E374 | Dan Bowyer & Mads Jensen, SuperSeed: Perplexity’s $8B Target, Anthropic’s AI Control, NHS Data Privatization, and Tech Giants’ Media Acquisitions

0:00What is happening in the land of European VC this week? there's lots of stuff i have got lots of stuff we are going to start with mad's my good man we're going to start with well i want to start with one of mine which is this uh this ai perplexity eight billion valuation and if memory serves they're only doing about they've only got about 50 odd million users doing in the low millions of revenue so this is mega massive is this just AI slough, do you reckon? What's going on with perplexity? You know, controversial take, I don't really get it. Well, let me peel back. I get it and I don't because it's a beautiful product and it solves a real problem, but that's a problem that now has been solved by others as well.

0:45I just don't see how they beat Google, to be honest. I know there's not much they can do in perplexity that they can't do with Google today. But will they take a chunk out? No, no. Will this go into their DOJ breaking up? Could we layer this in, you know? Okay, then all bets are off. You're absolutely right. If Google is killed, then all bets are off. Hey, listen, I can't see a world where it's anyone's interest to kill off Google, but it might be a counter from Google to say, hey, listen, don't look at us. Look, there you know, this is what's going on in the world of search. So there might be something.

1:17I saw another report today that Perplexity are doing a deal with the Wall Street Journal to get around some of the copyright issues. So they're obviously working angles and maybe there's more juice to be had. I just feel quite small and quite lowbrow for that kind of valuation, that kind of race. A billion and eight, baby. I wouldn't mind being on that cap table. Right. Staying with AI and staying with the big foundational crew, so Anthropik, not particularly European VC focused, I grant you, but Anthropik's new AI model and control your PC. Should we be scared, Mads? Terrified. It doesn't feel like it's not something that we haven't seen before, but it feels like the power that's waited behind the control.

2:06It's next gen. It's a natural evolution. I think enterprises might think twice before letting that loose inside their firewalls. Security implications are obvious. Yeah. But of course, it's what's going to happen. Look, we've been on the robotic process automation path for many years. Companies like UiPath, Automation Anywhere, Blue Prism and others have done kind of very static automation. You run a script of playing back, clicking on certain buttons and filling out forms. And companies have paid a lot of money for that, right? and Accenture and the likes have earned a lot of money implementing these solutions.

2:41And this is just totally next level. It's going to combine RPA with an LLM and there you go. So it makes perfect sense. It's natural evolution. And this is going to be just one more step towards kind of this AI-powered future we see in enterprise. So Dave no longer needs to turn out. Forget hybrid working. Dave's got his AI agent sorting everything out, chatting to your AI agent. There you go. fixing the world reading the reports it doesn't look like it's all but supremely useful just yet and and maybe this is in reaction to sales force now shifting into agent force and maybe there's a little bit of this kind of push to say hey we're here too we're doing clever stuff on the desktop too so maybe that's something that's all these things will start as toys yeah i bet this this looks and feels like it's an interesting direction to have i mean i'm assuming this is mainly going to be a coding tool so if i can work out how to build how to build and code as a co-pilot but we'll see let's see what happens okay so i've always wondered why if i'm on the nhs why can i not have my medical records and there's been a bit of noise on the social web that the nhs will now allow you to host your own patient records.

4:02You can have your own patient records, which makes complete logical sense. Now, what I did realize was this is a Palantir product. And my question for us and the room was more around, is there any angle for the startup community to get involved in building the services that run the NHS? Because every time I've seen a top-down service like this tried to be implemented anywhere in the NHS. It's failed. And obviously Palantir running the game, which upsets me. Why could we not do this within our own four walls? But do you think there are any opportunities here for smaller firms to get involved? It's so interesting.

4:43You saw earlier that in many ways, one of the most frequently used communication tools inside the NHS today is, do you know one of those? Fax machine. Well, that, that number two on the list. isn't the nhs the only buyer of faxes in the uk no i don't i don't know what's number two it's whatsapp no yes oh great medical staff uses whatsapp whatsapp to communicate about patients wonderful who doesn't want zuckerberg having more patient data that they don't say they're touching but you know to be fair i think whatsapp is probably better encrypted than any other nhs or government platform for that matter, I can think of.

5:25But it just begs the question, why is that where we default? It just feels like a kind of lowest common denominator almost. Isn't there a better way? You think about how you can use, if you have all the data, if you use all the data in the right way, how much you can enhance the NHS efficiency. We've got this huge behemoth that we just keep pouring money into. And it's like this big black hole and things don't get better. and there's more money and more money and things just don't get better. And then what we have on the other hand, there's absolutely revolutionizing technology and we're still relying on fax machines and WhatsApp on a good day.

6:00And it's just like something doesn't quite make sense. Well, maybe we'll forget Palantir for getting involved. And I think even though they spent way too much money, they did a reasonable job on the COVID stuff and the roller. And I get that it was way too much money for what it was, but actually technically I was, even though the use case didn't really work but we've never had a pandemic like it so but i thought it was actually not not a disastrous rollout i was very surprised that they managed to achieve what they what they can achieve yeah the last the last time there was a huge program to try and revitalize the nhs using it was the np fit right the national program for it there that was rolled out over over over new labor in the in the noughties and it was just an abject disaster the way the government procurement process ran the way they split up the regions many of the big providers ended up no bidding they're like nobody's ever going to make money doing this and some of the firms that ended up winning then totally tanked the projects and it was just a disaster billions were wasted on absolutely no productivity improvement you have this huge as you see these huge problems with these top-down procurement processes and would it be better if we could adopt technologies bottom up, having a more agile approach, right?

7:13Instead of saying, let's invest, you know, 20 billion pounds and trying to transform everything in one go. Let's try and take snippets and pilot different approaches and then see what works and lean into that. I've never understood why they don't just create the framework to say, hey, market, here's the framework for what we're trying to achieve. You provide the tooling and then we'll open a marketplace for the best and the brightest it just feels like it's a framework play rather than a top down here we're going to impose x absolutely useless system but it does it means into the state of play now a lot of a lot of companies do bring your own kit so maybe the nhs will be whatever you have you know there'll be an infrastructure that can load it on let's see let's see where they end up what have we got up next so this i thought was really interesting so I'm reading on CB Insights.

8:01So Stripe, HubSpot and JP Morgan are buying audiences by acquiring media companies. Now, my immediate thought when I saw this article and this research paper was, is there an angle? Because we work with many founders who are doing offline events. Obviously, they're doing the online community building. They're building the offline communities. And I'm not saying that any of our founders can go out and buy a media company, but it just feels like this is a really smart thing to do, is to go and buy either the immediate or the adjacent audiences to the people that you want to serve or the people that you want to be aligned with.

8:35Have you got any thoughts on this? I think it leans into the whole notion of the creator economy and of media and audiences and content as the primary next generation marketing vehicle. We're invested in companies that work in this space. We know how much better it is to acquire new customers using content than just to buy ads. And so I think it leans into that. Absolutely. I think it's just a general trend. You know, next generation businesses are fundamentally content businesses because that's what we want. We want content. We don't just want a dumb ad, right? I was at an event yesterday for VCs and LPs and Stefan from Crandon was talking about venture capital.

9:16And someone in the audience was asking about, you know, what's important about building a VC firm. And he basically just said brand. And I think it leans into, and you can break that down in a multitude of ways. We won't go into that now, but I think it does come down into this whole content creator economy. But I'm always intrigued about how many more yards there are left on the field. We're already hyper-connected. Everybody is a creator. And I'm wondering within that noise, what the next evolution will be for people like us building a VC firm, how, what levers will pull. I'm sure AI will be a part of that.

9:56Right. So Noah Smith, how long can we sustain economic growth is next on the docket. And his byline was thinking about the very long term. So what was on your mind with this? I mean, it's just it leans into something we've talked about before in terms of growth and how do you create growth and how do you invest? And what is it that create the wealth we all eventually want to create so that we can build successful and flourishing societies? Something that's sometimes said by people that are very well-meaning is, listen, this whole notion of growth is stupid because perpetual growth in a world with finite resources doesn't make sense.

10:35And so we can't keep growing. So we have to find other ways to make our economies work. I think what Noah, he did very well in the article is he sort of takes that notion a little bit apart by saying, listen, the way we create growth today is not just by mindlessly consuming more resources. We've just talked about content. We've just talked about the creator economy. Now, if somebody writes a book or creates a painting or music or a podcast, you can create things that will add millions, hundreds of millions to our GDP and our wealth as a society, but without consuming very many resources at all.

11:11And we know this because what we invest in is business technology that enhances efficiency of companies. And so, so much of it is about enabling businesses to make more with less. And so in many ways, kind of the notion that there is a limit to growth, I think it's just, it's a fallacy. Growth is infinite. There's nothing we can't create with human ingenuity. And you can get into kind of resources and fossil fuels and all these things. Look, these are almost solved problems from a technical perspective. Yes, it's going to be another decade or two until we have fixed the energy problem on a global scale.

11:46But again, this is something we will fix. You know, I just thought he very eloquently outlined all the sort of the scientific and the rational arguments that are underpinning this argument of why growth is a great thing and what we'll continue to enjoy it for the foreseeable future. Well, listen, even if he's got it wrong, Elon Musk will save us. He's going to go and mine some asteroids, get us some more raw materials. We've got it all fixed so we don't have to worry about a thing, right? There's that too. I want to stay positive. So Dario Amadei from the CEO of Anthropic has written a really lovely paper called The Machines of Loving Grace.

12:26So it's very easy in today's news cycle to get caught in the doom and horror and horrificness of humanity and everything that we do to each other. And there is a lot of bad noise around AI and it's going to steal all the jobs and it's going to take over the world and it's going to destroy society as we know it. None of which I think are true. So Dario has written a paper. It's a really lovely read and I highly recommend it. Yeah, called The Machines of Loving Grace. I don't know if you've had a chance to read it, but he's talking about how do we avoid this perception of propaganda, but also avoid grandiosity and the sci-fi, this whole sci-fi baggage he goes on in to talk about.

13:06And then he goes into these five distinct areas where AI is going to make a massive leap forward. Biology, health, neuroscience, economic development, peace governance and work and meaning. And it's funny how he bundles those two together. And I thought it was really real and thoughtful and considered. Next up, we've got one of yours. So Europe should take a digital leap across the innovation gap. I don't think this goes into much more than we have said a million times. But what were the key aspects that really caught your eye on this one? No, I think exactly. I mean, it leads back into something we've talked about before and how important it is that we lean into technology and invest in technology as a way to create the next economy for us.

13:51We have been relying on kind of legacy industries. Automotive has been such an important part of our economy, but we can see all the challenges our automotive sector faces globally on the EVs and the manufacturing side, efficient manufacturing side from China, on the technology and self-driving side from autonomous vehicle side from California. And so we're going to get caught between a rock and a hot place if we just lean back on how can we produce slightly better gasoline-powered cars. And the only way really to get back on the front foot is to invest more in technology. It leans into something else we've talked about a lot, which is how can we get more money into venture capital in Europe?

14:35How can we back more founding teams to go and create the businesses of tomorrow? And we need more capital because we need to make bigger swings. We can't just rely on us making bets at the early stage and then those companies growing up and migrating off to the U.S. and effectively ultimately benefiting the American economy. We have to be able to go the whole way and enable those businesses to stay here. And it's so interesting when you look at the difference between the rules for pension funds, pension trustees in the U.S. and Canada versus in Europe. Whereas in Europe, you have all the areas around fee caps and prudent man rules and all these things that constrain and hem in the ability of pension funds to invest.

15:21And so what we seem to keep doing is we try to address the symptoms of that. We've been trying to address the symptoms of that in the UK with this new long-term asset fund approach that's been implemented. But it's addressing symptoms rather than the root cause. And so I think we have to try and unlock some of that capital if we want to make a step change here. So I'm hoping that that rises back up to the top of the agenda. I think it's so important. Yeah, noises have been made at government and regulatory level and EU level. So at least that conversation is happening. But I think it's also worth pointing out that it looks like, and we spoke about this last week, that the UK is looking to raise 12 billion into venture capital this year, which is more than the heady peak of 21.

16:06So hopefully that's a good start for the UK. There'll be more coming from Europe and we can hopefully start to open up more funding channels and get some real stuff coming in. Next up, we've got Europe's lagging tech sectors. So this is a Chicago Booth report. This is one of yours, Mads. It's got all kinds of facts and figures here. Productivity, workload per hour, energy prices, EU compared to the States. It's fairly comprehensive in thinking, although it's quite a short report. What was on your mind with this one? I bet it's exactly a continuation of the same topic. We need to fix energy. We need to have cheap and stable energy because that's essential in the digital industries we want to compete in.

16:48Do you think that's number one? Do you think that's the first thing that we should solve? energy? I think it's always difficult when you have a system that needs to be addressed to just try and optimize one thing. We have to fix energy and we have to fix capital. Regulation. But I think the regulation is how you fix these other things, right? It's the regulation that needs to be fixed so we can actually build energy generation and distribution and so that we can raise more capital. That's the key for us, I think, to make a step change. Do you think we have the political will? Do you think Starman Rees will do it from our side?

17:20Do you think the EU, do you Do you think Draghi has had any impact on the EU? He's had an impact on what's being talked about. The challenge is you still have a fairly torturous process to pass new acts, to pass new directives, and to change European legislation and regulation. It's a long, long road. So we'll know in two years maybe whether there is a real impact, and then it might take even longer for things in the fullness of time to be rolled out. It's clear that we've started in the wrong place. Kind of the regulation around the artificial intelligence regulation, some of the ways GDPR was implemented.

17:54It's just some of that stuff is not helping us grow. And we desperately need to grow now. We don't have the luxury of leaning back and saying, well, let's shuffle the deck chairs over here and rearrange a bit and making things a bit prettier because wouldn't it be nice? We desperately need growth. i'm hoping that the thoughtlessness and stupidity on cdpr cookie management other other regulatory nonsense has been a bit of a learning for the eu i hope so and if not i hope the uk now not in the eu can make independent decisions but we'll we'll see because obviously we need to be aligned because there's 550 million of them right there heading back over the pond so this is a bloomberg article looking at how basically it goes into detail of how the the srp 500 gains the era of those gains are over a goldman strategist says so what's going on here is this is it the magnificent seven dropping up now no longer what's the deal it's on the back of a very well written well analyzed goldman's research paper effectively looking at a lot of different factors and saying net net on the back of this kind of this incredible decade we've had of s &p 500 gains of stock market gains are over and we're going to see much much lower returns over the coming decade and they go through a vector after vector of why that is and you point some of them out it's the concentration in the s &p 500 where you've got kind of the magnificent seven you've got these massive these mech companies having such a concentration of market cap around them it's to some extent relatively high valuations compared to historical trends.

19:34It's a level of high profitability of companies already where there's expectation of some mean reversion. So it's adding all of those things up. And they sort of say, we plug all of that into our big model and it suggests to us that returns will go down. That means that investors should be looking elsewhere for returns. So that's sort of one angle, but they have a very, very interesting paragraph in the report saying, if AI comes good and enhances productivity in the way it could, that could dramatically improve the profitability of companies, which of course will flow through to earnings per share, which means that actually there could be bigger gains ahead as well.

20:14And one thing's for sure, for my money, I would not bet against the S &P 500 right now. I think a slice of America when you have AI being rolled out to enterprises everywhere is probably a sensible part of any portfolio. So yeah, I just thought it was an interesting, interesting perspective. It is an interesting perspective. I'm always keen to explore how this will affect Europe and what the knock-on effect of any kind of reduction in growth or any potential stagnation, dare I say, recession, what then happens? And we talk about sneezing, catching a cold, but where are we at in the cycle where assuming Kamala wins, which you're going to come on to next and Trump doesn't, I imagine there will be not as much excitement for commercial activity as there would be under Trump.

21:06So I'm just intrigued as to how this slowing potential reversing would actually affect Europe. Any thoughts on that? Yeah, I think a lot is down to how well we are setting ourselves up to compete. You know, we talk about relatively low European growth rates, but it was not always like that. Europe was absolutely bombed to pieces after the Second World War. And then there was a huge element of catching up. And European economies had periods over the last 60, 70 years when they grew very rapidly. But that is now, you know, we are now seeing all these systemic issues I spoke about before. In many ways, we've been powered by this kind of this German export engine, this incredible manufacturing and where automotive sort of is the centerpiece.

21:51And if that automotive sector comes under threat, it can threaten the full supply chain and machinery and tooling and kind of all the other things we're doing. You know, for us to benefit from AI, we have to actually be competitive in the industries we are in. That's not going to help. If nobody wants to buy our cars, AI is not going to help us. So we have to have industries and products where we are world leading and the AI will then enable us to be more efficient at making those products. And so if we can retain our leadership in some spaces, you know, pharma medicine is maybe one of the few areas where we are truly, truly globally leading, then it could be a good thing.

22:29But kind of other sectors where you sort of traditionally see European strength, fashion, is that really a place where AI is going to have such a big impact? Maybe not. You know, I think it's back to the article we spoke about before, kind of digital first, digital forward, we need to retool our economy to be more digitally native. We need to develop industries where we have leaderships in digital products. And then, you know, from that, you could see those kind of business efficiency gains and therefore earnings per share gains and therefore stock market gains that we would love to see for European companies.

23:00Maybe this is an opportunity for Europe. Let's, I mean, let's move on to the next piece, which is another Noah Smith article. He was basically asking the question, how much damage can Trump do to the economy is in the broader sense. He talks about the three main economic policies that he has, mass deportation, high tariffs on China, and very large tax cuts for the rich. So they are going to be debt and inflationary triggering. But what else is coming out of here for you? I think that's your spot on and pulling those out. Actually, I was putting that up and maybe to save it for us to discuss on the fifth, the day of the election.

23:40Now, my expectations are on current polling that Trump will win. And so I think the question is, what will happen? And the challenge is, there's been a strength we've seen in the business community. I think many people have been a little bit worried about some of the things Kamala has said and done. And there are concerns around kind of the immigration policy that's been implemented. But what we know is that kind of all the immigrants that have come to the U.S., they've helped keep labor costs down. And so what Trump is proposing is, as you say, it's a reduction in immigration, which is going to put pressure on wages, which is going to fuel inflation.

24:17It's a massive expansion of the fiscal deficit in the form of tax cuts, which is going to put pressure on interest rates. And then inflation itself will put pressure on interest rates. And so could this lead to a reversal of the downward trajectory we are seeing for interest rates right now? And it's clear that if it does, that will be extremely detrimental, not just to the U.S. economy, but to the global economy. So I think this is, you know, is Trump going to do these things if he gets back into office? He's not going to win. I'm not having that. He's not going to win. okay let's let's play let's play with the idea that he's like is he going to do it i'm i'm not sure he will because i think that the policies he's proposing is so damaging uh and and i and i think people will um will work it out and i think he eventually will work that out i think to some of some extent he is saying right now what people want to hear as opposed to what he will do but uh but if he ends up doing it it could be a real issue yeah yeah well i mean we're vcs right so We're always going to look at the opportunities.

25:24Whichever way, up, down, left, right, there will be a startup to solve that problem. Lovely, my man. What a gift. I will catch you next week. All right. See you next time.

25:39Tear down this wall. It's more than just an alliance. This is a union of values. Let's stop acting. Thank you.