E440 | This Week in European Tech with Dan Bowyer, Mads Jensen, and Lomax Ward

0:00Welcome to Upside, where we dig into the real stories behind the headlines affecting European venture. Joining today is Lomax from Outsized, Andrew from 7 % and myself. I'm Dan from Superseed VC. Nomads today. He's on his way to see the cherry blossoms in Japan. Lucky chap.

0:27This show is not investment advice. and the hosts of this episode may be invested in the funds and companies featured. On today's show, well, I don't know how we're going to avoid Liberation Day in the US, pulling it apart, discussing how it'll affect Europe, and we'll pull that back to startups and investors. We're taking a look at European defence stocks and specifically Germany's spending plans and looking into the real right now opportunity with AI across Europe. Andrew, what's new with you? You've been a bit of a travelling dude. How's your week been? I have. I went to Japan and Singapore and South Korea for the first time, all in five days.

1:10Quite a long trip. What did I learn? Well, America is allegedly the home of great customer service, and I have to say they could learn something from Japan and Korea and Singapore. What a wonderful welcome. Everyone's so polite. Most of the places are so clean. And, you know, UK and Europe, well, we're so far behind on the customer service front. I don't know what to say. So I had a really great trip. Met some fantastic people. Lovely, lovely. Lomax, what's happening? When you run a venture fund as a GP, sometimes, you know, you're waiting around for a whale to enter the bay and you're all hands on deck, load the harpoon.

1:49And this happened this week. So Tuesday, met a founder who was closing around this week. So we've dropped everything to run hard at it, do as much DD as we can in a short time. I'm going to be giving him a call later tonight. So it's been one of those weeks where it's all hands to the pump. You weren't expecting it necessarily. And time is of the essence. So it's been one of those kind of weeks, you know. So let me get this straight. As far as you're concerned, the founder is aware of the whole harpooning. Is that? Well, the analogy, when you put it like that, is not the best one. Or you Jonah.

2:23You're Jonah, maybe. Maybe you're Jonah. I'm in my oil skins. It's a robo, I think is what it is. We're so going to get lost in an allergy. what um what what attracted you to the deal i don't know we can give away but what was what was the oh my goodness we've got to do this what was the yeah like um seed late seed bootstrapped by the founders taken in no money over a million of revenue seemingly 10x better technology than the state of the art and um a big us well i gotta be careful about what i can say but it's in uh ev in EVs. It's a supply chain for EVs. And it had been identified by a big US player as the leading tech in the world.

3:08So we've been doing a lot of kicking the tyres and various things, but exciting. And all positive. All positive. But, you know, TBC, whether we get in, whether we do the deal, etc. And what news has caught your eye this week? What are the quick newsy things that have... Quick fire, right? So look, big news i think was the isar um launch of its first rocket the first ever rocket on european soil from a launch company in norway it's a german company i mean it's nearly 20 years after the first you know spacex um trial and clearly this this is isar a munich based company and that has raised 400 million dollars had its first launch and the rocket did explode and blow up but this is all par for the course you know as you guys may recall it took spacex four attempts to get a rocket to get a rocket they expected it from what i can read they expected it and this was a measure of success it got off the launch pad it just didn't go that far and then flipped over and exploded but this yeah it was reported as the first rocket to be launched from uh european continent that could reach orbit but that's not actually true those who know their space history all know that in 1971, the United Kingdom launched the Black Arrow, launched the Prospero satellite into orbit in 1971.

4:28And then we promptly cancelled our entire space program and handed all the technology over to the European Space Agency, as we're so good at doing. Apart from that, it was great news. Launched from Norway, wasn't it? A spaceport in Norway, northern Norway. Launched from Norway, blew up, but as we say, shouldn't be a problem. Lots of valuable data. One interesting thing is it took these guys 400 million of funding to get to this point interesting spacex in 2006 when they had their first launch it took them 100 million of funding to get there so 20 years before um with a quarter less money spacex was already in in place so interesting kind of delta there normally europe is sort of renowned for being more capital efficient isn't it but not in this case we can't we can't put the down to inflation i guess 20 years no not quite no i think i know it's the opposite but i think you know that there's a number of things that feed into that including a ready kind of talent base that they had in the u.s probably to accelerate things with the funding that they raised.

5:20A couple of things caught my eye this week. So watching Amazon last minute try and buy TikTok, I thought was quite interesting. I know Trump's got his own plans for that. I've been wondering why Bezos has been cozying up to Trump. And maybe if I was a dirty cynic, I'd pose that. The other thing that caught my eye this week was AI models. Now, H &M have bought the rights to 30 models to not take photographs of the models for their campaigns, but to use their digital twins. And it kind of made me think, oh, we're going to bring back Elvis. Are we going to see Tom Cruise still acting in Mission Impossible 97?

5:55The Hollywood business model is so broken. Could this rejuvenate it or would this kill it completely like streaming broke music? But it was only a matter of time before we saw this kind of proposition where don't have the model, don't have the actor, don't have the singer, but I will sell you the rights to my image or my voice or my moving avatar. and i wonder where we're going to go with this and i wonder what it's going to mean for music industry movie industry and whatever it might be the other thing that i know i've met a number of people who are part of the mansion house compact crew um bbca in the uk which is the the kind of the uk venture capital association and trying we've talked about this a lot and looking at how pension capital can be unlocked into private markets into the UK mainly.

6:44But I know that there are many European countries that are looking at smarter ways of using pension capital cash. And I think my sense is there is going to be some actual movement. There is going to be some actual kind of pension cash. The model will be matched. There'll be some way of matching the VC2 and 20 into the fee structure that pension funds will find palatable. So my sense, I don't know if anyone's got any opinions or thoughts on this, but my sense is that there will be some breakthroughs this year. My bet is there will be some breakthroughs this year with pension cash. Andrew, what do you reckon?

7:20There's a lot of people been talking a lot. I want to see some action. So for sure, we'll see some movement. Will it be enough? Will the money find its way to the right type of investment firms? or will it all end up in mid-cap PE or growth sort of venture into de-risked enterprise software? That's the big problem. Will this money end up in the hands of VCs and general partners who are happy to really go the extra mile and follow the moonshot? Well, that's where the innovation lives, isn't it? The real innovation happens down in our world. And the big returns and the huge returns, right? the huge returns to own some of this technology over the next 10 years.

8:03It's interesting. I was staggered by the... There's a very low base, right? So 0.007 % of pensions money in the UK is invested in VC. 0.007, right? But look at how much goes into US venture. Look how much UK pension capital through various vehicles. It's like nearly in double digits. It's quite astounding. But sorry, I interrupted. Carry on. No, no, no, no, no. So I think under Mansion House Compact, they've targeted 5%, right? So they've got to get going by 2030. Until 2030, 5 % by 2030. My sense is that the vehicles they're going to be using, the funder funds vehicles or the vehicles that they're going to be using, are going to be more open and mindful of where the money goes rather than the pension funds themselves being able to be much more risk averse.

8:56So that's why I'm a bit more hopeful. That's got to be the way forward. Like, you know, in Europe now you have a cohort of emerging funder funds that actually understand the VC product much better than anyone in Europe has over the years. And I don't think pension funds going out, investing directly in GPs is the best idea. And what Andrew said, and we talked about it on this on this on this pod before, is it will end up going to the kind of boring people who aren't really investing in true innovation and are probably more like, you know, what we call AUM kind of fee generating machines. I think the funder funds do have the ability to distribute that capital down to the more proactive entrepreneurial kind of people at the more cutting edge of our industry.

9:37So that's got to be the way to go. I think my sense is that this year these things will unlock. Right. So let's talk a little bit about. By the way, hasn't TikTok become the ultimate party round? I mean, you've got A16Z there. You've got Silver Lake. You've got Blackstone. And you can see why this is the, as you alluded to with Bezos, all these guys hanging out with Trump. This is now getting probably access to an app that has 170 million users in the US at probably a very nice price, you know, with ByteDance having a gun put to their head by the Trump government. By the way, the deadline is on the 5th of April, so we should find this out tomorrow probably.

10:13It's the ultimate party round auction that's being run now, right? And then if you factor in the people who've already invested, like Code Chew, etc., you're going to have the kind of who's who all-star cap table of growth equity and VC firms, right? I totally get why Amazon's interested. TikTok shop, I didn't realize what a behemoth TikTok shop is. And it's only going to get better, worse, depending on your point of view. There's only going to be more throughput. put. This kind of shopping, it's absolute billions. So I can see why Amazon is keen. But I want to move to Germany. I want to talk a little bit about Germany's defense spend and defense stocks in Europe in general.

10:57Everything is on the up. Now, a number of defense stocks now trade on forward earnings multiples that exceed those of luxury and tech. So we've got Ryan Mattel, Saab, Kronzberg, Gruppen, SA trade at around 40 times forward earnings multiples. And now personally, I believe the defence will be the driving force that unlocks and saves Europe from irrelevance. I've also got this conspiracy theory. I was chatting with you guys earlier in the week. I've got a conspiracy theory that the lifting of the German debt ceiling was effectively an incredibly smart move from Schultz to rejuvenate and rescue the German automotive industry.

11:35So you have Volkswagen factories now producing tanks, subs and munitions. my take is that this will be a massive unlock what do you reckon will it be a wider gdp european breakthrough will it be this kind of growth hack that we need if so where how who when low max what do you reckon unleashing the fiscal shackles in the biggest economy in the eurozone is a great thing right so i think the cup this is an increase huge increase in defense spending coupled with i think 500 billion that they pledged to spend in infrastructure right so it's not just pure defense so i think this can only be a good thing i mean clearly it's pushed up their borrowing costs but ultimately everyone thinks that this will long term be an investment in growth and in gdp i think that's great i think it's a possibility like i don't i don't think it's a slam dunk you know we have to see what the asset allocation is like we have to see the execution of that i mean the last time i went to berlin i think then how long did it take them to refresh the berlin airport 10 15 years like these things you know in europe can get scuppered by you know the issues we talked about here before like excess regulations so you know like in what we can go we've been talking about in the context of the uk is you can pledge to make all this money and invest in infrastructure but you need at the same time to be cutting red tape and accelerating that and removing the roadblocks to that i think when i think about it on a startup level i see people running around like headers chickens um trying to invest in defense companies or dual use companies in venture when i see vcs flocking towards things i i worry um it feels like do you remember years ago everything was a blockchain app then everything was an ai app now it feels like everything is going to be a dual use or a defense tech yeah and per and per your point like you know you know pivoting volkswagen factories to defense it's the same we're seeing you know infrastructure ai companies pivoting to defense or dual use and you see that across a number of different companies I think I would worry here because you need to take a number of steps.

13:29It's like, yes, you increase by tens of billions, if not hundreds of billions, defense spending in Europe and the UK. That ipso facto, to me, does not lead to a big kicker in our world of startups, right? You know, you're selling into government. Not even in dual use. I think it will help, but I don't think it's slam dunk. Well, I'm just kind of thinking, what is defense? I mean, is energy defense? is cyber defense. It feels like you could, there is quite a broad bucket underneath defense that I think is going to be extremely meaningful. And if you think about all of the subsidiaries, all of these other businesses that are going to need to get involved in producing whatever it is that might be absolutely whatever they call it in that nowadays, peace technology.

14:18They don't call it war, they call it peace technology. So Andrew, what do you reckon? I think a better word is resilience. A part of that is kinetic you know so bullets bombs part of that is um equipment that's directly related to fighting a battle but you're absolutely right it's also cyber it's also energy i mean we still haven't got to the bottom of why heathrow went dark a couple of weeks back from one substation but i suspect if you did an analysis of critical infrastructure across europe and the uk we'd all be pretty horrified at how weak and fragile some of that stuff is so it's actually a super a very large area uh to look at and generally defense spendings i think probably a good thing if it's a rebalancing towards sort of defense and industrial autonomy within europe that's got to be a good thing we discussed before how 80 of the current spending goes outside of of europe that's kind of crazy when we're industrial nations that can build this stuff and we have the technology to build this stuff.

15:23It requires an alignment of ESG policies from investors. And it also, to your point, Max, it depends where the money goes. If all the money just goes into the pockets of the existing primes, BAE, Leonardo, TAL, Saab, Rolls-Royce. Which is what the market has bet, right? Those companies are up nearly 200%, this is 150, 200%. Yeah, they're up between 70 and 170%. I mean, all defense and all defense ETFs are up. You could argue that they were underpriced before, though. Yeah, of course. I would argue they are. That's not necessarily a bad thing, but I hope some of the money gets into the innovation because that's what's going to help us compete globally.

16:05That's what's going to be good for the ecosystem as a whole and obviously the stuff we care about, which is those startup companies. But you want to be not just investing in the capacity to create artillery shells just in case Putin loses his mind even further and invades Europe. But you want to be investing in the next-gen technology because that's what drives growth, drives innovation. That's what has a trickle-down effect into the rest of the economy and into other areas of dual use. So generally, I think it's good. Let's see where the money goes. It could be all the boring stuff, right? And actually, just like, for example, there's one big point.

16:43Germany, I think, the leading general in Germany came out, saying they need to recruit 100 ,000 more soldiers. He said for that to actually happen, they will need some form of conscription, right? And that will be the same across a number of economies. And I know we all think that warfare is going to be autonomous. I'm sorry, but like in the short term, you're going to need boots on the ground. So that, a lot of the money will be going in that. I actually think it's a very interesting article by a friend of ours, Nicola, who runs a dual use fund. And she was saying actually, because a lot of what I've seen in startups is what she calls battlefield tech, which is like these kind of very incremental point solutions, which actually, you know, you would kind of think if I'm investing in defense now, those are the things that you think might actually be the, you know, the kind of place to invest.

17:28But actually, I think it's a much more boring, the kind of infrastructure of defense that sits behind that, which might include hiring people and, you know, like trading people and all this kind of logistical stuff. This is my point, Lomax. This is my point. It's infrastructure. It's energy. It's recruit. It's operations. It's there. That's why my personal take is we are going to see an absolute reframe of lots of AI companies becoming peace tech or whatever the whatever we want to call it. And it's going to be quite a broad rejuvenating kind of set of scenarios where these startups have a lot more investor interest because our LPs and, you know, LPs, LPs, and there's going to this kind of broader push for a focus on defense, which I think is fairly, fairly broad and fairly deep.

18:16But I wanted to give one anecdote, right? I heard this this week that it kind of encapsulates what, when we think about it at the startup and investor level, what's happening at the moment, right? So I remember a climate software company four years ago during the climate boom raised a pre-seed at, let's say, 5 million. And about 10 minutes later, it raised the seed round at four to five times the price from a tier one VC fund with no change in substance, right? I've just seen the same thing happen with a defense company for no reason. Right. You know, went from kind of pre-seed at five to a 50 million seed.

18:48And when that stuff happens, I get worried, you know, because I feel people have like slightly lost their minds. So that kind of stuff is happening on the ground in defense at the moment. It is not straightforward. It is a largely B to G sale. I mean, unless you're selling into the prime. So it's a very, very complicated distribution route for a lot of these companies. So I think for tech and startups, there is opportunity. And don't forget that roughly a quarter of defense spending is on R &D. So there is definitely opportunities. But I think, per your point, a lot of it will probably be in the slightly more boring infrastructure side of things.

19:18One of the big advantages for the startup sector, though, is that a lot of this new technology is cheaper. So the primes are used to charging perhaps two, three, four million for a storm shadow missile that we've cruise missile that we've seen shipped off to Ukraine. the startups, even if they are working with some of the primes to ultimately deliver these, in conjunction with the primes to deliver these systems to the MOD or Ukraine, they're looking at price points of 15 % of that, substantially lower. And that's driven not just by the fact that some of these startups can create this technology more cheaply, but it's driven by demand from the primary customer today, which is the Ukraine, but also a recognition that actually in a world of drones and low-cost munitions, that your defense, your interceptors, the countermeasures have to be priced at a point which is you can produce enough of them and which is acceptable to the customer.

20:21So every single military, if it wants to be prepared, has to buy this stuff because they're all looking at Ukraine and going, wow, we don't have any of that stuff, we don't have any of that stuff, we don't have any of that stuff. So there's huge opportunity. And that's one of the really positive developments for the whole startup ecosystem in defence and dual use. I'm going to pledge to do a bit of a track on this and have a look at what's happening with dual use, what's happening with defence, as with how many startups are coming through, what the intrigue is, what the raised quantums are, what the value, and see if we can see any patterns of behaviour.

20:56And I'll report back. Final thing on the European side of things is that, per Andrew's point, the European governments have now pledged to basically buy European. Funnily enough, if you're an investor in Anduril, Anduril in theory has had half of its market just cut away from it because the Europeans won't be buying that technology anymore. So actually, I think it's very interesting for the Europeans because they actually in theory won't be competing against the US tech companies. And the Germans have just set themselves up as the central point of production, if I'm reading between the lines. So let's see how a lot of this unfolds.

21:32The Europeans won't even buy UK stuff at the moment, right? So it's very sovereign driven at the moment. The European's done a fantastic analysis of the resilient sector across Europe. And it's been up 30 % over the last two years, even though overall VC is 45 % down. So this is something that's already started and is only increasing. I want to talk about a report from Nicola Collin, who's the publisher at Drift Signal. He writes a blog there. He's also the co-founder of the family over in Paris somewhere. He's somewhere in France. Now, he's written an article called Who Will Be the Japan of the AI Era?

22:12And what he's trying to do, he's trying to draw a parallel between Japan's pioneering role in revolutionizing manufacturing with lean production in the 70s and the 80s and this current transformative potential of AI and reshaping production methods. And I think it's quite a worthy parallel and quite an interesting story. So as AI becomes deeply embedded in society, the key question is who will develop the most advanced AI technologies, but which countries or regions will possess the institutional frameworks and strategic vision to actually bring all this alive. Now he's suggesting that much like Japan's unexpected rise in manufacturing, the leaders in AI-enabled production will be at the application layer and the countries that will emerge as the winners, they will only win if they can embed these organizational structures and get it right down at the core level of how they apply AI and how they fully harness the capabilities.

23:07So to me, this feels like our opportunity. This feels like Europe's opportunity we're at this pivotal moment where China America are fighting on the foundational level we can harness this application layer and if if governments can get the act together and large orgs and institutions can get the act together we can really start applying and get Europe back on the map can they step up if so how can we step up where when so Lomax I know you are not in full agreement with with Nikola's thinking here where where would you take it no I think firstly I like I love his writings although they are um they are very bullish on Europe which is which is great but I I think sometimes erroneously so right so I just think when I I digested this article which is incredibly complex I think and it's a long one and one of the points he was making is that this lean production methodology that was pioneered by Toyota in Japan after the Second World War, in the context of manufacturing, was actually adopted much better in Europe than in the US, because that methodology was predicated on treating your employees well and fostering a culture of sharing.

24:20And what they'd said Toyota was, was you can have a job for life, right? So it encouraged people to kind of be in it together. Whereas in the US, it's a bit more transactional that people change jobs much more regularly, like they tend to be elbowing each other out of the way etc so in europe in a way was better at adopting that and the sort of symptom of that or the outcome of that was that our the european automotive industry actually did a much better job of adopting that and became a world leader in a way that the u.s actually ended up struggling right i don't know it's a long kind of complex point but i don't think that applies in ai basically so that's why i think i disagree with him and i was trying to think about why and i was like what i must have much as i love the idea of what you just said dan And it's like, if you just conclude that AI is better software, right?

25:05Let's just start with that as a premise. Even if it's a step change, it's still better software. Like, let's just take Europe versus the US. Like, the US, you know, the data is there. Like, the US enterprises have been a far quicker adopter of, for example, cloud technology than their counterparts in the EU, right? There is, okay, there is early, but I don't know. I can quote a report by McKinsey that shows that European companies are between 50 and 75 percent behind their U.S. counterparts in adopting AI. I mean, who knows that that's probably a bit vague, but I kind of can agree with that. So I'm sort of thinking, you know, if you take Nicola's like philosophy, you'd be like, oh, maybe the European kind of fabric and institutions are going to be better at adopting these tools and more collaborative.

25:51And, you know, everyone's kind of in it together. I just don't agree with that. If I look at, you know, you look at, just look at the numbers and the productivity. You look at productivity numbers. I think since 2019, the European, like US productivity per hour per employee has gone up 7%. In Europe, the equivalent figure is 1%. If you look at GDP growth since 2000, the Gulf is like staggering, especially if you, you know, you look at GDP growth, EU versus the US since the global financial crisis. And I think a lot of that is like adoption of the technology. Don't forget, guys, that the US spends three and a half percent of GDP on R &D.

26:25And Europe is maybe half that, just over half that. So adoption of technology, et cetera, et cetera, et cetera. We've already been doing it. We've had an experiment in the last 10, 15 years. So I don't see anything there that suggests that it's going to be any different this time with AI in Europe. Maybe I'm wrong. Call me the doom monger. but I don't see a kind of special reason for distinction the way that Nicola has drawn in his article. I hope and I know Andrew's going to agree with me. I'd love him to disagree with me. Andrew, where are you going to go? I'll agree with you that I thought it was an in-depth piece which was interesting as a thought experiment but there's a line in there that really caught my eye which was what chat GPT represented wasn't a technological breakthrough, but an interface revolution.

27:16And I just don't think that's true. There were LLMs before chat GPT, but nobody had scaled them to the feature set and competency that even chat GPT-3 and then GPT-4, with its reinforcement learning, sort of represented in that recycling of that network effect. So I think it's just fundamentally flawed to suggest somehow that this was just a great interface plugged into sort of run-of-the-mill technology. And I think it underestimates the value in owning the compute and the technologies and infrastructure around that compute. the processes which were applied from the you know it was the correct system or whatever um which you could sort of take and take off-the-shelf robotics and apply the same system of manufacturing ai is more than just a process of course if you've got access to the ai it's true that you need to embed this into every facet of society to to create the improvements that lomax has talked about but you've got to have access to the ai to do that and you've got to have access to those frontier models and whatever the next generation of technology coming down the coming down the tube is so i think it sort of simplifies a slightly more complex and fast moving ai landscape versus just a revolution and approach of how to build something which you know you can you can take as a as a process and a management technique and apply in your own factory you can't apply ai unless you've got access to the compute you can't build a model unless you've got access to the the processors um and unless you've got access to the software and the technology so it was a bit flawed from that point of view even thinking about the analogy with the pc you know this sort of the the pc world everything else um you know i think the ai stack is just evolving so much faster and you look at all the different people in different places trying to build these these frontier models you know in china you've got tencent alibaba and baidu all building frontier models you've got the us europe's hopefully finally waking up to the fact we've got to throw billions at it.

29:21So yeah, I thought it was pretty flawed, but an interesting article. I might regret this, but we debated this a lot and Mads loves talking about this and he's got some great insights in this, is that I think the AI, the sort of LLM level gets commoditized, right? So I think models get commoditized. I think there are though two things I didn't mention in my little diatribe was structural problems in Europe, which also cut across what Nicola was saying. Number one is the energy prices, which you talked about here. So even if, you know, computer models get commoditized, you still need to actually power this stuff.

29:58And Europe has a major, major, major barrier there. And the second one is this like regulation and red tape and everything. And I feel like that is a major structural barrier to the adoption and this sort of utopia, Dan, that Nicola is talking about, where the EU can suddenly become better at implementing this stuff and being more productive. Can I give maybe one glimmer of hope? we've been talking about software and ai there is the next level which is this physical ai which is the roboticization which is basically ai touching the physical world right and that is that is that is a whole kind of next level of um disruption and unlocking of like gdp growth potential etc europe might have if you if you take what nicola is saying might have an opportunity there because with this slightly more socialist kind of like supply kairetsu type integrated supply chains and long-term thinking and he's basically saying what a lot of the kind of italian german mittelstand has is very similar to japan that the u.s hasn't got you could maybe argue that with physical ai as something in a kind of more factory based um kind of uh industries maybe Europe does a better job integrating for those reasons than in the US.

31:13But that is very hypothetical. I don't know. But, you know, I'd posit that and I may regret that in five years time. The energy problem will go away. I think there will be new chips that will come along. There's going to be new power. I think it's a temporary, it's a short to medium term problem. The regulatory issue, I think you're bang on Lomax. Because if you constrain innovation early on with too much regulatory burden, You just can't move as fast as other people who don't have that regulatory constraint will. I think you're also bang on with the issue of, let's call it physical AI. That will become far more important, far more powerful than LLMs are today.

31:53I honestly think the current set of LLMs are a stepping stone. And there are a whole bunch of startups, both in the US and in the UK and in Europe, working on um from all different directions on solutions to this physical world ai problem and for people who haven't thought about it that much it's the difference between uh talking to chat gbt and understanding what you say through crunching of words and pattern recognition and chat gbt being stuck in a robot's brain walking into a room and saying go and pick up that mug and understanding what you mean and being able to recognize the mug and having all this, you know, the computer vision.