16They don't have democracy. And so all the resources are funneled into, you know, making the country more productive, keeping the currency down and exporting things and just winning more and more market share. So we're really caught in between the United States that no longer seem to care much about us and a China that is actually out to get us on a competitive basis and are producing great things. And we have to think about what does that mean when we think about how to position ourselves for the future. Some of the big trade surplus also stem from a lack of Chinese domestic demand for international goods?
25:51Well, of course, that's right. I mean, the two things go hand in hand. When you have a country without pensions, without a welfare state, people, they tend to save more because they are concerned that there might not, you know, if there is a rainy day, there might not be a safe connect to catch them. And all those savings are then funneled back into the economy or into exports, right? Instead of the Chinese domestic economy consuming the goods, well, they're exported abroad and sold to others. So you're absolutely right. And the only way to really fix this is you have to fix Chinese domestic consumption and get that up so they start buying more stuff.
Read the full transcript
26:24But how are we going to do that from Europe? Yes, really the question. I don't think we have many lemurs there. It sounds like a massive cultural shift. No, I mean, we also got beaten over the head by Elon Musk, right? We mentioned on the pod last week that the European Union had just fined X$140 million for infringements of our dear friend, the Digital Services Act. This prompted Musk to call for this dismantling of the European Union or the European Commission and how everyone's been betrayed, how we all need to throw resource weight behind the real parties that say how it is which is the you know the afds of this world how orban is great how maloney is great i think actually maloney is a really good bridge potentially between the eu and and the us and is much more credible than than orban but yeah with the us like uh defense strategy the nss the national security strategy and musk's not curveballs but his his grenades that he's been throwing in off the back of this big fine.
27:32It's been a pretty chastening week for Europe, I would say. Could I pick up on that for a minute? I mean, one of the other things that was discussed was this lack of a single lock in financial services. And there were sort of talks out of Luxembourg about how they've probably been blocking the idea for capital markets union, this able to create a monster. We don't want central regulation. The flip side of that is instead we have, you know, dozens of regulators today and no single market in financial services. I just mentioned it. I just mentioned it. Like, you know, Portugal has a stock market of 100 billion market cap.
28:06Like, what's the point in everyone having all these things, no? Yeah. So there is the stock markets, but there's also the ability of fund managers to raise capital across the continent. To use a single set of rules as opposed to being forced to use the national private placement regime, for example, in each country where they are long to do business. That's a huge thing. And in fact, specialists are saying that this is costing European business as much as$100 billion or euros every year in lost productivity just for all those trade barriers we're facing internally across Europe. That's a huge problem.
28:42And back to Musk, you're absolutely right. He's lobbying grenades. One of the things he did say was the European Commission and European president should be elected by a public vote. And there are good reasons for why there are not. but there are also good reasons for why it's probably the only way out of the mess. Because one of the reasons why it's so difficult to harmonize is because there's no real democratic or public mandate behind it. Who is the person that's going to stand up and champion this? And today it is, yes, sadly, unelected bureaucrats. And why are they unelected? Well, that's because if you are a national regulator or a national politician, the last thing you want is for somebody being centrally elected to have a bigger mandate than you to actually say, no, guys, we've got to think about this, not just from our own backyard, but think across the continent and think, how do we do things better?
29:33One of the benefits of the American model is you've got constitutionally, you've got the federal government is in charge of commerce. So if you have something that falls under trade, well, it is the purview of Washington. And until we have something similar in Europe. It's hard to see how we get the last bit of harmonization done so we can enjoy the benefits of a true common market, a true single market to benefit business and therefore also consumers. Is there any way that could happen? Look, sometimes it takes crisis. Sometimes things have to be worse before they get better. We've had decades of lots of progress around European integration.
30:13Then we've had decades of standstill. I think Brexit really took out a lot more of Europe over the last decade than maybe we sometimes realize, not just the UK, but also everybody else that were busy trying to make it all work with the UK, leaving the union. Yes, I think there's a scenario where this could happen, but possibly things have to get worse before they get better. It goes like a lot worse. Very quickly, chaps, defense spending in the UK. I saw some reports this week pulling that apart. Lomax, have you got any insight on the expense? I think this is a small point. And the small point is just like, let's just remember this.
30:47because there are lots of people in our industry at the early stages. I think some people, some people are very smart, some people running around like headless chickens talking about defense and wanting to throw money into defense startups. And that's the current VC consensus. As we know, consensus in VC is a dangerous place to be. But the point is this. Everyone has been excited about this because they see the opening up of defense budgets in Europe prompted by what's been going on in Washington and what's been going on in Ukraine. However, and prompted by the actual commitments from these politicians to say, we are going to increase defense spending.
31:25In the UK, it's to, by 2027, to 2.5%, right? At 2.5 %? Anyway. Sounds about right. And potentially to grow that to 5 % by 2035, which is just, you know, so far off that I think it's meaningless. but the but what you're hearing from the ground from the armed forces is actually that these budgets are not you know are not opening up now as quickly as you would like which basically means that we're still kind of stuck and so actually startups selling into this you know new government money are going to be waiting around for a bit which you know especially when startups are you know raising raising money every 18 to 24 months doesn't really work so much on the startup time so i think the point is the messaging from the armed forces in various countries is that this money is a bit slower to come through i mean they're probably maybe they're always going to say that but i think there's been a bit of inertia before though hasn't there there's been no but if you look at the uk i mean yeah exactly yeah there's like they'll grab the headlines with the with the commitment, but the commitment will be down the road and it will often be fudged.
32:35Word of warning that. Yeah. Okay. Well, let's move into AI corner where I would imagine there's going to be some more kind of defensive resiliency conversations potentially. So another kind of week of AI bubble popping conversations in the press. We've got private and public capital is moving out of big tech moving out of AI. We've smelt before, we've seen before. Mistral, some new stuff out of there, Mads, Nvidia can now sell H200s to China, but do they need them? And obviously there's now a kickback mechanism for Trump and his cronies, but China's developed a new chip company called MoreThread.
33:14So loads an AI corner. Maybe let's start with the AI bubble popping, not popping. Yeah. I mean, look, people have been talking about AI bubble for months now, and this This week, we've seen UK pension funds saying, look, we're going to cut US equity exposure because we have bubble fears. We know that the Magnificent Seven, I think they're like 25 % of the MSCI world. There's an incredible concentration in those large companies. We've seen Standard Life reducing US exposure and boosting UK. We've seen Aon also selling some equities. Many of the pension funds have been out and saying, look, we're going to lighten kind of our exposure a little bit.
33:51But we also know that NVIDIA has been doubling since April and really nobody wants to miss out on this rally. So the sense is, look, institutions might be hedging a bit, but they're certainly not fleeing and we're probably not in a sort of overhang of a bubble yet. I think this rally has got longer to run, even if we're in a slight risk off period right now. I think we're going to see 26. I think we're going to see further upside. Obviously, this is not investment advice and always do your own research. What's happening? So we've got Meta now going proprietary, Mistral, open sourcing. What's happening over at Mistral?
34:31Yeah, this is the good news of the week. I mean, we've been a little bit hiding away here in Europe on the foundation model for most of the year, actually, since DeepSeat came out about a year ago. It was Christmas time. We saw the first D3 model and then their reasoning model a month later. and Mistral, that was a pioneer of open source AI models. They released Devstral 2, which is a new open source model focused on development and coding. And the benchmarks are stunning. Very, very strong, performing, you know, almost as good as some of the Anthropic models, but obviously much more cost-effective.
35:06It's an open source model. On Schumann evals, it's beating DeepSeq. It's got, you know, relatively light. It's about 123 billion parameters, so much lighter than some of the Frontier Lab models. So a lighter model, easier to run, a big, nice context window. And it looks like our friends over in France at the Mistral team have really been pulling out all stuff to create a phenomenal model here. So great to see you're back in the game on the open source AI front. I thought they were moving to the product side and moving away from the foundational side. Not that we would ever know what goes on behind closed doors, but this is a...
35:42They are doing that too in the sense of they're doing more work with enterprise. We saw the big deal they did with ASML earlier in the year, and they're going to be doing phenomenal work there to implement AI models and help the ASML team become even more world-class at what they do. So loving that. But the fact that they're also out there in the open source game and playing that game and sort of in with a shout is great to see. Meanwhile, Meta, they've been pivoting. Going the other way. They've gone the other way. They're talking about a new closed model. I think they're calling it something as silly as avocado.
36:22So DeepSeek's model copied Lama's architecture. We know that Lama 4 then turned out to be a flop and that triggered a major shakeup in the organization. It seems that, you know, they've been piling so much money into CapEx and they're probably now thinking, how are we going to recover some of that spend? that are now looking at making avocado not an open source but a closed source model so the irony is really here while while europe is back in the open source game meta seems to be pulling out meta being the original champion of open source so yeah all the pieces are moving around with the chessboard i'm intrigued to see where yan lakun ends up and what he ends up working on and i i hear that he's back in europe opening some labs so it'll be really interesting to see how that unfolds and nvidia can now sell h200s right so nvidia can sell trump gets his kickback there's 25 kickback there but china has invented through more threads their own chips what's happening here well when there was the trump she kind of china us deal done you know a few weeks back i think we said at the time look they're probably going to find a way to get the nvidia chips open up for exporting those again because Trump wants it and can see a way to get a cut out of it.
37:36And lo and behold, they have. The Chinese, meanwhile, have decided that actually we'd rather grow our own homegrown silicon semiconductor infrastructure and economy. And so they are kind of Huawei, they've got their Ascend chips and we've seen other more threats have surged in IPO. That's another kind of X, kind of NVIDIA spin-out team that's created there. So there are new competitors springing up in China. And, you know, I absolutely think there will be NVIDIA chips coming into China because they need them and want them right now, the entrepreneurs and the corporates that are building AI. But it also seems obvious over time that China, they are absolutely focused on creating a domestic economy and domestic capability.
38:20Own the stack, right? Own the full stack. You talked about Otaki on our last part, and I think that's absolutely what we're seeing specifically when it comes to semiconductors. Yeah, why not? If you can be self-reliant, then just own the whole stack. Well, not only that, think about how the West has weaponized semiconductors. This is the rational response. Well, we're living in a no-trust society now where we're all decoupling, lack of trust. Everyone's going back to their own sovereign capabilities. I totally get it. We reap what we sow. Anything else that caught your eye this week, Mads, in AI Corner?
38:56Maybe if you want to go to deal of the week, there is the unconventional AI. They've raised a$475 million seed round. What is seed anymore? Who knows? Who knows? Well, I mean, look, the company is two months old. So it's not like they've been around for years and years and the rents have raised a Series F. I mean, it is literally a seed. It's right off the block, but it's very, very exciting stuff they're doing. Again, in semiconductors, kind of the view is, look, the traditional semiconductor, The traditional way to make CPUs is very energy intensive. They want to take something they call, I think they call neuromorphic computing, which is effectively saying, look, if you think about the way our brain works, it runs on about 20 watts of energy, which is just a fraction of what a GPU needs.
39:44So if we can make processes that run more like biology and less like the syndicates we know today, we can make them far more energy efficient. And the efficiency gain they're targeting is 1 ,000-fold. So, you think about some of the stuff we're talking about with data centers in space and decentralized computing, places where it's hard to get energy. I mean, it's kind of that topic again, right? Energy is the bottleneck for data centers today. It's the bottleneck in space. Can we fix that by creating much more efficient semiconductors? And that's exactly what unconventional AI is targeting. Very nice.
40:20Hello, Max. What was your date of the week? I actually have three deals of the week. I'm very, very greedy. You greedy swine. They cover different areas. So one is a straightforward series B,$40 million by Sol Intelligence, which is a legal tech company building the operating system for IP by leveraging LLMs to revolutionize, as they would say, the patent lifecycle. So basically it's LLMs for drafting patents. YC company, three British founders. It's now kind of UK-US, which makes sense given how big the legal market is in the US. Mr. Stebbings at 20BC has led the Series A and he's now teamed up with the good chaps at Visionaries Club to lead the Series B.
41:05It's an interesting company. I mean, it's a classic use case for LLMs, isn't it? Like patent, drafting patents, you know? I'm surprised there's not a company sort of similarly doing this for grants as well. In fact, there probably is those kind of things. I think the other deals of the week for me are one is a commercial deal and one is an exit relation therapeutics has signed a big deal with Novartis so relation therapeutics another AI for drug discovery company as we talked about earlier signed a deal 55 million dollars up front which is a pretty big deal when pharma is doing deals with these platform drug discovery companies potentially big royalty payments out the back end of multi-billion relation is a very interesting company to track it's a london-based company it's raised money for some pretty marquee international investors including costler nvidia dcvc really really good team i think it's a company to watch there's a long lead time on this because they're basically you know developing you know assets and drugs but i think it's definitely one to watch and then hot of the press today is the exit of go cardless by molly did effectively it yeah i mean it's not cash exit it's actually a kind of share for share.
42:13It's a combination, effectively, a merger. Two companies in the same industry, in fintech and payments, merging together. So it's a paper deal, basically, to try and give themselves scale to go to market and build a more global company. So I think that's an interesting, literally hot off the press today. Because GoCardless was founded in what? 2010, 2011 or something? It feels like in dog years, it feels like an ancient, an old school business. Yeah, but, you know, these things take time. Yeah, baby. Well, New Quantum have been no slouches, so I'm pretty sure they're a Cambridge spin-out. They are.
42:50They are. Yeah. And they have just announced their 60 million Series A, so their Quantum Networking just announced their 60 million Series A. So I wanted to give them a massive shout-out and say a good speed to you, my dear Cambridgeites. Well done to you. If you want to look for the upside, as we've talked about this before, but in quantum the uk is a really really interesting place yeah i mean europe quantum fusion europe feels feels quite substantial because there's lots of activity in both across europe so but in quantum yeah i mean clearly in france but in the uk it's very very interesting and even some of the biggest us players came out of the uk so i think um more to be more to be seen here all right my lovelies well that was it short and sweet i'll catch you wonderful people next week.
43:38See you on the flip side. Bye-bye. Distribution decides who wins. And in today's market, the teams who communicate clearly, the ones who own their narrative, are the ones who get ahead. At EUVC, we've spent more than half of a decade helping Europe's best founders and investors shape stories that open doors to capital, customers, partners, and even talent. To help more European teams communicate with impact, we're giving away one free bespoke communication strategy session led by one of our sharpest comms experts. Get ready to sharpen your narrative and strengthen your distribution. Scroll down to find the link and apply now.
44:19One lucky team will be selected.
44:30I'm sorry!
From the publisher
Welcome back to another episode of Upside at the EUVC Podcast, where Dan Bowyer, Mads Jensen of SuperSeed and Lomax Ward of Outsized Ventures gather for a holiday-home special to cut through the noise around Europe’s tech, geopolitics and AI shifts.
What begins as an innocent debate about whether DeepMind is “still a UK company” quickly spirals into a tour of sovereign AI strategy, the SpaceX mega-raise, Europe’s increasingly uncomfortable place between China and the US, defence-spending reality checks and a surprisingly uplifting set of deep-tech deals across the continent.
It is classic Upside: the takes are sharp, the geopolitics gets spiky, and the optimism… well, it arrives eventually.
What’s covered:
04:36 AI-for-Science, robotics and the new “AI scientist” era
06:50 A national-curriculum Gemini and the vision of a tutor for every child
09:39 The SpaceX 2026 IPO: what investors are actually buying
14:00 Starship, orbital compute and the trillion-dollar imagination gap
18:07 Why Europe missed the space race once again
19:43 Portugal flips the script: “Economy of the Year”
22:58 Europe between China’s export tsunami and America’s cold shoulder
32:07 Defence budgets: the hype, the delay and the reality for startups
34:25 AI Corner: bubble fears, Mistral’s comeback, Meta goes closed, China goes full-stack
Comms Strategy Expert Session
Apply or share the opportunity with a founder or investor in your network: https://luma.com/euvc-comms-expert-session




