Tech analyst Philipp Klöckner in conversation with Conor McNamara

Transcript

Philip, welcome. My pleasure, thank you. Awesome to have you here. What do you think of that video? I would agree that it depends on how you define singularity. So is AI taking off, becoming totally independent? I don't think we're there yet, but what I do see are the hints or the bellwethers of a huge paradigm shift. If you look at new business formation, if you look especially at how AI accelerates itself, how software builds the next version of the AI and that version builds the next version and the production velocity that companies like OpenAI or Anthropic have. If you look at the revenue expansion that is unheard of of Anthropic. In our podcast I said Anthropic will be the first company who will have a revenue expansion rate from north of 200% and it turned out it's north of 500%. That means the average Anthropic customer spends five times more compared to last year with them, which is completely unheard of. Every software vendor would kill for such a revenue expansion and that's what I mean with paradigm shifts. Stuff that was true before needs completely new benchmarks and metrics. So I do see this moment in time that 2026 is a year that changes a lot of things for sure. For sure. 500% well. So on, maybe just, we could go many ways with this conversation so we'll stick to maybe software development to start with. We see it at Stripe a lot. What's your perspective on software development and the changes that have happened as a result of AI? And do you think that the world as a whole has digested the implications of the pace at which software is getting built today versus 24 months ago? [SPEAKER_00] I think most people don't. I think there are a few people, mostly working in startups or founding startups building new businesses, who are well aware of the possibilities and make maximum use of it. And then there are also people who maybe work in incumbent companies in the German Mittelstand who still try to find excuses why not to, why their piece of software couldn't be built by AI or their job couldn't be done partly by AI. And I think that will continue to exist, this dichotomy, or this you could say a world of AI have and have-nots. We're gonna stick with that for at least a few more years. So we saw the same with the digitalization and the internet. For some people, usually the younger generation, it became completely normal. They were digital natives, they were internet natives. You will have AI natives. People leaving university right now are AI native. They have done their exams with AI and then there are people who still try to find excuses not to use AI, which is stupid. [SPEAKER_01] Yeah. I think, especially with software development, probably, and that makes Anthropic such a good company, that they focused on the kind of perfect entry market from a go-to-market perspective, and software development is for sure the stuff where you will have autonomous agents building pieces of fabulous stuff. Doesn't mean it's perfect yet of course. You still need a human in the loop. You still need to check stuff, redo it, etc. But you can't deny you are three or four times more effective or efficient building software. If so, you're just doing it wrong. And I mean, just if you look at how people, you personally use AI, all the people, there are a few skeptics now that say that AI is a bubble that will burst and almost act like it's going to go away again. We have the same with the internet and digitalization of course. Don't fight gravity. Exactly. It's not going to go away. People have started using AI. I mean, a good test for what a good product is, is would you take it away from me? Would I regret it? And if you would take AI away from me again, I would be mad. I would be crazily mad. And I would be begging to pay you 2,000 euros per month to let me continue to use it. And we see that actually in our data at Stripe. So we see quite a lot of users moving from one AI tool to another. But we don't see people who start on one AI tool never go back to non-AI tools. So there's a lot of tool hopping happening but not necessarily like I'm going back to the way I used to do it two years ago. That's just not happening at all. [SPEAKER_01] Which I think reinforces your point. That's what I envy you for. That data I would obviously give my firstborn to get that kind of data. I'm very happy for the ramp data which gives us some indication about B2B usage. But obviously I would love to see who's cancelling the contract. Maybe go for a point later on and I'll tell you some more. Awesome. Just to go back to your point around, you know, this. We often talk about technology being a democratizing force in the world. I mean, I worked in cloud for ten years before joining Stripe and that was a big story. But actually, when you look at a lot of the data, in some cases, technology cannot necessarily be an economic equalizer in some ways. At least in the short term, we can create a divergence in the economy between the haves and have-nots. What's your view? Where are you on the whole stance of AI? Is it going to create further divergence in the world in terms of economic prosperity? Or do you see the future is where it's going to lift all boats? [SPEAKER_01] Excellent questions. I would say both. So I would always argue it's probably a very low barrier technology. So almost everyone potentially has access to free AI if you use Google Gemini or something or Meta, you can basically access it for free. And access to information and to knowledge was never so readily available for people at almost zero cost. So that should be a huge equalizer and at least offer chance for everyone. And then still, I think that there might be monopolies in some of the AI markets again and lots of the value will accrue at very few people potentially, unless we find ways to distribute the value more evenly. But what you can say for sure is that even if the economic value might accrue, basically the money that is earned, accrues at very few people, the kind of consumer benefit that you have is very evenly distributed. Because everyone can use it, most people will be able to use it for free, and everyone who wants to can reap enormous benefits personally from it. You basically have everyone has access to an executive coach now, to a therapist, stuff that's really hard to get, at least here in Germany, is in your pocket. Now at a decent quality by now and it's getting only better over time. And even if a few more trillion dollar companies will be minted by this, everyone of us is able to benefit from it a lot. So these things can be true at the same time, I think. Agree, agree. No, it doesn't make sense. Yes. Therapists, fitness coaches, financial coaches, who knows. [SPEAKER_01] Awesome. I want to talk to you a bit about, like, again, I think Patrick has said this before. Everyone who wants to can reap enormous benefits personally from it. You have everyone has access to executive coach now to a therapist stuff that's really hard to get at least here in Germany is in your pocket. Now at a decent quality by now and it's getting only better over time and even if a few another trillion dollar company will be minted by this everyone of us is able to [SPEAKER_01] benefit from it a lot. So these things can be true at the same time I think. Agree, agree, no it doesn't make sense. Yes, therapists, fitness coaches, financial coaches, who knows. Awesome I want to talk to you a bit about again I think Patrick has said this before he's said we the SaaSpocalypse it's a discussion where AI is software at the world now almost AI is eating software in some ways and so you see this and then Patrick has used this analogy of a pizza recently so he's saying software should work almost like a pizza you know make it once use it once and move on to the next thing. Where are you in this whole SaaSpocalypse thing? Is it overdone? Is it reality? Where are you on that? So for some pieces of software it will be reality but on average generally I think it's an exaggeration. The sell-off of software companies. I don't think certain companies are worth half or only 25% of what they have been worth before. Stuff like Monday.com or ServiceNow. I think it's always overdoing some emotional reaction if you sell off a company by 75% in a few weeks or a few months. And I think people are not aware how software purchasing in corporates actually works. Because of course I could write code my own ServiceNow or ServiceNow is a bit harder. Let's take Monday.com or Asana. You could write code that even a bit tailored to what your own needs which would be good. [SPEAKER_01] However you still need someone maintaining it. You need someone to care for the security and safety of the product. [SPEAKER_01] You need to continue to develop it. So it will still take some cost. And of course that will be with AI and write coding it will be much cheaper than it has been to build the first Monday.com. [SPEAKER_01] But you don't need to be cheaper than Monday.com. You need to be cheaper than one license of Monday.com to make it economically feasible. [SPEAKER_01] If not it doesn't make sense to build your own piece of software. So as long as one license of ServiceNow or Monday.com is cheaper than write coding it. Which it will be for a long time. Because again you have the maintenance that you need to maintain the software and continue to improve it. [SPEAKER_01] You have to adopt to tax regulation, to compliance, etc. So I don't think most companies will stick with their existing software products. [SPEAKER_01] Then there's the thing that agents, why wouldn't agents use software as well, right? [SPEAKER_01] An agent that tries to shop something needs to connect to APIs, use other services, etc. [SPEAKER_01] So Jensen Wang, that's Jensen Wang theory that as we have billions of agents they will actually increase the use of software. And that brings me to the last point which is ultimately this separation between SaaS software and AI that doesn't make sense, might make sense in the short run. But ultimately all AI is software obviously. And all software will include some kind of AI. [SPEAKER_00] So I think the so called SaaSpocalypse, I find the word in here. [SPEAKER_00] I just kind of moment through it, hopefully no more. [SPEAKER_00] So I think that's a rather buying opportunity for the good pieces of SaaS. [SPEAKER_00] For example those that basically store the data for the clients. [SPEAKER_00] The SAPs, the oracles, the sales forces, the systems of records. [SPEAKER_00] It will be really hard to get those other companies. [SPEAKER_00] If in Germany if you want to change your SAP system, you know that better than me probably. That's a two to four year project and usually one C-level person loses their job doing that. Only one? [SPEAKER_01] I want to see the companies switching from SAP to a write-coded ERP now. [SPEAKER_01] And obviously the value of SAP isn't the ugly front end but it's knowing the tax regimes of 210 countries in the world. [SPEAKER_01] You might argue that as well has become a bit easier with AI to maintain that. [SPEAKER_01] But I think especially the more complex parts of software have a good market. [SPEAKER_01] However, and that's a big difference and there we are with the half and half-nots in the K-shaped economy. [SPEAKER_01] I think there will be less startups ever purchasing in SAP or growing into that stage where they would typically use an SAP. [SPEAKER_01] And they might never need one because they will never have such a large HR. [SPEAKER_01] So SAP success factors they probably not use because their organization will be 12 people and not 1200 people. [SPEAKER_01] So these companies that are built AI first from scratch, they will need all the processes that are only built for humans to exist in these processes. [SPEAKER_01] And then you might never need an ERP. [SPEAKER_01] A friend of mine reported to me about her first AI use case and I said what did you build? [SPEAKER_01] And she automated a report that she would get every Monday to check on ad bidding or something. [SPEAKER_01] But the thing is that report only exists for her. [SPEAKER_01] So in an AI native company, no one would ever build that report or process. [SPEAKER_01] And that's what's going to be the huge advantage of companies that are being built right now. [SPEAKER_01] That they will use the technology to build companies in an entirely different way with less overhead, [SPEAKER_01] less people just doing measuring, counting, more automatization from the very beginning. [SPEAKER_01] You don't need to constantly think of the human in the loop. [SPEAKER_01] You can just build an ecosystem that works in itself. And for those companies, they might never buy certain software products. But they still need a Twilio or a Cloudflare to host the stuff. So even those will continue. Maybe just, I mean, completely gone off topic here, but that's always a sign of a good fireside, right? We go all over the place. But maybe just workforce displacement in general, right? So just on the topic of that, it sounds like you see the average size of a business, headcount wise, getting smaller over time as a result of AI. Again, we certainly see that in our data. What do you think when you look at the future of the workforce and just again, bear cases and bold cases around what AI will do for the future of work. What's your view on it? And what are going to be the emerging areas of growth for people that will require more human workforce, etc? Yeah, a few things. So first, it will workforce displacement will be much slower than expected. Because there are these young companies that would never hire certain people. However, we always underestimate the inertia, the beharrungskräfte, German word, of the large enterprises, basically, and the public enterprises. So, and again, the analog is again digitalization where you had 20 years and it's still ongoing until people were fully transformed or something. And the same will be true for AI. So, the big companies will be relatively slow in adopting AI. So the job displacement can't happen that fast unless the startups would actually take away huge parts of the markets from them. Yeah, a few things. [SPEAKER_00] So first, workforce displacement will be much slower than expected because there are these young companies that would never hire certain people. However, we always underestimate the inertia, the beharrungskräfte, German word, of the large enterprises and the public enterprises. So again, the analog is digitalization where you had 20 years and it's still ongoing until people were fully transformed. And the same will be true for AI. So the big companies will be relatively slow in adopting AI. So job displacement can't happen that fast unless the startups would actually take away huge parts of the markets from them, which then won't happen as well. The other thing is this AI and automation robotics era couldn't hit us at a nicer time because we have this strategic gap in the German job market of 500,000 people, which means we're lacking 500,000 workers more every year because of the demography of Germany. So we need 500,000 new either immigrants. If you try reproduction, that takes a few years to build workforce. So either we need to automate stuff or we need to let in a lot more immigrants. So in that regard, if AI automates 1% of jobs per year, that's perfectly fine because that's exactly the people we need to replace anyway. So I'm quite optimistic that we, at least in the beginning, won't have riots or huge displacements. And apart from a very few key industries, like I wouldn't tell my daughter to become a fashion model anymore. Probably she doesn't have the genes for it. And on the other hand, if you look at jobs for catalog models at Otto or whatever, that is down 63% already in a few years. Customer service may be the same, photographs may be the same. But especially stuff like developers, lawyers. Lawyers, Goldman Sachs said 44% of work could be automated in the short term, which makes them the second most exposed branch for AI. However, lawyer hirings are at an absolute high. And you won't find a lot of scientific data that supports job displacement right now. And the studies that you can see from Brynjolfsson from Stanford only show that this is early. You can listen to it in the podcast tomorrow, they will explain in more detail. But that was a plug for the podcast. Very subtle, very well done. No, but in short, there is this narrative that early, the job entry jobs, the junior jobs, the early career steps, those are falling away faster or they are falling away when the others, the seniors are still hired. And I think this misses several points. There's a data selection bias because it's only small and medium sized companies where the data is. And I think this is the first pull from mostly companies that embrace AI very early. And I think what actually happens is not that companies don't choose to hire young people anymore, but that young people are not choosing these jobs anymore. You can see the computer science enrollments in the US is already going down from 660,000 to 600,000 in the last two years. So I think the actual problem is not companies not wanting to hire junior people, but people coming from university using a lot of AI and deciding I will never become a software engineer or a customer service agent because that doesn't have a future. So that's what's actually happening. Interesting. Yeah. Okay. So it's on the supply side rather than the demand side. Okay. Switching gears, agentic commerce. What's your view? Agents becoming actors, trusted ecosystem on the web where people, nonhumans are purchasing on our behalf. Where are we on that curve? How bullish or bearish are you in terms of the future outlook for us? And what use cases, if any, will be the ones that will emerge first or earlier? Yeah. I think we're in the very early days and this will also move much slower than you would think. I'm the worst keynote speaker of the world because usually people try to make these paradigm shifts and tectonic shifts look like they would happen within two or three years. But if you look at data, the truth is every big shift, the mobile shift is still going on. One percent of people shift from desktop to mobile every year. And that's not because they shift their behavior, but because old people who would never use a smartphone for certain things die. And young people who use a smartphone for everything get into the market. And that's why it shifts by one percent or one and a half percent every year. Right. And this is how usage shifts happen. It's usually. And that's why I think it all will be much slower. So I'm. Change does happen, but you will find very few technologies that penetrate a whole market in a fast manner. Like WhatsApp is a good example. Our grandparents are using it. Our kids are using it. So that has really quickly penetrated lots of people. But people are not good at adoption. Distribution is much more important. Coming back to agentic commerce. I think if that would address a natural need of humans, then first Alexa would have been much more successful. Like the use cases you hear now sometimes like ordering your consumer staples, toilet paper, whatever. That was well possible with Alexa already. The new LLMs are not a game changer in terms of talking to Alexa. The payment infrastructure has become better. I admit that. But I don't. I think people just don't want to shop that way. If they want to shop that way, then our high street areas in our cities would have belts and not windows. [SPEAKER_01] People want window shopping. They don't want to talk to someone and say I want to order something. They want to window shop. They want to compare. To meet a decision what you want to buy, you need to see the options that you don't want. You want the luxury stuff that you can't afford. You want the cheap stuff that you say I'm better than that. And then you decide for the option that you love. That's why these chat interfaces right now rebuild exactly that. They basically live craft you a comparison website in ChatGPT to rebuild a price comparison, a comparison shopping website, because that's how people want to shop. [SPEAKER_01] People want window shopping. [SPEAKER_01] They don't want to talk to someone and say, I want to order something. [SPEAKER_01] They want to window shop. [SPEAKER_01] They want to compare. [SPEAKER_01] To make a decision what you want to buy, you need to see the options that you don't want. [SPEAKER_01] You want the luxury stuff that you can't afford. [SPEAKER_01] You want the cheap stuff that you say, I'm better than that. And then you decide for the option that you love. That's why these chat interfaces right now rebuild exactly that. They live craft you a comparison website in ChatGPT to rebuild a price comparison, a comparison shopping website, because that's how people want to shop. And if that has been done, people at some point might continue to then use an automatic way to not go through the shopping basket anymore, but then have an agent do the rest. [SPEAKER_00] That I believe in. [SPEAKER_00] I think the really pesky stuff like consumer claims. [SPEAKER_00] Imagine your train is delayed. [SPEAKER_00] You have a claim against the Deutsche Bahn. [SPEAKER_00] I don't want to do that myself. [SPEAKER_00] I want an agent who just automatically checks every plane and train travel that I have and gets my claims automatically. [SPEAKER_00] I'm happy to share 10% of the results just to get rid of it. But today I can handle it myself. I don't even need to share with flight ride or something. So I think consumer rights might be something where it will enter rather early. And the whole services and appointment stuff like having the table on the window with your favorite Italian restaurant, booking a treatment with a massage parlor, hairdresser, stuff like that. A dentist, doctors. That makes a lot of sense to do that in a generic way. And sometimes it involves commerce. [SPEAKER_00] So if it's treatments, there is a payment included. [SPEAKER_00] If it's doctors, maybe not. [SPEAKER_00] But we shouldn't have people sitting in the doctor's offices that are just there to answer the phone. [SPEAKER_00] So that is the stuff that I see first. [SPEAKER_00] Booking travel or doing larger purchases, I don't see that yet. [SPEAKER_00] And if so, even for a small fraction of the consumers. [SPEAKER_00] How do you think about B2B? [SPEAKER_00] So, obviously, we're very stablecoin-pilled at Stripe. [SPEAKER_00] We see stablecoins as this complementary rail to existing rails in terms of promoting economies that just weren't financially viable in traditional rails. [SPEAKER_00] And we believe that there's whole sorts of agent-to-agent transactions that will happen online in the future. [SPEAKER_00] Again, small percentage today. [SPEAKER_00] What's your view on that and the degree to which that will happen in the future? I see that as much more likely. Generally, businesses are faster in adopting new technologies. You can see that with AI. Consumer adoption is much lower compared to business adoption right now. So, technical spend, procurement, I think that makes a lot of sense. That's a cohort of processes that are made to deal with certain human limitations and compliance requirements. And that is much better done by agents. If I want to, I'm a keynote speaker. If I speak at a conference of Deutsche Bank or SAP or Robert Bosch, to be 30 minutes on stage, I have to go through a four-hour procurement process. [SPEAKER_01] That obviously doesn't make sense. Just so they can procure my solution. Yeah, but all the small stuff that a company buys, again, technical spend, but even larger stuff, I think that might well be agentic in the future. Maybe the margins will be a bit smaller there given high repeat rate and large volumes. But the total gross margin is much bigger. For sure. For sure. Okay, let's keep going. Let's talk about Germany, the German economy. I was debating, I know I'm on dangerous ground here, asking you if there's any similarities between the state of the German football team and the state of the German economy. But I wasn't sure if that's the right way to go or not. But let's go there and see. [SPEAKER_01] The only consolation I was told on mentioning the German football team was that at least the Dutch also lost. [SPEAKER_01] So that's at least something, and I speak, I'm an Irishman, so I know I'm not on higher ground here. [SPEAKER_01] We haven't been in the World Cup since 2002, I think. [SPEAKER_01] Anyway, what's your perspective on the German economy, where we are right now, the role AI can play? [SPEAKER_01] You already alluded to it a little bit when you talked about in terms of supporting what is a gap in terms of the workforce availability to support economic growth. [SPEAKER_01] But broad perspectives on where Germany is and the upside and downside case for Germany. Where to start? So there is, the example with the national team might hold up quite well. This morning my girlfriend asked me, haven't we been one of the best teams? When I was a child, we were someone. We would be playing the best teams and sometimes we would win and now we are not participating in the final rounds anymore. And the same is true for Germany. We've been leading in a few industries very early on, life science, engineering stuff, then car manufacturing and machines in general. But we haven't made the transition to the next technologies in certain industries or to new industries at all. And I think maybe it's because we are too well off. Most of us are still in a very comfortable manner. I think the incentive to build something new is really small in Germany because if you study at a good university, you did a good exam, you can get 80, 90,000 euros working for Porsche, for Mercedes-Benz, for Google. So why would you build something on your own? You can plan already when you want to build your own home after five years of working. So why take the risk? And we don't take enough risk, I think. The few people who really take enormous risks are oftentimes immigrants. You can see that much more in the US, where almost all of the pivotal companies were founded by immigrants. [SPEAKER_01] I mean, ultimately, everyone in the US is immigrant, or child of immigrant generations. [SPEAKER_01] But these are the people who changed the world. [SPEAKER_01] And we have another mindset problem in Germany, I think. [SPEAKER_01] We are trained, I studied business administration as well. [SPEAKER_01] So I'm one of these guys and I may say that. [SPEAKER_01] But if you study business administration, you are basically trained to improve some process by five to ten percent. [SPEAKER_01] And extract five to ten percent more productivity or efficiency from some process. [SPEAKER_01] But engineers are trained to build something entirely new. [SPEAKER_01] And we need more of these engineers who don't think, how can I extract ten percent more efficiency from an existing process? But yeah, these are the people who changed the world. And we have another mindset problem in Germany, I think. We are trained. I studied business administration as well. So I'm one of these guys and I may say that. But if you study BVL, business administration, you are trained to improve some process by five to ten percent. And extract five to ten percent more productivity or efficiency on some process. But engineers are trained to build something entirely new. And we need more of these engineers who don't think, how can I extract ten percent more efficiency from an existing process? But how can I rebuild this process entirely? Or how can I build a completely new business that makes this process redundant, after all? And if you look at the past in Germany, it has always been like that. All the big German companies, the car manufacturers, the Siemens, they have all been built by engineers. All the US, the ten richest people in the US are all engineers, by the way. They are not business people, but engineers by training. And the same was true for Europe. We had that before, the engineering DNA. Even Bernard Noh, the owner of LVMH, which is the richest person of Europe, is an engineer by training, not a business person. He's a very good business person, but he's an engineer by training. And I think we need to stop having business guys like me building and consulting and advising and optimizing stuff. We need people who build stuff from the ground up, who build it smarter using the newest technology. [SPEAKER_00] If we now use AI to transform companies bit by bit every year, that's not going to change the fate of Germany. I think we need to spend as much effort as we spend in preserving our incumbent car industry, etc., on new business formation, on building the next chapter or the next wave of companies that will use entirely different tech stacks, will use a different breed of people. I think that's the only way how you can reignite growth. But if you look at, and we talked about this backstage, EU Inc or the 28 regime, Draghi's report, which has been doing the rounds for a couple of years now. I was in Brussels recently as well. I see a lot of re-articulation of the problem statement, not necessarily solutioning towards the future. So when you look at, you referenced engineering talent and that DNA as being one area, but if you had to rank the most critical inhibitors to Germany getting back on track in terms of being an innovation led, founder led, DNA led country as opposed to some of the professional class, put you and I in, what needs to happen first? So I think we would benefit from a larger digital single market. The EU tries to build a digital single market, but we still have, I don't know, 20 tax regimes, languages, company formation forms, etc. So that makes it harder if we. One of the biggest advantages of the US is that they have this huge domestic market of 350, 400 million people with strong purchasing power. And if we put whole of Europe together, we had the same or even more, but we are basically still very separated. Yeah, small states, if you will. So that would be good if we have. It would be much easier if it would be easier to found a company in, I don't know, Benelux or in Switzerland or in Denmark. And you could from day one service every European country. [SPEAKER_01] That would help a lot, I think. [SPEAKER_01] Then it would help to mobilize more capital. [SPEAKER_01] However, I also think always saying we don't have enough capital in Europe is a bit of an excuse as well. [SPEAKER_01] Because if something meaningful is built in Europe, see Stripe, capital is here within milliseconds, right? [SPEAKER_01] If someone builds a really meaningful company, there are enough international investors who want to be a part of it. [SPEAKER_01] The hard truth is we are also not building a lot of exciting companies. Revolut, Stripe might be the exceptions from the rule, I'm afraid. I think getting talent into Germany needs to be much easier. That's what actually founders who I invest in tell me, that they would like it to be much easier to get someone from India or from the Philippines or from former Soviet Union into Europe to work here. So that's a process that still takes months, sometimes years. And we need the best talent, obviously. The very best people already go to the US. And if we can't even let the best people from the rest of the world in, that's a big inhibitor of growth, I think. Because we don't have all these people in Germany. We need talents from abroad and we need to make it much easier. If a company vouches for someone, they should be able to hire people within four weeks if they basically take the risk on their side. [SPEAKER_00] Yeah, no, I agree. [SPEAKER_00] Again, another Stripe Atlas, which is our incorporation service for businesses in Delaware and the US, we see a large percentage of that base of founders actually coming from outside of the US and a big chunk of them coming from Europe. [SPEAKER_00] So to your point, because they know they can get access to the single market overnight and it's a couple of hours as opposed to days or weeks here. [SPEAKER_00] If we look at which sectors of the economy, if you're in Germany, you think Germany has the best position to lead on in looking into the future. Or maybe areas of historic strengths that have since lapsed that you think, if Germany gets its act together, can return with the support of AI to being a global leader. Like, are there any particular industry segments that stand out? Yeah, maybe to lay the groundwork. So what we need to understand is that in the world, maybe in 10, 15 years times where AI has or is taking over more and more of the white collar jobs and robotics are taking over more and more of the blue collar jobs, then there are two ultimate resources, which is energy and natural resources like iron, copper, etc. Because basically robots will build themselves, AI will build itself over time. I'm not speaking of tomorrow, but 10, 15 years out maybe. So the only thing that matters is who has the cheapest energy and who has access to natural resources, iron, copper, coal, stuff like that. [SPEAKER_01] And as you might know, we're not in the best position in Germany. [SPEAKER_01] We don't have the cheapest energy prices and we don't have a lot of natural resources. [SPEAKER_01] And even worse, the two countries with the cheapest energy prices and the most access to energy and the most natural resources are China and Russia. [SPEAKER_01] So these are predestined to thrive in a world where business is mainly done by AI and robotics. [SPEAKER_01] So we have to find a way. I mean, we won't win robotics, we won't win manufacturing that for sure is going to China. [SPEAKER_01] China is already producing 80% of the stuff that is in robots. They are able to build robots for less than 10k euros or dollars. So they will win that, just like they won electric cars and software. AI will be hard as well. So we need to find something where we don't need a lot of energy and don't need natural resources. [SPEAKER_01] So, these are predestined to thrive in a world where business is mainly done by AI and robotics. [SPEAKER_01] So, we have to find a way. We won't win robotics, we won't win manufacturing. That for sure is going to China. [SPEAKER_01] China is already producing 80% of the stuff that is in robots. They are able to build robots for less than 10k euros or dollars. So, they will win that just like they won electric cars and software. AI will be hard as well. So, we need to find something where we don't need a lot of energy and don't need natural resources. And I think biotech, life science, pharmaceuticals would be a great industry because that lives mostly from great science, great human capital. We do have that. We still have excellent universities. Lots of the leading AI researchers are actually trained in Germany. Germany and the same is true for life science researchers. And Germany has been the pharmacy of the world in the end of the 19th, beginning 20th century. The times when Bayer, BASF, etc. and the predecessors of IG Farben, etc. were created. We have been doing most of the pharmaceuticals and chemical production of the world. Chemicals are too energy intensive, so we won't get that back. But pharmaceuticals, biotech, life science. We need to build the next 10 Biontechs. We had a 100 billion dollar company in Germany for a few weeks which was Biontech. And it was also built by immigrants, by the way. So we need another 10 Biontechs. And I think we can do that because there are so many new possibilities. Leveraging AI for pharmaceutical development and research. And I think that's the best chance we have. We should also diversify. We shouldn't bet everything on that one card. But I think that's the most promising industry. But it's also one that's threatened because Google DeepMind has their own life science department. And tries to build the next cancer drug as well. [SPEAKER_00] Okay, getting close to wrapping up here. [SPEAKER_00] I know drinks are waiting downstairs in the ExoPol for everyone. [SPEAKER_00] Two last questions. [SPEAKER_00] One startup in Germany that is potentially under the radar right now that you're personally excited about and that is going to set the world right, is going about it the right way. [SPEAKER_00] Under the radar is hard. [SPEAKER_00] Tectile just announced a new round led by Tiger. [SPEAKER_00] I think they raised 130, 160 million something. [SPEAKER_00] So likely a unicorn that not a lot of people know about. [SPEAKER_00] It's from the regulated financial industry. [SPEAKER_00] AI for KYC process, et cetera. [SPEAKER_00] I think that can be very impactful and could well become a billion dollar, maybe is a billion dollar company. Could become a decacorn at some point. If you continue to serve that specific industry very well. That would likely be. And I, via a friend, I did a tiny investment with them. [SPEAKER_01] Which unfortunately I sold part of in secondaries already. [SPEAKER_01] Because the previous round, before the previous round was done by Tiger as well. [SPEAKER_01] And I thought if Tiger is investing, I might cash in on a few of my shares. [SPEAKER_01] But I regretted it a lot. [SPEAKER_01] You still got some skin in the game though it sounds like. [SPEAKER_01] Awesome. Very last question. And we'll try to always leave people on a high note, right? That's what you're told in these fireside conversations. [SPEAKER_00] So if we're sitting here in Stripe Tour Berlin in a much, much, much bigger auditorium in 2030. [SPEAKER_00] With like 9,000 people in and around, something like that. What needs to be true if Germany is back at the forefront of the global economy. Both from an AI standpoint, but also from a German powerhouse traditional industry standpoint. What needs to be true? Yeah, it takes a lot to do it. But I think what will be essential is a very strong open source ecosystem. So lots of the technology we're using today is open source. Web servers, CMSs, databases, not all software is commercial. Cell phone towers are open source. Android browsers, all that is open source. And I hope AI, the LLM layer at least, the foundation model layer will become open source. Right now, most of the open source models are from China. However, I think as we don't have a sovereign German player in LLMs, we need to hope that there will be freely, publicly, openly available models that we can employ. And then we will create the value from employing open source models in our existing industry. I think that will be super important. If not, if we become more and more dependent on US companies, I'm afraid the same will repeat. History will repeat itself. And the same that happened in the digital advertising market will happen, which is not good for the tech situation in Europe, the tech and tax situation in Europe. So open source is important. [SPEAKER_01] I hope we continue to shift our pension system towards capital based, privatize it more so that people invest more in our own economy. [SPEAKER_01] That's good for people because they participate more in business formation or in capital, but also for the businesses that seek new capital. [SPEAKER_01] That would be important, especially in Germany. [SPEAKER_01] We are really terrible at that until now, but a lot has been done now to do that, which is good. [SPEAKER_01] And as I said before, I hope we invest almost as much effort in fostering a new class of startups, of companies that are just about to emerge. [SPEAKER_01] And I think that's not as much effort as we put into maintaining, preserving our existing incumbent industry. If we get that right, if we let people build the next version of this economy in parallel, because I don't believe in transforming large companies. I think you have to build a faster startup economy in parallel like it happened in the last years in the US over and over. Then we still have a chance. We still have excellent human capital. We have the smartest people. If you look at the AI papers in the US, these are all European names. Everyone in the US at some point came from here. So we have the manpower and we just have to get going and start building stuff again. Grasp the opportunity. Okay, awesome. All right, Philip, with that, thank you very much again for sharing your insights. We really appreciate it. What a great way to round out Stripe. Thank you so much. Thank you very much. Thank you. We still have excellent human capital. We have the smartest people. If you look into the AI papers in the US, these are all European names. Everyone in the US at some point came from here. So we have the manpower and we just have to get going and start building stuff again. Grasp the opportunity. All right, Philip, with that, thank you very much again for sharing your insights. We really appreciate it. What a great way to round out Stripe. [SPEAKER_00] Thank you so much. [SPEAKER_00] Thank you very much. [SPEAKER_00] Thank you. Because if something meaningful is built in Europe, see Stripe, capital is here within milliseconds, right? If someone builds a really meaningful company, there are enough international and international investors who want to be a part of it. The hard truth is we are also not building a lot of exciting companies. Revolut, Stripe might be the exemptions from the rule, I'm afraid. I think getting talent into Germany needs to be much easier. That's what actually founders who I invest in tell me, that they would like it to be much easier to get someone from India or from the Philippines or from former Soviet Union into Europe to work here. So that's a process that still takes months, sometimes years. And we need the best talent, obviously. The very best people already go to the US. And if we can't even let the best people from the rest of the world in, that's a big inhabitant of growth, I think. Because we don't have all these people in Germany. We need talents from abroad and we need to make it much more easier. If a company kind of vouches for someone, they should be able to hire people within four weeks if they basically take the risk on their side. Yeah, no, I agree. Again, another Stripe Atlas, which is our incorporation service for businesses in Delaware and the US, we see a large percentage of that base of founders actually coming from outside of the US and a big chunk of them coming from Europe. So, to your point, because they know they can get access to the single market overnight and it's a couple of hours as opposed to days or weeks here. If we look at which sectors of the economy, if you're in Germany, you think Germany has the best position to lead on in looking into the future. Or maybe areas of historic strengths that have since lapsed that you think, like if Germany gets its act together, can return with the support of AI to being a global leader. Like, are there any particular industry segments that stand out? Yeah, maybe to lay the groundwork. So, what we need to understand is that in the world, maybe in 10, 15 years times where AI has or is taking over more and more of the white collar jobs and robotics are taking over more and more of the blue collar jobs, then there are two ultimate resources, which is energy and natural resources like iron, copper, etc. Because basically robots will build themselves, AI will build itself over time. I'm not speaking of tomorrow, but 10, 15 years out maybe. So, the only thing that matters is who has the cheapest energy and who has access to natural resources, iron, copper, coal, stuff like that. And as you might know, we're not in the best position in Germany. We don't have the cheapest energy prices and we don't have a lot of natural resources. And even worse, the two countries with the most, the cheapest energy prices and or the most access to energy and the most natural resources are China and Russia. So, these are predestined basically to thrive in a world where business is mainly done by AI and robotics. So, we have to find a way basically, I mean, we won't win robotics, we won't win manufacturing that for sure is going to China. China is already producing 80% of the stuff that is in robots. They are able to build robots for less than 10k euros or dollars. So, they will win that just like they won electric cars basically and software, AI will be hard as well. So, we need to find something where we don't need a lot of energy and don't need natural resources. And I think biotech, life science, pharmaceuticals would be a great industry because that lives mostly from great science, great human capital. We do have that. We still have excellent universities. Lots of the leading AI researchers are actually trained in Germany. Germany and the same is true for life science researchers. And Germany has been the pharmacy of the world in the end of the 19th, beginning 20th century. The times when Bayer, BASF, etc. and the predecessors of IG Farben, etc. who is, etc. have been created. We have been doing most of the pharmaceuticals and chemical production of the world. Chemicals, too energy intensive, so we won't get that back. But pharmaceuticals, biotech, life science. We need to build the next 10 Biontechs. We had a 100 billion dollar company in Germany for a few weeks which was Biontech. And we need also built by immigrants, by the way. So we need another 10 Biontechs. And I think we can do that because there are so many new possibilities. Leveraging AI for pharmaceutical development and research. And I think that's the best chance we have. We should also diversify. We shouldn't bet everything on that one card. But I think that's the most promising industry. But it's also one that's threatened because Google DeepMind has their own life science department. And tries to build the next cancer drug as well. Yeah, yeah, yeah. Okay, getting close to wrapping up here. I know drinks are waiting downstairs, downstairs. In the ExoPol for everyone. Two last questions. One. One startup in Germany that is potentially under the radar right now that you're personally excited about. And that's that is setting, is going to set the world a lighter, is going about it the right way. Under the radar, under the radar is hard. Tectile just announced a new round led by Tiger. I think they raised 130, 160 million something. So likely a unicorn that not a lot of people know about. It's from the regulated financial industry. AI for KYC process, et cetera. I think that can be very impactful and could well become a billion dollar, maybe is a billion dollar company. Could become a decacorn at some point. If you continue to serve that specific industry very well. That would likely be. And I, via a friend, I did a tiny investment with them. Which unfortunately I sold part of in secondaries already. Because the previous, before the previous round was done by Tiger as well. And I thought if Tiger is investing, I might cash in on a few of my shares. But I regretted it a lot. You still got some skin in the game though it sounds like. Awesome. Okay. Very last question. And we'll try and always leave people on a high note, right? That's what you're told in these, in these, in these far side conversations. So if we're, if we're sitting here in Stripe Tour Berlin in a much, much, much, much, much bigger auditorium in 2030. With like 9,000 people in and around, something like that. What needs to be true if Germany is back at the forefront of the global economy. Both from an AI standpoint, but also just from a, you know, German powerhouse traditional industry standpoint. Like what needs to be true? Yeah, it takes a lot to do it. But I think what will be essential is, I hope for a very strong open source ecosystem. So lots of the technology we're using today is open source. Web servers, CMSs, databases, not all software is commercial use. There's the cell phone towers are open source. Android browsers, all that is open source. And I hope AI, the LLM layer at least, the foundation model layer will become open source. Right now, most of the open source models are from China. However, I think as we don't have a sovereign German player in LLMs, we need to hope that there will be freely, publicly, openly available models that we can employ. And then we will create the value basically from employing open source models in our existing industry. I think that will be super important. If not, if we become more and more dependent on US companies, I'm afraid the same will repeat. Like they say, history will repeat itself. And the same that happened in the digital advertising market will happen, which is not good for the tech situation in Europe, the tech and tax situation in Europe. So open source is important. I hope we continue to shift our pension system towards capital based, privatize it more so that people invest more to our own economy. That's good for people because they participate more in business formation or in capital, but also for the businesses that seek new capital. That would be important, especially in Germany. We are really terrible at that until now, but a lot has been done now to do that, which is good. And as I said before, like I hope we invest almost as much effort in fostering a new class of startups, of companies that are just about to emerge. And I think that's not as much effort as we put into maintaining, preserving our existing incumbent industry. If we get that right, if we let people build the next version of this economy in parallel, because I don't believe in transforming large companies. I think you have to build a faster startup economy in parallel like it happened in the last years in the US over and over. Then we still have a chance. We still have excellent human capital. We have the smartest people. If you look into the AI papers in the US, these are all European names. Everyone in the US at some point came from here. So we have the manpower and we just have to get going and start building stuff again. Grasp the opportunity. Okay, awesome. All right, Philip, with that, thank you very much again for sharing your insights. We really appreciate it. What a great way to round out Stripe. Thank you so much. Thank you very much. Thank you. Thank you.