Poets & Thinkers

“Liftoff” rounds, data moats, and trust barriers: How AI is rewriting the venture capital rules with Pascal Unger

Benedikt Lehnert Season 1 Episode 10

What if venture capital is finally getting the reset it desperately needed? And what does that mean for the qualities and skills required for future founders, startup leaders, and even investors? In this episode of Poets & Thinkers, we explore the future of venture capital and startup building with Pascal Unger, managing partner at pre-seed VC firm focal. From his base in Miami, Pascal brings a unique perspective shaped by his Swiss roots and global experiences spanning coding, consulting at BCG, and finance before diving into the venture world.

Pascal takes us on a journey through the evolution of software – from systems of record in the 1980s to systems of engagement in the 2000s, and now to systems of intelligence that can automate entire workflows rather than just optimize them. He reveals why this shift is creating what many VCs believe to be the largest market opportunity in history, as software can now target not just software budgets but headcount budgets and enable companies to do exponentially more with existing resources. 

Through compelling examples of how his portfolio companies are building data moats and reducing friction to adoption, Pascal illustrates what it takes to win in this new paradigm. However, this platform shift also challenges the VC model to its core because small teams can now go further and faster than ever, start generating revenue early, without requiring to raise a lot of venture capital.

Pascal challenges conventional wisdom about startup building, arguing that distribution and go-to-market strategy are now more critical than ever before. He shares his framework for assessing founders across six key dimensions – from learning speed to moral compass. His insights on the “liftoff round” concept and the compression of funding cycles offer a glimpse into how venture capital itself is being reimagined for the AI era.

In our discussion, we explore:

  • Why software is evolving from optimizing workflows to automating entire outcomes
  • How the trust barrier affects AI adoption and why humans still need to stay in the loop
  • Why data moats and distribution strategies are more crucial than ever for startups
  • The six dimensions investors should use to assess founders in an AI-first world
  • How building has become more efficient while the bar for initial products has risen dramatically
  • Why Europe risks becoming a “museum” due to lack of adaptability
  • Where the VC industry is struggling and how it needs to reinvent itself to stay relevant

This episode is an invitation to understand how the fundamental rules of software, venture capital, and startup building are being rewritten in the AI era – and what it takes to thrive rather than just survive in this new paradigm.

Topics

02:45 - Pascal’s journey from Switzerland to BCG to founding Focal VC

05:10 - The evolution of software: from systems of record to engagement to intelligence 

07:25 - Why systems of intelligence represent the biggest market opportunity in history 

09:50 - The role of trust in AI adoption and keeping humans in the loop 

13:35 - How startups can compete against foundation model providers with proprietary data 

16:20 - Building data moats through integration strategies and reducing friction 

20:25 - Trust-building measures for startups in high-stakes vs. low-risk use cases 

24:10 - Why the minimum bar for softwar

Send us a text

Get in touch: ben@poetsandthinkers.co

Follow us on Instagram: https://www.instagram.com/poetsandthinkerspodcast/

Subscribe to Poets & Thinkers on Apple Podcasts: https://podcasts.apple.com/us/podcast/poets-thinkers/id1799627484

Subscribe to Poets & Thinkers on Spotify: https://open.spotify.com/show/4N4jnNEJraemvlHIyUZdww?si=2195345fa6d249fd

Speaker 1:

Welcome to Poets and Thinkers, the podcast where we explore the future of humanistic business leadership. I'm your host, ben, and today I'm speaking with Pascal Unger. Pascal is a managing partner at early-stage venture capital firm Focal. Born in Switzerland, with global experiences, including working in India and the United States, he brings a unique cross-cultural perspective to investing in AI startups. Pascal and I first met when I was at Stark, where he is an investor.

Speaker 1:

I have a love-hate relationship with venture capital, mainly because, for a very long time, I felt that the industry had somewhat lost the venture and many firms have turned into faster banks. Long gone are the days of investors finding truly groundbreaking companies with a big vision that not only offered exponential financial returns but actually could change the world for the better. Now, as capital is a commodity and AI-powered startups can go to market at record speed, making revenue faster than ever before, the venture capital playbook has to change and the entire VC stack might need to be rewritten. Is this the reset the industry needed and, if so, what does it look like and where do we go from here? I've always appreciated Pascal's measured and thoughtful perspective on company building and investing, so I'm excited that we get to talk about how he sees especially the early stage market evolving, what it takes to build startups in the AI era, new venture models that are emerging and, most importantly, what skills are critical for founders and investors to develop so that they stay relevant in the future.

Speaker 1:

If you like the show, make sure you like, subscribe and share this podcast. Pascal, where does this podcast find you?

Speaker 2:

It's great to chat today, ben. I'm based in Miami, florida, amazing place to be, especially during the winter months.

Speaker 1:

Why don't we just get started and you tell us a little bit about who you are, what you do, what got you to where you are today, and then we'll just jump into some of the really exciting topics I want to discuss?

Speaker 2:

Very happy to. So. I'm Pascal Unger, born in Switzerland and most of my early years there. I was living in a couple of places across the world, coding studying in India, spending a bunch of time in St Louis Missouri as a 16 year old, and so on, eventually moved into the world of management consulting after my studies, specifically the Boston Consulting Group, and as one does in Switzerland, within the financial industry. After a while I transitioned to BCG San Francisco office and did a couple of years in tech with a specific focus on go-to-market, which then led me to team up with someone that I had known since we were little kids and start a venture firm called Focal, where we exclusively lead and co-lead pre-seed rounds. So think the first one or two million that goes into startups try to be as early as we possibly can be, and we exclusively focus our investing on pure software companies that sell into other businesses and that are started by strong technical founders.

Speaker 1:

You know I've always described you and your partner, Daniel, as one of the really good ones out there. Thank you. Having worked with you and having seen many other VCs out there, it's really an interesting time. I wanted to talk with you for quite a bit because I've been reading all the posts and newsletters that you've been writing over the last, especially last year, as so much has been changing in the technology industry and venture capital, specifically in startups, so maybe a good way for us to kick things off would be just for you to give your perspective on what's happening in the world right now. Of course, we're going to talk about AI probably quite a bit but what does it mean specifically for the world that you specifically invest in the world of software, the way you look at the world, the way you look at value creation in that new world, and then we'll use that as a jumping off point.

Speaker 2:

That's great. I guess writing is the best way to really sharpen the thinking, and so I try to do that quite a bit. Unfortunately don't do it nearly as much as I would like to, but that's a different story. If you think about software I'm not trying to give you a history lesson Software started a lot of it in the 80s and software is just basically a system of record. Software is mainly here just to store and process data. Because a lot of these systems of record were super clunky, software kind of evolved into a system of engagement that sat on top of these clunky kind of interfaces of the systems of record and they came out in kind of the early mid 2000s with the sales forces and so on of the world that invented the software as a service model.

Speaker 2:

To me, software as a system of engagement mean the primary focus is to make workflows work productive, and so you kind of look up data inside that system of engagement, you do something with that data and you kind of re-upload it, which worked well. There's a lot of interesting great companies started in around that time and so on. Now, in kind of the last one or two years, software moved from a system of engagement to a system of intelligence right. So meaning that increasingly all of software now has an intelligence layer to it and with that intelligence layer to software we can now move from making workflows more productive to actually automating outcomes. And that's an incredibly different value proposition versus and how it comes with a very different market opportunity versus software as a system of engagement.

Speaker 2:

Because, for one thing, like one of the downfalls of software as a system of engagement is that they the more data kind of you upload into these system of engagements, the more likely they are to break down Systems of intelligence.

Speaker 2:

They get better and more intelligent the more data they have. That's one thing. And then the other thing is, with software as a system of engagement, you sell it into organizations and sell it in a way where humans should use your software, and the more time humans spend in the software, typically the more valuable it gets. With system of intelligence, you're increasingly taking work away from humans and you're basically not just going after software budgets anymore, but increasingly also either going after headcount budgets or you allow companies to do so much more than they've done to date with existing budgets, because us humans can move away from doing a lot of tasks that probably very few of us enjoy, and so a lot of firms, including ourselves, have to the thesis right now that, right, this is probably the biggest market opportunity we've witnessed at any point in time, because software can now increasingly go basically after all of knowledge so the conceptual description which I had read initially, I think, in one of your last articles, makes a ton of sense to me and I think it's interesting.

Speaker 1:

On the one hand, we're still seeing, currently at least, basically the old mindset applied to the new AI world. So there's a lot still of workflow automation or optimization happening. But at the same time, I think we're seeing more and more of this idea of a system of intelligence starting to become real. And then I think the immediate next step that we're seeing is seemingly, as you said, companies then specifically going after well, now we can reduce headcount and then ultimately improve our bounty to improve our profit margins and all of that. But what do you think if we're thinking this a little bit further and say, okay, system of intelligence will be the default, right, how many systems of intelligence will there be? How do you think about that? Let's play this all the way through right. Right now we're in this transition phase, we're in between. There's a lot of AI just being used for workflow optimization still. But let's say we get into this world where system of intelligence is the default, how many systems will there be? Will there be just three or four? What does it look like?

Speaker 2:

That's an incredibly tough question to answer, but I'm happy to give you my take. So I think you should very strongly differentiate trying to make workflows more productive versus trying to automate steps, because everything that comes with automating steps to me is starting to be a system of intelligence and even at the most basic level, it will be a while until we can automate workflows end-to-end. Part of the reason is technical or technical capabilities we're still very early on but the other part is also trust, and one of the biggest barriers to adoption of a lot of these AI systems is trust, and the only way to overcome the trust challenge is to still have a human in the loop for a lot of things and have the human check things and have the human oversight, which means you can't automate too much of it, because then you take the human out more and more. Now, the more we start to trust the machine, the more we're happy to get out of the way and let the machine do its work. It's similar to humans, right? If someone works for you, if you're managing people early on, you're checking everything. Once you trust them, you go into a much more passive mode and eventually you move into a mode where you don't touch anything unless they proactively surface something to you because you trust their judgment and you trust that they actually at least in my opinion the best way to manage people is once you get them to a point where you can basically just completely turn it off and they come to you whenever they know your input is required, and we'll kind of move in a similar direction with the machine. But it takes time and again. The best way is just to start to automate small pieces and then eventually expand from there and early on have human sign-off and eventually you don't even need the human sign-off anymore to proceed to the next step, which then means you can do increasingly end-to-end automation.

Speaker 2:

Now there's different camps out there with regard to what's the future of these systems of intelligence or what are they right? Gonna gonna look like there's some people in the camp of like, uh, we'll have one, whatever agi, that rules it all and we can all, we'll all go to universal basic income and like, basically, the machine does everything for us. I'm not necessarily in that camp. I strongly believe that we're moving into a world where there's a whole, a very, very large set of different agents out there. Right, so we're moving into a world where there's a whole, a very, very large set of different agents out there, right. So we're moving. Or let me start a bit earlier as long as there's humans in the loop, right.

Speaker 2:

It's hard to have fully kind of agentic systems because as long as the humans are in the loop, you can't really automate tasks end to end because the trust isn't there but increasing again.

Speaker 2:

We're moving to a world where the machine can like very basic examples that people use is like hey, I want to go to New York, book me a hotel with these criteria at this price range and like it books it for you, or flights, or whatever it is.

Speaker 2:

I strongly believe we'll move to a world where we're increasingly moving into an authentic world, meaning that as a human, you will have an increasing number of different agents that are trained for different tasks and for different outcomes, that do work on your behalf.

Speaker 2:

Like one of the basic examples of it's not fully an agent yet, but what I use in my work is if I have five calls with new startups tomorrow, what I'll do the night before is I'll use OpenAI's deep research and I have kind of my master prompt for like meeting prep, but basically just shoot off five parallel deep research, almost like analysts, at one time and they come back the next morning. I can read all of those and I typically give it a prompt where, like, I can read in about 15 minutes and it's as if I had five analysts that did two or three days of work for me to help me get ready for that meeting at such a high quality, and so I basically have kind of like five analyst agents that do work in parallel for me that's going to be increasingly common for almost any and all kinds of tasks.

Speaker 2:

And like to your question is will there just be three or four or five? I don't believe so, partially because, right the more, the more specific data you have on specific use cases, the better the outcome will be that you can automate one and then two. I actually also believe that we're moving into a world where we'll have a warm-up period for software, meaning that if you start to use a new system of intelligence, it'll probably take some time until it learns how you do certain things and how you want certain things to be. And if you use it for a certain while, it will actually get better and better and better for you because it is, to a degree, personalized.

Speaker 1:

But, as you were just describing I think you know the amount of data that a system has of you or of a certain company or a certain market the more valuable it will be in reproducing that data, especially if it's historic data, behavioral data, whatever you have right, prompts conversations the higher the lock-in effect is. And then on top of that, you add the fact that it just still, to this day, it takes immense amounts of compute, that there's really only so many companies that can today produce that, at least on a platform level. And there may be ways to think about agentic experiences or agent experiences that are built on a platform level, and there may be ways to think about agentic experiences or agent experiences that are built on top of those. But eventually you want to own that data. I have some more questions about you know how do you then think about startups in this world? So I wanted to get a sense for how you think about this will play out. Yeah, look.

Speaker 2:

I mean, if you believe that our general purpose large models will always be better because they can be trained a lot more data and everything then smaller, much more purpose-built models, then you probably shouldn't be investing in startups.

Speaker 2:

If you believe so much of the world's data is still proprietary and these models don't have access to, and so on.

Speaker 2:

So if you believe in a world where small, purpose-built malls are actually better, then right, there is a place for a lot of different startups and where they'd write a lot of different type of agents, very specific for industries, for specific companies, for specific individuals, and so on, and I guess maybe a transition to how we look at startups, one of the things that has changed.

Speaker 2:

So we, as a firm, we've always believed in backing strong technical founders because, like, speed matters more than anything else in startups, and having strong technical people on the team means you build better products faster. Now what has changed, though, in the last couple of years is that nailing distribution from the start so meaning go to market is now more important than ever before. Right, because, a if you have a lot more people that can build products, but it's harder than ever before to break through the noise. And B if you think that all of software has an intelligence layer to it and if you believe that everyone's building on top of the same technology, second foundation models to get started, then in order to be able to build differentiated products from the start and train differentiated models. You now need to get access to proprietary custom data and workflows even earlier than before.

Speaker 1:

That's right, and you need to have actually even more of a unique insight into a specific use case or market that you can then leverage, or just a way to get in marketing work with customers and gather that insight partially with them.

Speaker 2:

Then the next step of that is then and we think about this with founders quite a bit at the start even is there a path to data mod. So how can you get access to customer data that others don't necessarily get or that others don't necessarily think about, and that can mean you do things that are harder at the moment and that may take a little bit longer to get off the ground, but where the impact compounds over time. If you sell into businesses that have ERP type systems or EMRs in healthcare or anything else, no one's going to rip and replace their existing systems for a new solution at the bar to MVP.

Speaker 1:

It's super high.

Speaker 2:

But if you get in with an initial wedge and basically make it very easy for people to kind of integrate whatever you build into their systems of records and basically start to not just use all the data that they have in there but also everything that comes on top, then eventually, once you go back to trust, once you've established a trust, it's very easy or it's a lot easier to start to replace these systems of trust. Once you've established a trust, it's very easy, um, or it's a lot easier to start to just start to replace these systems of record once you've proven that you can actually do it and once you have all the data in there that they have as well and potentially give it out for free. So you very significantly reduce the friction to adoption because, again, right, getting access to data matters now more than ever, and reducing friction to adoption of whatever you built to get access to data is increasingly incredibly powerful to then build a durable and compoundable moat on the back of it. But we've also invested in a company in a different space that leveraged a lot of open source tools, which is, by the way, incredible right now how good open source has become.

Speaker 2:

Yeah, it's amazing, basically built on top of open source to make it very easy to replace a lot of existing vendors that they pay a lot of money for with open source, and the two components that they built is one, an easy layer to interact with different open source tools and manage them, and the second one is a unified data layer, both of which you'd have to pay a lot of money for right now.

Speaker 2:

But again, because of the value of data, they're giving out all of that for free as a distribution hack because they know once they get access to data, they'll have a lot of insight in what all the outcomes are that you can automate, and that's where a lot of the money lies. But without getting access to the data first, you're not going to be able to automate as many outcomes to begin with, and whatever you're building is not nearly as strong versus if you get access to that data and very significantly reduce the friction to getting access to it. Get access to that data and very significantly reduce the friction to getting access to it, kind of longer term thinking versus shorter term thinking.

Speaker 1:

Yeah, I like that, and I was going to ask you what are then ways to create a moat, ways to differentiate and I think you gave some really good examples there, because everyone, especially in the startup world, of course, doesn't have typically access to a proprietary data set. When I was at SAP, we clearly had, or you have, openai and Microsoft and Oracle and all of the big players that have been sitting on a wealth of data for decades at this point, and that's an interesting point. And that's an interesting point Now. We've mentioned a few times trust, quite a bit actually, in terms of especially the speed to adoption, I would assume. And then there's trust in the company, right, and then there's trust in the system itself.

Speaker 1:

When we're looking at pretty much all the models still, especially the large language models, there's a ton of problems with them still, when it comes to trust, specifically around hallucinations, but also just overall data processing, it's fast, but not necessarily as precise and accurate as it should be. How do you think companies, especially startup companies, can and have to navigate those two aspects which, especially if you're a newcomer, if you're a brand new startup, you don't have either right. You don't have control over the models, you might not even have control over the foundational data, and you don't really have established the trust with the customer yet. So what do you see there, or how do you think about that in terms of, you know, company building? How early do you need to invest in trust building measures, whatever that is going to look like?

Speaker 2:

It's everything to get started right, and that's why so many companies start by building their products in a way even if they could automate more building their products in a way where it makes the human better and where the human is still the checkpoint. And if you can. With a lot of these companies, the question is always very significantly depends on the use case. What's the minimum kind of quality bar you need to meet that people actually perceive it in a way where it adds value and in, I don't know, like in sales use cases where it starts to pre-write emails for you or whatever. You're going to read those anyways. And so the bar to quality is lower versus, say, in compliance use cases where, if you have a screw up there right, it can come with very heavy fines and everything else.

Speaker 2:

And so, depending on what the use case is and what bar you have to meet, it's even more important to make sure that, right, the human is still in control, versus again, right if it's low risk use cases, high frequency use cases where you don't really care.

Speaker 2:

And if, if the the screw-ups are very non-costly, then you can probably try to automate a lot more from the get-go and just increase speed, versus, again right the heavy kind of one-off, very expensive potential like screw-up use cases. There you have to invest a lot more into accuracy from the get-go and you'll probably you'll run into a lot higher friction sales processes, even though everyone wants to adopt AI these days and has more budgets than ever before, which, as a side comment, it's pretty incredible to see that what used to be up to 12 month sales processes can now be a month or two, just because everyone's so eager to adopt these tools and because the value that these tools deliver has also, or it can deliver, is also incredibly high yeah, so on the other side, actually, some of the some of the startups, certainly in certain industries, can benefit from sort of almost like a trust halo from the big ones that are kind of pushing into the market and have that kind of upfront trust, and that's interesting, that's a really interesting key figure.

Speaker 2:

We spend a lot of time on the distribution side and it's similar to if you're a startup and this is a quote from somebody I've spoken to but if you're a startup, a, b plus first sales meeting is a failure, versus if you're at an established brand, that's a good meeting. Why is it a failure as a startup? Because you're already fighting such an uphill battle to get people to move on from the status quo and to trust someone who isn't proven yet that you basically have to blow it out of the park every single time you have a first sales meeting because otherwise, most likely the process is going to stall. Because we all as humans, it's much more comfortable and easier to maintain the status quo.

Speaker 2:

It's similar on the product side and it's part of the reason why a lot of established players have generated a lot of revenue very quickly, because once you have the trust right, then the bar of what you have to deliver partially because these tools are so powerful is a lot lower. But as time goes by and as people start to see what all is possible, one of the side effects of what's happening right now is that the minimum bar that software has to meet is actually going up very fast. Early on you could just hack a few things together and that was good enough to get to sales very quickly. But by now your initial product now actually has to be a lot better than what it used to be for people to pay attention to?

Speaker 1:

Yeah, because we've been so conditioned that has nothing to do with AI better than it would what it used to be for people to pay attention to. Yeah, because we we've been so conditioned you know not, that has nothing to do with ai, but it has to do. I mean, really it started with the iphone, probably that we got so used to software just working so smoothly, being so well made, being so intuitive by default right just out of the box, that I think that is just increasing now, probably with the deployment of of ai systems.

Speaker 2:

That that's just very very fast, partially because these systems get so much better so fast, and we also all get used to the speed of improvement that something that is not in line with what your expectations are anymore, even if it was in line three months ago, you may not even pay attention to it. And, as a side comment, one of the things we also look at when we look at startups is does their product get better and more defensible if the underlying technology gets better? Because, like we're, we're very strong believers that we're going to maintain this rate of progress, and so if, if, if your ultimate product doesn't get better and more defensible, then you're going to get steamrolled, which, by the way, is what happened very early on, once people started to build on top top of OpenAI. That's just going to keep happening.

Speaker 1:

The other thing is also on that note, I assume that you're looking at defensibility in the sense of not just building and essentially hard coding the reliance on one specific model or platform provider, but being absolutely flexible as to whenever. I mean, that was the whole DeepSeek thing, I think, when you wrote something about that as well, when DeepSeek initially hit the market and everyone was freaking out, and especially companies that were entirely built around or platforms that were entirely built around one model, and then immediately get, you know, disrupted, start sweating.

Speaker 2:

Yeah, and I think it's in general. The amount of adaptability and flexibility you have to incorporate and have as a human and then also incorporate in your business is now higher than ever, Because, again, every piece around you is moving so incredibly fast that you just have to be able to change. Even if you invested 12 months into something, a lot of money and everything time went into it, something new comes out within a few hours. You have to just like kill it and leverage whatever somebody else built, because it's just so much better.

Speaker 2:

So that's yeah, go ahead and like that is something, but I think that's pretty broadly understood by now. Again, like that was. We're only kind of two years into this and maybe in the first year people haven't fully understand it, but I think by now most people are building in a way where they can rip out and replace the underlying foundation models very easily, which also, by the way, happens at the app layer, with all the coding, assistants and everything that have all gained incredible traction in a few months. But the biggest question there is stickiness, because as soon as somebody else is better than whatever you used to, you move away, and that's a big question that everyone is grappling with right now.

Speaker 2:

How do we build defensibility into it? Probably the two best answers that I have right now is one memory, so that it gets better for your specific use cases, and the other one is, again try to not just be kind of one-off tools for certain actions, but try to basically automate more and more pieces across an entire workflow. Typically, that's another lens through which we look at startups that the more complex and the more specific to a certain niche or industry a workflow is that sits on top of the intelligence layer, the more likely are a certain niche or industry, a workflow is right that sits on top of the intelligence layer, the more likely are you to be able to build something by working with customers and understanding that well, to build something that is unique and differentiated.

Speaker 2:

Yeah, Now the last thing just because we talked about defensibility so much. The last thing I'll say here is like defensibility early on is incredibly hard as a startup and the most defensible thing is the founders and their execution speed.

Speaker 1:

I was going to ask you about that.

Speaker 2:

That is what ultimately makes the difference, and we don't spend a lot of time no time basically trying to figure out what the mode is at the time, because it doesn't exist, but we try to get people to think about what are things I can do now that compound over time in a way where we can build something that is defensible in this world that is moving so incredibly fast. But, again, the best thing to stay ahead of the curve is to just move faster than everybody else.

Speaker 1:

Yeah, no, I mean, it is a great segue also into the next question. I have A big, big kind of theme question that I have for you around how is actually investing changing? It's always been, of course, about the founders and the team right, the core team, but it seems like even more so important now, given that, as you said earlier, you could be building something and someone kind of sideswipes you and you have to completely change, pivot or scrap what you've been doing and kind of refocus very, very quickly, and you can only do that when you have the right team. And it seems like that is getting even more so exposed now, you know, and it becomes kind of the critical investment point.

Speaker 2:

Yeah, yes, I'll give you kind of three different types of answers. So one is what we weigh when we invest. So it's first and foremost, by far the team, then it's kind of the broader market opportunity, because you do want to be in a space that has tailwinds and by far the least important criteria is the actual idea at the time of when we invest, knowing that so much of it will change given how early we invest. That's one.

Speaker 2:

Two, I think the difference in terms of speed in general, the difference in speed between average and the best went from X to 10X, partially because you can now put so much leverage on the you with all these automation tools that you can execute so much faster than ever before.

Speaker 2:

If you're very good right, which also means, as human beings, the skills that are actually important right, they've moved up and almost like an abstraction layer, from used to be a huge advantage if we can just execute very fast, but now, like managing, like systems thinking and managing different agents and almost different junior resources on your behalf, has become so much more important because you don't need a lot of resources, you have all these systems of intelligence underneath you, and so it is interesting how that's changing human or like the skill set that is required and basically creativity, getting things done right, owning the outcome and kind of thinking ahead is now more important than ever, and it's basically the most important thing. Like as a side comment, it's not at all what our education system is teaching us, because our education system is teaching us how to follow a process and do well with it. That and that's a complete opposite of what is required now.

Speaker 2:

So I'm actually not very bullish on the education system as is, but that's that's not for what is required now. So I'm actually not very bullish on the education system as is, but that's not for a discussion right now.

Speaker 1:

No, but it is actually part of the discussion within the podcast. I actually, you know, part of what I'm exploring is how do we lead into the future and how do we lead in the future and this theme that you just put out here around basically in this world that we're already in right at the very beginning stages, of course, but it will require us to lean much more into these very critical thinking, creativity skills that are inherently human. But our education system, which is basically the product of the Industrial Revolution, is teaching the exact opposite right, and that's the tension and that paradox that we're currently in the middle of is huge.

Speaker 2:

No, 100%, and just being a curious person is worth so much more now because you can. Basically you can now learn anything because everything's available to you, like for me. I'm personally incredibly curious, but I hate doing the research. So now having these tools that do the research for me so I can just focus on learning new things and skills is amazing.

Speaker 2:

The what I'll say about education is like I tell this to a lot of people if I, if I was 18 again and had the majority that I have now, which I did not have as an 18 year old, I would not go back to university.

Speaker 2:

One of us probably did exactly but like, but like in, and I don't know if it would have been a good idea for me at the time when I was 18. But like I will not go back to university anymore because you learn so much more by doing, versus all the theoretical stuff. Now, what I recognize right, I grew up in Switzerland where education is free, where there's a social safety net and everything, so it is coming like from a position of privilege to say, hey, I don't need the tick mark and the social status that comes with university because I always have a fallback option, right, but that just from a pure skill set and everything perspective. Any and all information is available on the internet. You can learn anything, and you learn by far the most by just doing Again university. Really, unless it's a very specific science or if you want to become a doctor, I think it's still the right place.

Speaker 1:

But for me, who has a business background which I would actually, by the way, never study again like it's pretty much useless yeah, I know, I see it, I see it with my students um, it's absolutely incredible how they're making such amazing use of the technology that's available to them now, right, and then they're applying their creativity to you know, they're really starting to build things, um, and and that they learn so much more than they would kind of studying textbooks, right? So applying that knowledge is a lot more useful, a lot more valuable, and putting that creativity to use.

Speaker 2:

So the third dimension is just how we assess founders. There's kind of six dimensions that we assess. First one is learning speed and learning willingness. Right, which are different things but they're both incredibly important. Like the faster you pick up on the information process it do something with it, the more likely you are to move your company in the right direction. Then also learning willingness is just feedback seeking and reflecting. If you don't do that, you don't really get better over time.

Speaker 2:

The second dimension is execution speed and conscientiousness. Not every founder needs to have that, but as a team you need to have, especially the second piece. And so execution speed, I think, speaks for itself how fast can you get the most important stuff done? And conscientiousness is just about discipline, kind of organization, reliability. At least someone on the team, I still believe, needs to be very good at that, not every founder.

Speaker 2:

The third dimension is kind of adaptability, flexibility, performance, risk taking. For the reasons I discussed earlier, that's now more important than ever because you basically have to. You get new situations with the speed of progress. You face new situations so much more than before, so adaptability has become even more important than ever before as a startup founder. The fourth one is just emotional stability. Startups are a roller coaster and it really is important to keep a straight head because everyone's looking at you in tough situations to make the right decisions. The fifth one is classic and this is basically like grit, but grit to me is two things. One is kind of consistency of interest and the other one is perseverance of effort, and typically the combination or consistency of interest is once you like something and are passionate about it, you really go deep into it. Then perseverance of effort is just right, the sheer willpower to get things done over time.

Speaker 2:

Then the last one, and probably most important one, is just kind of integrity and the right moral compass, because we don't want to work with people that we don't enjoy working with and that we don't where we don't see the world kind of the same way, because life is too short, um yeah, to work with folks like that yeah, that.

Speaker 1:

I love the clarity of those parameters and those qualities especially the middle three, if I counted right are interesting because they're sort of intertwined right, like if things get really tough, which they often do in a startup, then you might not have the emotional stability to be flexible and you might be more stuck in your way. But I think it's a really great way to assess founders. In fact, as I was listening to you going through them, I feel like almost they're describing general human qualities to thrive in this world that we are currently more broadly, when we're looking geopolitically, that transformation that is not just a technology transformation, but actually really transformation of the way we define work, the way we create value, the way companies are being run. That is, everything is in flux, which is really opposed to the human nature of wanting things to be safe and stable and secure. So it was really, as I was listening to you, I was reflecting on that. That's probably what we need to teach more broadly if people want to thrive in this new AI everything world.

Speaker 2:

A hundred percent and, as a fellow European, right. What he said about stability and keeping things as is and like the past is a guarantee for the future that the longer we are or like the longer we are or like the the the longer we are in this period, the less like. That is something, um, that's helpful and, and I actually worry, right, and part of the reason why I'm in the us and why, um, I'm doubling down on being here and everything, is because I like for what we do for work there's like I had lived in Asia and India and in China, but like I don't really know those those worlds anymore.

Speaker 2:

So within the world that I know, the US is by far the best place to be, and I'm actually incredibly worried about Europe just becoming a museum for Americans and Asians at this point because of the lack of speed and willingness to change things Like the US approach. There's a lot of challenges with the US right. To me, europe is still the best place to live, but for work the US is the best place. Europe's general approach to things is just to talk about them for 10 years, versus in the US, the general approach is to just do something, even if it's not always the right thing, but at least people do things and they learn and they adopt and, like some of the startups right, the more you have a lean towards doing versus thinking, the faster you actually learn over time.

Speaker 1:

Yeah, I was going to ask you about that because you know I have a personal probably love-hate relationship with venture capital, especially tech venture capital.

Speaker 1:

But I've always appreciated your and your partner Daniel's perspective, because you have a global perspective, you have something that is beyond the just move fast and break things perspective, which I think, if nothing else, the European sensibility or sensitivity for at least risk assessment is probably useful, because we have, as a tech industry, not just made the world a better place right, and I was wondering how you see the venture industry more broadly, kind of grappling with it or not grappling at all, because the speed is just increasing and with that I wonder if there's a dominant narrative that would just keep doubling down on what we've been doing rather than maybe assessing of, okay, well, this is working, but this is not working, and I was always hoping that Europe could play a little bit of a counterbalance there. But I'm with you, I think so far it hasn't. So I was wondering, kind of, what your perspective is more globally, given where we are geopolitically right and there's a tech and geopolitics is now, if nothing else, so intertwined. So I don't know if you have any thoughts on that.

Speaker 2:

Yes, I won't comment on the politics these days, but when I describe the US versus Europe, I always say the US is a very individualistic society and Europe's a very community-based society. On the upside of a community-based society you have social safety net and people are much more embedded into their communities, which means you have a lot more friends and a lot more friends and everything and a lot more kind of rewarding social relationships. But on the downside of it you also have to, as an individual, you have to fit into community meaning right, you have to fit into a small box, you have to do certain things a certain way. So many people back home that I did, I know, right, they have to have a certain job, they have to drive a certain car they have to go to certain places on vacation, because those are the social expectations and you don't really want to change anything because, again, right, things always continue the way they are.

Speaker 2:

The US is an individualistic society. Right On the downside, you don't have a social safety net. People are a lot more lonely. Interactions are so much more transactional. There's there's a lot less trust amongst human beings. But on the positive side, as an individual you can do almost anything and people encourage that. Folks get so much more done and if you do get something done, everyone will applaud you, whereas in Europe, if you stand out of the norm, people look at you in a strange way.

Speaker 2:

And I always say, as a European, it's incredible to spend time in the US to learn what individual excellence means and to learn that everything's possible. Individual excellence means and to learn that everything's possible, but for every American. It would also be amazing to live in Europe to learn what it means to actually live as a well-functioning society and respect each other a lot more as human beings and also not just optimized for themselves but like the people around you kind of the greater good. And so I think that is also reflected in the industry, and in the US a lot of things are just very monetary driven and if it makes financial sense, people will do it and they care less about everybody else around it. Again, the good side of that is, if you can combine social good and monetary incentives, then the US is by far the best at actually having a real impact and changing things, because, again, they just build things versus talk about it for 10 years.

Speaker 1:

Absolutely fair. Absolutely fair. Yeah, I have a couple more questions for you, One being, I mean, everything is being tested currently, pressure tested in terms of you know, how does it fit into this new world, and that includes your you know your industry, venture investing. What do you see as the biggest risks and the biggest opportunities that you know the venture industry, specifically, is seeing right now?

Speaker 2:

Again, good question. So I guess there's a couple. The biggest opportunity is just like now, everything's possible, right, more so than ever, and again, with the intelligence layer, this is the biggest, I believe, like everyone says it, this is partially self-serving, but I believe that this is the biggest platform shift. What is interesting? A couple of things, I think, in venture. So the industry has changed quite a bit on many aspects. For one, the biggest players in town have raised so much more capital than ever before and they're sucking up a lot of oxygen, which is how the kind of bad industries evolved and everything, and I think it was always going to go that way.

Speaker 2:

But it's interesting if you look at the venture or the venture industry as a whole.

Speaker 2:

So, if you, we talked about kind of startups and individuals and more leverage. So what I very strongly believe is that, right, startup building has become a lot more efficient than it's ever been. Again, right, individuals get so much more efficient. And maybe, as a side comment, because of what I mentioned earlier, what we look at or assess companies through, one of the questions I ask everyone is what is your individual and your company's AI tech stack for across every function, and what have you tried and experimented with, not because I care a lot about what they actually chose, but because I want to see and understand what have they tested and what are they doing, because I strongly believe that we've kind of crossed a chasm with regard to these tools being so powerful that if you don't leverage them and use them for every part of your business, you're just not going to be able to keep up right so? But if you do leverage them, it's insane how much more efficient startup building is, which means right.

Speaker 2:

if you believe that three people can now do the work of 10, and soon 50, then you also have to believe that founders can now get a lot further with every dollar that they raise, especially early on and if you believe that, then right, there's a lot further with every dollar that they raise, especially early on, and if you believe that, then right, there's a lot fewer early entry points than today's pre-seed, potentially multiple seed rounds until you get to series A.

Speaker 2:

Interesting yeah, and so as an investor, I actually I'm quite skeptical of being kind of an early traction on the rider, which the classic seed round is, versus being laser focused and being first, because if you believe that founders can get a lot further, then there's probably going to be an early kind of we call it internally a liftoff round, for lack of a better term. Again, this is partially self-serving, but that's the one to two million that we specialize in and focus on and we try to get in as early as possible. The last investments we did was ahead of the company. We signed the term sheet before the company was incorporated, signed the docs, so later the company was incorporated and with that one or two million they can now get a lot further until they eventually raise a larger quote unquote expansion round to really double down on what they're doing. And that's typically the round where a lot of the large funds come in. And so if you're not incredibly focused on investing first right, then you kind of miss, increasingly miss the middle entry point, especially for the best companies, which is what venture capital is all about. Then you have to compete for these larger rounds with the largest firms and that's just an incredibly tough place for kind of the middle of the pack. So that's one risk to the venture capital industry. Then the second one there's a lot more.

Speaker 2:

But another one is, with the large firms raising so much money and sucking up so much of the oxygen, they have to deliver incredible outcomes to even pay back their funds.

Speaker 2:

Venture as an industry was basically going to go through an incredibly rough time after the 2021 peak and everything that came with it, because people just invested valuations that didn't make sense and a lot of these ventures are not going to be very good. But AI saved them because there's a new hype and platform shift and everyone wants to be part of it, which allowed so many people even if they put up mediocre performance it's too early to tell, but like very mediocre performance between 2020 and 2022, they were able to raise such gigantic funds and they're hoping that ai will like correct all of that right. That allows the industry to thrive further. But if ai doesn't deliver that, then we're all basically screwed, and the larger your fund, the bigger the outcomes have to be in order for you to actually make money and for the industry as a whole to make money again, given the concentration of capital into just a few firms with such large funds.

Speaker 1:

Yeah, and that's probably love the way you phrased the new terms for these new types of rounds, which really reminded me when you talked about how you're looking at your own business and where you want to invest and how you have been invested now in these last few companies, which is there is actual venture in that investment strategy and being so early and taking real risk into founders, into ideas and the ability to pull something off that will have an outsized impact and, in turn, an outsized return which I think for the last maybe two decades we certainly haven't seen as much, because it was such a basically just kind of stacking of rounds and it was very predictable and basically spreadsheet exercise more than it was an actual venture discovery exercise.

Speaker 2:

It's hard for me to comment on because I haven't been in the industry that long and I think one of the biggest challenges that I faced early on hopefully is also a big strength. I didn't come from the industry and so I basically thought we'll see right. Part of the hardest thing about this job is not knowing if you're going to get it for 10 plus years. But we try as a firm to basically look at everything from kind of first principles and just see how do we want to do things and how do we believe things should be done, even if it comes with the risk of spinning cycles and having to experiment a lot and not all of these experiments working out. But I do strongly believe, like we're a startup ourselves and I strongly believe that that will lead to better outcomes.

Speaker 2:

One of the differences between how we invest now versus how we got started is that we are pushing very heavily to be much earlier than we were before we got started.

Speaker 2:

Our website everything says your pre-seed lead Again, if the right point was, founders are a couple of months in that they're actively going out to raise a pre-seed and so on.

Speaker 2:

I don't think that works anymore, like we really try to very actively be involved at the very start, because that window between having nothing and having meaningful things to show for went from anywhere from six to 24 months to now at times like one or two months, and so if we're not very early, it just doesn't like our phone model and everything doesn't work anymore, which then means right, and I spend a lot of the and like thank you to deep research and all the other tools that helped me with a lot of the information gathering.

Speaker 2:

We're kind of transitioning from one phone to the next and as part of that I took a step back on a couple of weekends and spent a lot more time in psychology papers and trying to understand how to do a much even better job at truly assessing human beings and making that an even bigger part of our assessment. Again, because of how early we invest and how much of a. It's always been a people bet, but it's very different but actually putting a lot of weight on the human assessment and using psychology frameworks versus spending a lot of time on the idea and the company and everything else and kind of having that then reflect on humans. And we still spend a lot of time on trying to unpack the idea and everything, because I think it says a lot about how a person thinks. But also doing a lot more specific, deep dives on the person itself is something that's become a much bigger part of our diligence process in recent weeks.

Speaker 1:

Really interesting that the shift in technology platform shift in technology is leading you to assess the human qualities of the person even more so. That's fantastic. We're way ahead of time. I appreciate you so much, pascal, for the very honest conversation. Thanks for having me and great to chat. Great to see you again. Thank you All. Right, that's a wrap for this week's show. Thank you for listening to Poets and Thinkers. If you liked this episode, make sure you hit follow and subscribe to get the latest episodes wherever you listen to your podcasts.

People on this episode

Podcasts we love

Check out these other fine podcasts recommended by us, not an algorithm.

Pivot Artwork

Pivot

New York Magazine
The Prof G Pod with Scott Galloway Artwork

The Prof G Pod with Scott Galloway

Vox Media Podcast Network
Science Vs Artwork

Science Vs

Spotify Studios
5 Year Frontier Artwork

5 Year Frontier

Daniel Darling