E428 | This Week in European Tech with Dan Bowyer, and Lomax Ward

0:00Welcome to Upside, where we dig into the real stories behind the headlines affecting European venture. This week, it's Lomax from Outsized with Mads and myself from SuperSeed VC. And we're talking UK Doge, Project Europe, IPOs and M &A up, or are they just as the American markets melt? Plus, we've got the latest in AI and deal of the week.

0:26Tear down this wall. It's more than just an alliance.

0:40This show is not investment advice and the hosts of this episode may be invested in the funds and companies featured. So Mads, you were a bit of a traveling Wilbury last week. What's going on with you? What are you up to? Yeah, it's been great to come back to London this week. We've had some super cool boards. We back early stage kind of seed and sometimes really early stage kind of pre-seed companies. And we backed a few kind of very early stage businesses a few years ago. And it can take some time. It can be quite hard work for those teams to fight their way from kind of those early ideas in the early tech to actually launching into the market and starting to see the revenue ramp.

1:17And it's just, we had a couple of meetings this week of folks that after some years of really hard work, things are really starting to come together. And it's just such a fantastic thing to see. I mean, I'm so, I walk away from those meetings and I'm so excited in, you know, what founders are doing and what we're doing in our industry. It just, it's a great thing. Yeah. It's a long-term game. That's for sure, though. It's a long-term game. Got to be patient. Lomax, what's cooking with you? All good. meeting founders uh working with the founders in the portfolio personal wise all good another clear health scan this week so you know onwards and upwards yeah celebrate that and yeah all good how's the office on the beach how's the office on the beach the office on the beach is great we have wi-fi now we have a big table three chairs and we are going out for partnerships as we scale it up to um the surf lodge on the beach for you know founders gps and lps in europe come join come and work.

2:15I'm going to. I mean, it's nearly the season for Lisbon, right? It's middle of March and we should be there. This week, what's been on my mind, I saw Northvolt went finally, finally, after their Chapter 11 debacle and they finally closed doors after raising$15 billion. That was sad, interesting, thought provoking. I've been speaking to a bunch of founders this week. Nearly all of them shouldn't be raising venture capital. So I've been talking with them about what they should raise. And we've been discussing and noodling what if there's something that lives between Angel and VC and what that looks like.

2:48So that's kind of been a really interesting discussion week on funding types. And I've seen that the Food Standards Agency in the UK might be warming up to the prospect of lab-grown meat. So I've seen that the whole lab-grown meat, which kind of fed out of favour not so long ago, might be back on the table quite literally. league. So those things I found quite interesting this week. I think that's interesting, by the way, for a number of reasons. I think that food tech has definitely been challenging in the last couple of years. But I think this is an example of the FCA, the regulator actually taking the bull by the horns, running this sandbox program, which has been very, very successful within fintech.

3:27And we now see that for food. And it's an example of the UK being a little bit unfettered by European regulation post Brexit. Is that what you think is driven? Is that what you think is I don't think it's driving it. I think the UK has a bit more room to operate now. That's it. As you know, I'm not pro-Brexit. But in that respect, I think it's an example where the UK is trying to be a bit like Singapore, which has really taken the lead on early regulations for food tech. They have to, right? I mean, you've got to take something out of Brexit. It's been such a freaking disaster on every single level.

4:00We've got to take something good out of it. And if this kind of ability to be a bit more free form with regulation, I think, is a bad thing. And actually, Foodtech is a place where Europe can be a leader, already has startups, which are some of the best in the world, right? So actually, look at companies like Most of Meat, Gourmet in Paris, Hoxton Farms in the UK, all going into this sandbox. These are actually world-leading companies. So it's an example where EU could actually lead the way. Well, we're going to stay with the UK. there have been some interesting news reports stories kind of ideas coming out and almost wrapped into a UK doge so are we seeing the start of one has Elon given us the permission slip or the broader permission slip for governments to start cutting this week Pat McFadden who's the I love this title who's the chancellor of the duchy of Lancaster now I didn't realize I thought that was just a schmancy title.

4:57But actually, he's the most senior cabinet minister after the PM. So he was recently talking with Laura Koonsberg at the BBC and talking about cutting across the civil service. So contextually, the civil service employs just over 500 ,000 people and costs the UK taxpayer around 16 billion pounds annually. Interestingly, the UK government spent an additional$7 billion last year on temp staff. So would cutting make the difference? I know that 2016 was the low point in civil service employees and mads. You're going to go into some of the more detail in a second. And one other thing that caught my eye was that public sector pensions looking forwards, that bill has been estimated to be anywhere between£2.5 and£5 trillion.

5:44pounds. So that's a big, big chunk of change. So lots of noise and movements on the cost cutting and the civil service. Will AI play a part? Will startups and the investment community get involved in here? Mads, you've got some more numbers, thoughts and drill down into this. What's going on behind the scenes? Yeah, look, it's the good old, I guess, if something isn't sustainable sooner or later, it has to change. And it's clear that the public finances are not sustainable in the UK, in Europe, in the US. We've been running our economies in ways that might have worked in the past, but certainly don't work now or on a look forward basis.

6:24It's sometimes said that it's easier for the Tories to raise taxes and it's easier for perhaps for Labour to reform public services, because to some extent, you know, those sides are not expected to do that. And so when they do it, it's sort of, you know, They had a little bit of air cover, perhaps. And I think we saw some of that today with Kiesdama effectively saying, look, we are going to close NHS England. This doesn't mean that the NHS as an institution is going to close, but it's a massive reform. And it's really taking on one of the most powerful guangos in the UK. If you look at the UK government, it's got about 6 million employees.

7:03So more than 1 in 10 adults are employed by the government. NHS alone is more than 2 million people. Local government, another 2 million. And then the civil service itself is more than half a million people. And that's up a lot. You mentioned 2016 as the low point or the high point, perhaps, you could say, depending on how you look at it. Just nine years ago, there were only 384 ,000 civil servants. So we're up 130 ,000 in nine years. That's 34%. It is nuts. Yeah, I would say, Matt, sorry to interrupt you. If you take the 6 million, right, which is central government, local government and civil service, actually looking at the data recently, it's been flat at about 6 million.

7:47I was shocked by this for the last 15, 20 years. Now, clearly, we had 10 years of austerity from 2009 to 2019 under the Tories, and they would have reduced the numbers and they've come back since then, which is probably what you're referring to. But I was actually staggered by that. Yeah. So I think what you're saying is it's at a similar number now as it was 15 years ago, but it's been much smaller in the interim. And what you've seen is that there's been a massive increase in the size of the civil service, up by a third. A lot of this has been driven by Brexit because we've got all these value-adding people that we've now had to employ to think about how we can put customs and controls and all these things we didn't used to have, which, of course, are not creating any value for anybody at all.

8:36At the same time, we are seeing productivity in the NHS down 20 % in five years. So although we've been pushing a lot more money into the NHS, we're actually getting less out of the NHS in terms of productivity when you measure on the operations and the procedures and the actual health outcomes. If you think about the net, the UK tax burden is the highest since 1950 and is projected to increase further to World War II levels. Meanwhile, welfare benefits and pensions, they cost more than 300 billion pounds a year. Today, one in five people are retired, and this is projected to increase to one in four as we get closer to 2050.

9:20So it's obvious that the country is broke if we don't take action. And it's just as obvious that we can eat more productivity out of all the money that goes into government. Well, how do we know that? Well, that's just because just a few years ago, we were able to deliver more with less. So I think the good news is AI is an incredible opportunity. It's the opportunity of a generation to transform the public sector. Starmer, Reeves, and Peter Kyle, who is the head of DSET, Secretary of State there, they all say they want to do it. And we'll see in the spring statement, which will come in a few weeks, March 26th, whether they are going to put some tangible steps in there around actual cuts, because they will have to find a way to make efficiency gains.

10:07Otherwise, the numbers ain't going to stack up. Lomax, what can they cut? Where can startups get involved? What can we do? okay this is interesting by the way because hot off the press the government has announced today that they're going to find 45 billion of spending through efficiency drives using ai literally peter kyle i think mad's your friend you're with him recently right um has announced this right so i think that's a very interesting and a higher number than i would have expected right out of context the government spends 1.2 trillion every year they think they will find 45 billion of savings on an annual basis right now what's interesting before that was announced today that was also announced at the same time as getting rid of nhs england which is part of these efficiency drives um i've been thinking about this because on this podcast before i've actually been very skeptical of doge i think you've i in terms of the outcomes that it will achieve and actually funnily enough if you look at the u.s numbers recently u.s government spending has actually gone up in the last two months despite doge but we all know in an organization when you conduct a RIF, you know, reduction in force, you know, start laying people off, your short-term costs tend to spike and then, you know, will go down in the long term.

11:19So I wouldn't read too much into that. But what I was looking at it, I think Mads made a very good point that the UK spends 350 billion on welfare, welfare benefits, pensions, et cetera. You know, that's an absolute number of like cash transfer to people effectively. So, you know, that's a quarter of the spending. So in a way you can't really automate that away, right? I think one of the obvious places you could um use ai to drive efficiency in terms of cost is the six million workers that we just talked about right so did some calculations and i roughly worked out that we're spending about that payroll is about 220 250 billion a year so let's imagine you can reduce that by 10 percent right so therefore you'll find that will save you that will find you 10 20 billion of savings right Right.

12:06And then one of the other things I've been thinking about, the problem with that from a political point of view, clearly, it's very, very unpopular. Right. You know, you're you know, the politics, it's all very well for us to say you run this like a business. But actually, politically, that's going to be very, very hard for the government to do. And don't don't remember that. Don't forget that Labour, you know, most most of the people who work for the government are actually Labour voters. Right. So they're actually cutting off, you know, for them, it's very hard for them to do it. and also politically just to get it done.

12:34And also, of course, it's not an absolute saving because when you lay people off, if they don't find another job, the government has to pay benefits to them, job seekers allowance, et cetera. So there's all these complexities that go into it. But clearly there is ways to streamline things. But there are things that I was thinking of that actually AI could really, really help with. Benefit fraud. 10 billion a year, the UK loses to benefit fraud. I bet you with more efficient screening, tooling, et cetera, you could save a lot of money. You could a lot of money through using AI and trying to prevent for that and screen for that.

13:12So when I did the numbers, I thought I could get to 15, 20 billion a year, you know, with more efficient, more efficiency in the NHS, et cetera. Clearly, the government thinks they can do double that, which is great news. They point to 45 ,000 letters opened every day by the DVLA, 100 ,000 phone calls received by HMRC every single year. So clearly, if those. Yeah, which is pretty amazing. It's good. It's good. It's funny. You watch Peter Kyle. I just watched him before this, before this, before us getting together. And it's funny. He's almost talking like a tech founder, like talking through. Like how automation can actually impact your business.

13:48And of course, he's presenting this to the UK taxpayer. and who you know non-technically initiated so it's very very dumbed down but it's very compelling and actually you know this is not new though and actually you know you know i don't know if you came across this um gov tech accelerator public.io which has been in london they set it up in 2017 a friend of mine set it up god i've known for a long time they actually sold it this year and that was an accelerator working just on gov tech initiatives and actually i wonder whether we should get them on the show because um you know to actually um listen to they will know the next level of detail than us about this uh but i would say i was surprised pleasantly by the 45 billion that they think they can find because it was higher than i back of back of the enveloped and um of course let's see if they can actually do it and it will it will be difficult for them to do it because they're going to expend political capital to drive this through but let's go for it now's the time as mad said you know we need to yeah i think it's uh i think it's super exciting i want to stay i want to stay really positive and upside true to the name of the pod because i want to talk about project europe so this is an initiative by harry stebbings of 20vc fame it's a european incubator for under 25s to stop the brain drain to the us they're offering 200 000 euros to kickstart a startup idea which has to be a difficult to solve technical problem i feel that this is incredibly timely.

15:17I think they've absolutely nailed the timing of this. It is time for European dynamism. So I absolutely love it. It's also great deal flow for Harry. So massive, you know, if I was in a high school gym, I'd stand up and do the slow clap about now because this is bloody genius. His mentors and investors in the program, he's just turned the who's who of founders of European startups. I'm Matty from Eleven Labs, Stefan from Synthesia, Sebastian from Klarna. He's turned these kinds of characters into scouts for him. So it's bloody genius. So I guess my only key challenge here is for this whole program is how will it stop the US coming in from hoovering up the best and shipping them off to the States anyway?

15:58So it kind of makes the whole program a bit so what? So we've got to fill that, you know, growth funding gap to get them out of this incredible, you know, positive European initiative and get them to stay European, if that's even in any way possible. But Lomax, Project Europe, what do you reckon? I think it's a great initiative. And I think that, you know, it follows the precedent of the Peter Thiel Fellowship in the US, where they were giving, I think, 100 grand. Admittedly, that was in a grant, not this program, Project Europe. They take 6.6 % equity for the 200K. But actually, looking back, I did some, you know, reflecting on the Peter Thiel Fellowship and what's that achieved?

16:36What's been the outcome of that? Right. Because that's been going for much longer. interestingly vitalik from ethereum fame and dylan field from figma which was a you know is a big big company that um we're going to talk a bit about later you know virtually exactly they both went through the program so you know clearly um you know that's a that's a great result if if the program is the reason why you know they ended up going on to do what they would do i would say you know this is a great great timing great initiative um you know the only thing you know let's give a balance for you on this right the only thing i would say is very bbc is just in in a way they're using their platform to to solve something that actually doesn't need to be solved like precedence that we don't need more precedence seed funding in in in europe we need is as you as you do the growth funding but that i totally agree with but do you not think there's space for more early like founder dating pre pre pre my sense is there's a lot there's still a lot of money there and the second thing i'm not again i'm not never saying no to more to more money this is great but i'm just i'm just making a point and the other point i would make is which is why i'm always that this is me as an investor always a bit skeptical of accelerators and incubators because i'm like oh you know is this actually making a difference in the sense that if there are people out there young enough crazy enough like foolhardy enough to jump into building a company does you know does this kind of you know this kind of program is it needed i mean they're going to do it anyway is what i'm saying do you know what i mean like i don't will this actually change but for this program you know will there actually um be more um be bigger outcomes that's the only thing i was saying but you know if one individual in their 20s ends up founding a unicorn and they don't take a job in a bank then that's a good result for the 10 million invested here right so it's a very low um low bar in terms of roi i'd say i just love the i just love the pr of it i love yeah you're right and sentiment it feeds exactly this there's absolutes and there's me being an anorak about it but i would say there's the sentiment and and you know if if this kind of thing was happening every week every month in europe and by the way we're seeing more of these kind of initiatives and you know we have the eu company initiative now to try and solve that kind of structural problem like this stuff had been happening in the u.s over the last 20 years and you know actually in a way like we're starting to drink a little bit more of the of the um what's in the water in the u.s i'd say so this is great i'm loving that and um mads you you know a bit more of the history and you're going to give us a bit of a history lesson now i was going to go down memory lane for a little bit because no because i think so here's what's interesting.

19:26Venture capital was invented in the US, in New England, but not by an American, by a Frenchman, Georges Doriot, who was an immigrant, moved to the US and was a professor at HBS. And he set up the American Research and Development Corporation in the 40s. And that really was the first institutional venture capital firm. And his first big deal, kind of the thing that made the industry was he backed DEC, Digital Equipment Corporation, in 1957. And he bought, he put$70 ,000 into DEC for a 77 % stake in the company. Cool. And when he, it's very punchy, right? And when he exited a few decades later, it returned$355 million.

20:17So it's 5 ,000 X return. Can we have one of those, please, Matt? Just the one. Just one. And obviously that got people excited. But that was all on the East Coast. That was New England. And so what about Silicon Valley, which is the thing we associate with venture capital today? Well, that all started with Fairchild Semiconductor and the Traytras 8 that broke free from Shockley. And that was also in 1957. And they were backed by Arthur Rock, who was the first VC investor on the West Coast. But you're in the 50s there. And the first truly institutional firms were set up in the early 70s. So that was Kleiner Perkins that invested in Google and Amazon and Sun and Sequoia.

21:08We all know today investors of Atari and Apple and Cisco and Oracle and anybody you can think of. And it was Don Valentine who set up Sequoia, who was an ex-Fairchild employee. He had been in sales there, and he understood this new tech industry extremely well. So we are going back to the 50s, 60s, and 70s. And that's when the VC industry took off in the States. Now, George Dorio, he eventually came back to Europe, but not to set up a venture capital firm. He thought what Europe really needed before anything else was the equivalent of… Regulators. We need more regulation. Well, don't forget who he was.

21:51So he was a business school professor, but he was also a venture capitalist. And he wanted to set up the equivalent of HBS. And that was what became INSEAD, which he set up in France. And it's a fine place, and it's a truly European institution, and it's been very, very important for European business. But because his focus was there and not on VC, that just meant that Europe was extremely late to the venture game. And it wasn't until 1996 that the first really true, what we call modern European VC firm was born. And that's Index. Right. So we are 25 years behind Silicon Valley in terms of developing our ecosystem.

22:38And I think this is what we sometimes forget. In the 1990s, Europe produced about 5 % of the global tech enterprise value came out of Europe. But in the past 10 years, that's increased to 17 % as our VC industry has matured. And we can do so much more. And I think the initiative here from Stebbings is a wonderful thing. For me, it's just about we need more. And especially we need more of the cyclical things you always talk about, Dan, with successful founders back in the next generation of founders building the ecosystem. And that is what this is about. That's why I want to keep them. That's why I want to keep them and get that recycling, recycling.

23:23Absolutely. Otherwise, it is super brain drain. In fact, Project Europe will totally backfire because everybody will learn their true skills, get that early stage funding and then disappear over to the States. which I hope doesn't happen. But I want to move on to AI Corner, another one for you, Mads. So we want to talk a bit about Anthropic. They've obviously just raised, we've talked about that before. They've released their latest revenues. I want to have a look at Manus.im. That looks like a super cool operator-style agent. Looks very impressive. Larry is following Sergey and getting off the bench, looking at his new gig.

23:55And Microsoft is competing in its own backyard, which might be a bit problematic with OpenAI. So Mads, over to you and AI Corner. Yeah, thanks, Dan. One of my favorite segments here. You know, I'm a huge Anthropic fanboy, and the business has just been running, kind of running on all cylinders and firing away. They've grown annualized revenue from$1 billion in December to$1.4 billion in February, so 40 % up in three months. They've had a very strong launch of Claude 3.7, the new Sonnet model, and Claude Code, and it just keeps getting better. And what's super interesting is in parallel, we've seen Manus released last week.

24:36It's a new agentic platform. It's a little bit like some of the things we've seen from OpenAI, but it's just very, very powerful. I was having a play today. It's brutally interesting. I've not even scratched the surface. It is. It could do really, really cool things. Effectively, just for context, you give it some work, a project, and it will go off and do a two-week research project on a company in an hour and produce you with just a beautiful, beautiful output. What's interesting is the platform has come out of China. So China is making a really, really cool application. And what's under the hood, under the hood is Sonnet, Anthropics model.

25:17But they're open sourcing them, aren't they? They're open sourcing this, right? Well, yeah. Well, obviously Anthropics model, they can't, but a lot of the platform itself, they are. And so it's just wonderful to see, you know, all the innovation that's happening and how everybody's learning from everybody else and leveraging and using technology. And, of course, you know, as a kind of with our outlook, right, being a techno optimist and being, you know, having a liberal outlook of the world. I love when you see, you know, the West learning from China, China taking the best from the West. Can I borrow that?

25:54Sure. Yeah, look, they can't steal the Sonic, they could possibly steal the Sonic code, but I think actually what they've done here is they've leveraged it fair and square, which I think is beautiful. You spoke about Larry Page and Google of Google fame. He came out of stealth with Dynatomics. It's a company we've been aware of for a little time, but it was finally released, kind of the scoop was released by The Information, which is the Silicon Valley blog. Now, what's so interesting here is He wants to build a new AI company in manufacturing, and it's all around creating models for manufacturing to help make manufacturing better.

26:33Now, obviously, Larry Page is the founder of Google, and they have a fair amount of AI power inside that business. And so the fact that he is involved in a new business outside of Google to build AI specifically for manufacturing is evidence that even the frontier labs don't think that these foundational models can solve all problems. This is the thing we've been saying for a while. You're going to see a lot of specialized models. You're going to see a lot of applications because there's so many things you can't do with a general model. So I think that's an interesting development in the industry.

27:09Why do you think it's not being done within Google, Mads? I don't think this is on Google's focus or model. I mean, so when you look at what Google's business plan is, for them to be involved in anything, they sort of have a rule of thumb. It has to have the potential to get at least to 1 billion users. And that's what you have with Search and YouTube and Android, etc. So a billion is the number of users they're looking at for their business model. And I think they've looked at Dynatomics and they've looked at manufacturing and said, well, it's not that. It's a completely different thing. And so we're going to develop it somewhere else.

27:44But it's super interesting that both Larry Page and Sergey Brin have come off the bench for AI. They don't need to work. They don't need the money. So I find that this is obviously such a transformational world. They're stepping back in and starting new businesses. I think that's fascinating. As anybody who is in the industry can see, that we are living through the most exciting time, at least since the 1990s, if not forever. The speed of development is just so fast. So this is what's exciting us every day, right? Everything that's going on. Just a few more things to wrap up on. So Microsoft had a very close relationship with OpenAI.

28:23They were a huge early backer of Sam Altman. They are experimenting now with DeepSeek in their new model, specifically in their co-pilot. It's potentially a blow to OpenAI. We know OpenAI has gotten a lot closer to SoftBank. So interesting to see how these alliances are moving back and forth in Silicon Valley. On the home front, or at least closer to home, across the channel, Mistral looks like they're pivoting from being an LLM company to being more of an application company. Interesting. Where Anthropic has been perfecting coding, Mistral, they've been tackling OCR and specifically tables. And what are the places in which we need lots of tables?

29:07Well, that's finance and that's accountancy. and so we're not investors, we're not on the board, we're not privy to any sort of inside information, but it's interesting to see how they are repositioning kind of what they're doing and what they're releasing. So much happening in the AI space, still lots of gyrations. We'll talk about NVIDIA and the general stock markets in a minute. It's clear we're still so early on in this journey and it's just an exciting place to be. I want to stay with AI because obviously the Mag 7, I'm pretty sure that most people will have seen that because of tariffs and all of the Trumpian style economics, there's been some market corrections.

29:48So all of the newsletters I've received this week with market sections have been marked red. Most noticeably earlier this week, we saw effectively trillions wiped off the Mag 7 since Jan. So NVIDIA and Tesla have been taking the biggest losses. The only one stock that didn't take a fall was Meta. So the combined loss was around$2.7 trillion. Just to put into context, the UK's GDP last year was around$3 trillion, just over$3 trillion. So this is a massive market loss. What's coming to Europe? What does this mean more generally, this side of the pond? What's going on behind the scenes? Mads, can you take this one?