Managing Partner at Chemistry: Ethan Kurzweil

What you need to do to be a startup founder or be a leader of any type of a church, of a company, of a startup, you need to imbue followership in your flock. This is Ethan Kurzweil, founding partner at Chemistry and former Bessemer VC who backed Twitch, Twilio, PagerDuty, and Intercom before they blew up. In this episode, he reveals how great founders think and what separates the ones who change everything.

If you see from Brian Johnson or if you see from even Elon or Sam Altman, every person and every organization that started building this cult-like opinionated version of the world usually starts from the Bay in San Francisco. I think Silicon Valley, you can come, people want to know what your ideas are, what you're doing in the future, and they don't care so much about are you some established thing or not because we know that the best ideas always come out of left field.

Would you say that this is a theme across good founders as in being very opinionated and having a view of the world that is very different than the others?

I think so, yes. The best venture capitalists don't necessarily dream of the best companies, you know, companies that put rockets into space. What I'm really attracted to are the companies where they've built something that can't easily be replaced by a third-party tool, and that's pretty hard right now because...

So you're saying that basically you need to build some sort of...

Followership. Likeness. Cult leadership, followerness. I mean, it's the same thing we were talking about earlier about what you need to do to be a startup founder or be a leader of any type of a church, of a company, of a startup. You need to imbue followership in your flock, but once you do it, it's a very powerful place to be. I think the social media age has enabled more people to do it because it just amplifies their views, and so they'll find an audience or people that maybe are susceptible to that chain of logic.

In some ways, it can be good; it's not all bad. The bad is sort of, you know, I think we're aware of that from our political system, but the good is that you can take action and mobilize really quickly. It's a very powerful tool that a lot of people have used for good and bad purposes, just like all technology is.

I see. But that requires you to share your opinions, and most people avoid doing that online.

Right, but I think there's enough people that start to do it and get a positive feedback loop from doing it that they do it more and more. In the past, you couldn't really just share your opinion via the old traditional media forms. There wasn't a way to do that beyond your local community or people that you could do one-to-one. So now anyone can share their opinions. Most people are reticent to do it or get ignored, but enough people are doing it, and then people will kind of pluck them out of the crowd and say, okay, this person has contrarian hot takes.

Think of the most extreme members of Congress now, the ones that are in the news all the time, or some of the podcasters or some of the conspiracy theorists that have big, big audiences. I think it's a lot because they tested their ideas in small scale, people liked it. People that were susceptible to it liked it, and then it just fed itself.

Hmm. So that sounds like the Brian Johnson story a lot, isn't it?

Yeah. Maybe a little bit. I mean, maybe in the past he would have just been, somebody would have vetted him out. People are like, oh, that's too wild. Who is this guy? He doesn't have the normal health credentials. And so he wouldn't have been able to get his ideas in front of enough people that he could have gotten a following. And now, I don't know his rise, I don't really know, but my guess is that it built on itself.

He happened to be at the forefront of a movement around biohacking and longevity that I think a lot of people were interested in. And then he was just the most extreme version of that. So the people that were already excited about it were excited about him.

I guess the first question I'm going to ask is going to be relevant to that before we jump into your story. If you see from Brian Johnson, or if you see from even Elon or Sam Altman, or even Y Combinator, and every person and every organization that started building these cult-like, if you like, opinionated version of the world usually starts from the Bay and San Francisco. And you spent a lot of years here. Why is this happening?

Well, it's a good question. My observation about the Bay is that it's the place that's the most open to new ideas and new people. I'm from Boston, Massachusetts, which is a great city. It's a great place to have grown up. But I would not put Boston on the list of places with a culture of openness to new things. There's a reason Mark Zuckerberg left Boston and came to Silicon Valley to build Facebook.

And there's a reason that Google was started here. A lot of the really transformative technology companies of our age are from this area. I think Boston has a little bit of a closed-mindedness to new approaches that's kind of embedded in the Puritan-ness of the culture. A kid hacking in their basement wasn't quite accepted as a viable person that should get the credibility of someone that had worked their way up through Harvard or MIT and spent a dozen years working their way up to the top or didn't come from a certain sort of family lineage that was well established.

I think Silicon Valley, you can come, you can be here. It's a very transient population. I've been here 17 years now, but I don't feel any more like Bay Area than someone that got here last week. People want to know what your ideas are and what you're doing in the future, and they don't care so much about are you some established thing or not, because we know that the best ideas always come out of left field.

There's kind of an openness and a willingness to accept that reality here that I don't think exists in many other places. That's my guess as to why, despite a lot of people having said over the years, oh, there's a democratization on ideas, the internet makes communication kind of free and open. There really isn't another place like this. I don't think there ever will be, despite whatever the trends are. I think that is a very magnetizing, powerful pull, probably will persist for a long time.

I'm wondering what's the philosophical reason for that. I have exactly the same experience. I think the one thing I noticed in the Bay is I spent almost 10 years in London. When I came here, my company became relevant from day one, which was not the case in London. In London, to do that, you'd need to just basically start meeting people, climb the ladder. It has these steps sort of effect. I'm wondering what's the axiomatic philosophical reason behind that.

Well, it's probably born in the Bay Area's counterculture ethos. Maybe San Francisco specifically, it was the site of the hippie movement. A lot of those various fundamentalist religious cults were based around San Francisco and the Bay Area.

Was that the case, by the way, which cult?

That was the one that was down in New Guinea where they had this sort of tragic ending. I forget the name of this, but it took its roots in San Francisco and in the Haight back in the seventies. It'll come to me the name, but I do think there's a weirdness to San Francisco and the Bay Area around acceptance of kind of new radical things that transcends a lot of the other forces of society. Part of it is kind of like openness to that.

There's also kind of like a constant reinvention of specifically San Francisco. There've been many waves and the cycles go kind of fast. It's not a static place. Maybe philosophically that kind of feeds itself. There's an acceptance of radical thought that I think will be kind of, it's hard to inject that into a society like London or in Boston where I'm from. Even New York that also has a kind of openness to new people.

There's a ruling class in New York of financial institutions that I think, I'm not sure you'll ever sort of break into that. Maybe in the fullness of time, but it may take a while.

For the audience that is listening and seeing you for the first time, can you give them an introduction?

Sure. Ethan Kurzweil, one of the founders of Chemistry. Chemistry is a new early-stage focused venture capital firm. We lead seed and series A investments in a sort of high conviction focused way. We do five to 10 new investments per year across our entire partnership. We have a very focused, concentrated approach to working with startups. I spent 16 years before launching Chemistry with my two partners, Christina and Mark at Bessemer Venture Partners.

I was one of the managing partners there. I led investments in developer platforms, developer tooling, as well as consumer companies, led investments in companies as diverse as Twitch on the consumer side and Twilio and PagerDuty, Intercom, and other companies like that. I'm really excited by new tech trends. I love early-stage hands-on investing. I like investing according to a theme, but I'm very swayable by new technology, by a passionate founder that makes me think about things differently.

While most of my investments have been in a couple of different areas, I'm always open to something frontier or out of left field or new that has the power to be revolutionary for business or for consumers or for some process in the world.

You mentioned Twilio, you mentioned developer tools. I've seen an interview that you've done in the past, it's been many years, I believe, and you were mentioning that Twilio was one of the, probably the first developer tool or API-like business. How does one spot a Twilio, what did you see at the time?

I think it comes back to the founder question in that case. The traditional wisdom when Twilio was getting started was that selling to developers was a bad idea, that it was kind of like a middleware approach and you'd never get big being a middleware company because developers didn't have the ability to tap into big budgets in their enterprises and that they would be like experimenters or tweakers and that developers weren't a particularly important function for businesses.

This is like 2008 when Jeff got started. I think what we did well in spotting Twilio was spotting Jeff disagreed with that orthodoxy fundamentally, that developers are actually the most important role, the most important class of people in businesses because they have the ability to build transformative things in a way that any business line, sales, marketing, even product, finance couldn't do, that they could allow businesses to build the future in a way that no other function had the potential to do.

And then he saw that that was going to change rapidly because of cloud computing because cloud computing allowed developers to build full-featured functional applications without kind of permission from anybody. They didn't need a big upfront expenditure of capital like provisioning servers and things like that to get permission for their projects. They could just spend $8 on AWS and get enough capacity to build something.

And he had this sort of radical, what now seems like obvious, but this is, we're talking about 17 years ago, it was a little bit radical contrarian point of view. And he also had an ethos about usability and functionality that was very consumer-like. I'd sort of thought when I started my investing career I was going to be a consumer investor. So it appealed to me, this consumer-like mentality that Jeff had.

And that was about the developer experience being great. No one thought then about business user experience being good. It was like, did the software do these things? Does it have this functionality? And Jeff was like, no, no. Is it wonderful, beautiful to use? What is the DX like for developers? And so he built a very simple product that had five primitives, could do five things. And he boiled it all down to those five primitives.

And this was even before their SMS application, it was just for calling. So it was things like say, it was like you could speak out loud, receive, you could hear what somebody said and record a clip with it. And it was just this sort of different ethos that he had. He was coming at the problem of communication applications from an orthogonal point of view. And he had a much bigger vision than just communication apps.

It was about empowering developers to build really beautiful things in the world. And so I think just being kind of like open to, wow, this person is really compelling was at the root of it why we were attracted to making that investment.

So would you say that at that time he was very deliberate about this or would you say that just over time created this picture?

He was very deliberate and opinionated from the first moment, as I recall it. He was a compelling showman. Think like Steve Jobs or something like that. Jeff's very famous for coding on stage in a not pre-planned way in front of thousands of people. And it doesn't always work because code is brittle, right? But he has that show and he's even in a one-on-one where he's pitching the company or I mean, there's a few people there, a couple of people.

It's impressive. But he always had this point of view around empowering developers. I mean, he got the idea for Twilio, as I recall, because he ran an extreme sporting goods retail store. He was briefly at AWS product manager, but before that he ran this business, I think it was called Nine Star or something that was a sporting goods store. And he wanted to build apps for people to check on the progress of when their snowboard was going to be ready to be picked up.

And so, he wanted to build like you could call and put in your customer number and it would spit back out from the system status or something like that. And so, he built for his own very minute problem. Interesting Shopify is another example of that Toby built for, he was a snowboard, he had a snowboard store, e-commerce store and he wanted to make the process of setting that up easier. So, he had this experience and had this root of it, but he also saw that he could empower a lot of people to solve their own problems.

That was what was really compelling about it. He didn't want to just solve one problem. Imagine if he had just built a module for building response applications to retail store checkout procedures, that's not a very big market, it wouldn't have been that interesting. That was just my little problem at Twilio, but I can empower all the people like me who know how to code, who have experienced coding to solve their own problems.

It was unique. A lot of people talk that way now, but as I recall, this was very early in the company's life and it was just four people, him, his two co-founders and Danielle, his first employee and he was already using that language. So I don't know when it came to him, if it was even before he started it, but that was a lot of how he talked about it.

Would you say that this is a theme across good founders as in being very opinionated and having a view of the world that is very different than the others?

I think so. Yes. Every good founder has a sort of version of the case that they think they want to prosecute that's different than the way things are now. I think actually like PagerDuty is a good example. Alex and Andrew started that company, they were DevOps folks on the Amazon team. So DevOps like responsible for keeping things running, keeping the server infrastructure and the applications going.

And their view of the world was, it's not that radical. There was just that you should be able to control the experience of being a DevOps engineer so that when alerts were generated, you could determine who got them and when with what severity escalated to whom. It was not radical, but it wasn't possible before without a lot of custom work. And so he's like, you know, this is the way this job should be done.

Let's build a piece of software that's at its core, fairly simple, but complicated to pull off in a way that really meets the needs well and bring that to the world. So there's always an opinion there that's in the best, best companies. There's probably some exceptions, but what I would look for is an opinion that the way things should work, that needs to be enabled by technology, that without that technology is really hard or difficult or requires skills that those companies or people don't have.

So Strikes maybe this is opposing to the YC philosophy. Maybe like you have on the one side that it's a very opinionated, a very contrarian version of the world. And on the other side is, I'm going to be iterating with customers until I find the way.

Yeah. It's an interesting dichotomy there. I think there's a consistency though. So it doesn't have to be contrarian, I actually just got in this debate with Will at Slow Ventures about to be right, you don't have to, to make it for a good investment doesn't necessarily have to be contrarian. Will's argument was it has to, but the best are non-consensus. So non-consensus is broader than contrarian, contrarian ideas are non-consensus, but there's other ideas that are just not the consensus.

They're not necessarily opposed to some consensus, they're just maybe out of left field. And I think first I sort of disagreed with him and argued with him, but then as I kind of reflect on it, I actually think non-consensus is a better framing for this.

But maybe independent.

Independent and non-consensus, I think that's maybe a good way to say it. And so you can get to a non-consensus independent point of view by iterating with customers. You can also do it just by being brilliant and having this idea. YC would probably say that you're going to be wrong more often, too often, unless you're like on the ground with customers, but there's outliers to that. There's Steve Jobs style thinkers where they don't really, they just sort of figure out, create a need for people or tap into a need they already have that they don't know about.

That's one way to get to an independent thought. But I think you can do it the YC way and a lot of companies do. But I think about the best YC companies, Airbnb, Dropbox, Stripe, it's not really a YC company, but it is associated with them. Is Stripe a YC company? No, their previous company was. But they have a very tight relationship. I don't know that they iterated. I could be wrong. I think they had those ideas and got them out in the world and then iterated with customers.

So it's like both things are true. They had a kind of vision and direction as to the way things should work. And then they were very hands on about it, but they weren't hands on about coming up with the idea. Now, you should fact check me on that because I don't know those histories well. But in the retelling, it's usually like more of an epiphany.

What I'm trying to is like, you already invested in many billion-dollar companies. From your view, did you see the one side of the argument or the other? Or is it the mix and match?

I mean, it's certainly a mix, but I'm always attracted to the founder where their argument for the way things should work is super compelling and differentiated and maybe even contrarian. And I've used the word contrarian deliberately there because that's a little bit my sort of renegade-ness. But I've also invested in founders where it's just sort of very, they've sort of incrementally, they've pivoted into something or they've incrementally iterated their way into something that created a very nice, good business outcome for them.

I think about the Sengerton founders, a company I was involved with. And they had a vision for making kind of like algorithmically generated email easy. But it wasn't like they came out with a world-changing point of view. It was very much a like, solve my problem and people like me kind of, and it just so happened that they solved it the best of all the companies of that particular era and eventually went public on the backs of that.

So that is a perfectly viable approach. And I've done both. But I do think when I look back and I look at the arc of companies I invested in or wish I'd invested in, I think through the founder ethos back in the kind of earliest stages, the moment that chemistry seeks to partner with them, there's usually an independent point of view there that isn't influenced by traditional orthodoxies. And sometimes it's pretty heretical.

If we go back to your time in Bessemer, can you walk me through how did you learn about Bessemer in the first place? And then how did your career change through the years? Did you see the fund evolving in different ways?

Oh yeah, it evolved a lot. So I joined Bessemer because I pitched Bessemer, I pitched David Cowan specifically on a startup idea. I had worked as a, I had had an internship consulting engagement with Linden Lab, which is a company behind Second Life, if you remember that company, Virtual Worlds. And I was coming off of that with an idea that I pitched David on that had some similarities to Second Life.

And he said, you know, this idea is kind of terrible. You shouldn't pursue this. I don't think it'll work, but I'm looking for someone to work for me here and you'll kind of learn about great ideas and then, you know, give me two years as a colleague at Bessemer and then you can go on your way and pursue some better ideas because you'll have learned from the startups that pitch you over that time period.

And it was an appealing pitch and I took him up on it, not really thinking about venture capital as a long-term career path because you didn't, it wasn't a very, very common career back then. There were very few venture capitalists. It wasn't clear how to break in. Most, I was kind of came at it from another.

That's interesting, by the way, today, everybody was.

Very different, but this is 2006, 2008. This is like 17 years ago. It was a very different point of view, but it just, it was like a small industry. It's like a cottage industry. And I joined and I learned a tremendous amount. I'm so happy I met David and got to learn from him. You know, he had by that point was almost 20 years as a venture capitalist. He's still now whatever, 30 plus years as a venture capitalist.

And he had a certain style and a certain approach, some of which I'm similar to and some of which I'm different. But it was a really interesting and fortunate position to be in for me. And I never thought I would stay in the career indefinitely, let alone stay at Bessemer for 16 years, but I got to experience the post-global financial crisis era of venture leading up to the Zerp era, the COVID era, the Zerp era, and the boom and bust cycle that came from both of those things.

And that was a really formative, you know, bit of learning for me and a set of experiences. Into your question, Bessemer evolved over that time period in many different ways. There's many different cycles to the firm it is now where there got to be a certain point where I actually felt like stylistically, I wanted to do things differently. And that was kind of a bummer because I'd been there for so long and was so close and still am to so many of the people there.

But I felt like the best way for me to bring my innovation to the world was through starting a new venture firm. And I didn't know the path exactly for that, but it just so happened I had two others that we have pretty similar visions for what this could be and how we could start something new and take an old approach, but sort of give it a modern twist. And that actually, in a funny way, that was different than the way other firms were approaching things today, including all of our predecessor firms.

And so once the idea was kind of staring and screaming us in the face, it just became so exciting and interesting and fun to work on to the point where I had to do it. And so that's what led to the founding of Chemistry.

To answer your other question, the industry evolved to where a lot of the firms that had been kind of early-stage specialists decided they wanted to become multi-stage firms and move into different asset classes and different stages. And my view and Mark and Christina's view, my co-founders at Chemistry, was that focus is going to make you the best, just like a startup. It's hard when you're defocused.

Why would you say that this is defocused? The way I understand it, that funds become bigger over time, they go deeper and deeper in the stages of the bigger companies. Why is this defocused?

Well, I think there's two ways to become bigger and bigger over time. There's what you just described, but it's not exactly how people are thinking about it anymore because then they wouldn't have gone quite as big as they've gone. I think that would be an argument for sort of growing the asset class more incrementally according to the kind of opportunity set of your portfolio. What people have done is be gone into multi-stage investing where they want to invest everywhere from kind of seed to IPO to kind of pre-IPO and even beyond IPO.

There's some firms that have hedge funds associated with them where they're buying and selling public companies too. And so that's just a different... You have people doing lots of different things. You have people looking at companies from one lens, you have people looking at companies from another lens. And it becomes not insurmountably difficult, but harder to manage an organization where you've just got things that are a little bit sometimes at cross purposes.

Because you do have people looking at later stage companies that may conflict with the point of view of the people looking at the earlier stage companies.

How is that? What do you mean?

There's only going to be a few transformational companies of every era. Whereas there may be companies that have not quite transformational point of views, but incremental ones that are quite successful, but conflict with the transformational ones. And so if you're looking at companies at a later stage, you only want to be in the data bricks of this era or the name your company that's been a kind of revolution snowflake of the three or four years ago era.

That's kind of what you're looking for. And there may be perfectly good approaches that are not quite that, but are very, very good investments for an early-stage investor. And those might generate conflicts because they compete with each other. There are different approaches. Team-wise it also is kind of interesting in that the way you look at a growth stage investment and the way you look at a very early-stage one are different.

And it can be hard to apply the same lens. A growth stage investment is more of a rigorous kind of due diligence process and set of financial analyses to come to a conclusion as to whether something is a good growth and a market analysis. Because in early stage, we've talked a lot about that. It may just be about a world-changing founder point of view, or there's lots of theses that may apply there.

And you really can't go from one to the other and back very easily. That creates that defocusing that I was speaking of. Again, I don't think it's insurmountable because I think you can have different teams doing different things in different ways and be very successful and some firms have done that. But for me and for Mark and Christina, we felt like to be the best at early-stage investing, it was going to be very hard to do anything else well because that's already a Herculean task to be a magnet for all of the best early-stage founders.

That was a hill worth climbing that if we're successful or partially successful on that journey is enough for us. We don't need to do anything else to achieve what we want to achieve with our careers and our aspiration. And that purity was kind of clarifying. And when you have an organization that's smaller and everyone kind of has the same goals, it's just there's more alignment. There's more alignment amongst each other and more alignment between us and our founders too because we're not thinking, once we make an investment, we're not thinking, well, let's make sure we make enough bets in an industry that later when it's clear who a winner is, we can make very, very big growth investments.

It's like a mentality that wouldn't work for us. We don't have the fund size to do that. Our mentality is always like, who are the founders we want to bet on for this era and go all in with them and then that's where we're placing our bet then and only then. And across the whole organization, there's no other point of view to wrestle with. And that's nice. That's freeing. And it's a simpler set of contingencies than you might have in an organization where there's just different incentives.

But how would you go from the problem idea, if you'd like, you raise the fund very fast as well. Can you maybe give me an idea of, you mentioned you knew the other partners. What was the discussion like maybe the early days and then how did you actualize this? What's the story to what we are today?

Yeah. So the three of us started working sessions together around the ideas that became Chemistry kind of four or five months before we actually decided to do it. And it was a casual kind of, oh, that would be interesting because the big firms have gotten bigger. There's this little bit of this gap in the market for early-stage focus. There are some firms that would say that's all we do, but there's not as many.

Even there, there were distractions. And so we got to this market realization, internal realization about ourselves, what we enjoy and what we wanted to do with our careers. And then we kind of translated that into a pitch to LPs. We haven't talked about the LP side of it, but we also felt, we didn't know, but it would be a compelling pitch to LPs to say to LPs, if you're going to invest in venture capital, you want a high alpha product.

You want a fund that could be a 10X, 50X potential or not work out. Because you're going to be illiquid. You're going to be investing in this fund for a long time. You're going to be illiquid on that. And so the potential return better be worth it. Now, if you're investing in a whole class of things that are billions and billions, it's not realistic that that's going to be an outlier return. The math doesn't work.

And so we felt like we would go to LPs and express to them that an early-stage purity is more in line with perhaps what some of them were seeking from venture capital. Not all. I mean, different institutions have different goals. And so we worked a little bit on our pitch to LPs, which started as its foundational premise. We're going to be early-stage forever. And we're going to be focused on that.

We're going to be, and by the way, we're experienced people. We've been doing this for a long time, 35 years of collective experience, dozens of unicorns that we backed before. We kind of knew how to do the job and we thought we could do it even better in this formation. And then we got to where we were just personally excited to do it. And we felt like the LP pitch had a chance. We didn't test it.

We tested it with one XLP who was now doing indirect investing because we really couldn't talk to the LP community until we had fully committed to it. But it was a little bit of a leap there that there'd be receptivity to it and there'd be enough interest that we could raise the fund size, our thinking in those days was 250 million, to be able to enable the strategy.

And then what's the thesis of the fund? What are you going to be investing in?

Yeah, thesis-wise, we're generalists with thematic areas of interest. But there's a big openness to new ideas as we've talked a lot about already in this discussion. There may be a theme that comes at us that, and I would invite any founder listening to this who doesn't have a conventional theme that is VC ready to try us, try me on it. Because just like that Twilio moment, I realized that the best venture capitalists don't necessarily dream of the best companies.

Companies that put rockets into space and build self-driving taxi services. Those come at you in unpredictable ways. Thematically, I'm always interested in what enables developers and technology class people to be more productive, as well as what enables business users to be more productive without the need for developers. So it's like, how do you make developers better? Then how do you make the rest of the people that are like knowledge workers have access to far more functionality than they can do on their own without being coders themselves?

I call that theme the democratization of development, bringing development tools to the rest of us. And companies like Cursor kind of do both, right? They allow people that aren't necessarily deep technologists themselves to be able to have lightweight tools and they make devs far more functional. That's the theme of interest to me. You're also interested in fintech application software, consumer applications.

How is this class of technology going to lead to revolutionary improvements in health, in life, and in business? And then we start there and then does the company kind of line up with one of those kind of hierarchy of needs? And that generally is a theme that's of interest to us.

We spoke a bit earlier about the AI native company. The way I understand this is that we see a lot of companies right now incrementally trying to improve their current offerings. And it seems that very few companies are thinking fundamentally what's AI enabling right now that we never had before and building these foundationally. What's your view on the new type of companies that are coming?

Yeah. I mean, it has partially because AI is so revolutionary in terms of its impact. You can build something incremental pretty easily, which is great. I mean, that's really, really good that you don't need to even have a company to build something. You can build an open-source project that enables a piece of functionality or a tool that has global reach and really makes things better for the world in some way.

OpenAI started as a nonprofit or maybe still is a nonprofit. But to build a fundamentally durable company and make a really good venture capital investment requires a lot more than that. You have to be able to apply your technology in a differentiated way to something that's not easily copyable or easy to do. If there's 10 ways to do something and yours isn't a lot better than the other nine, it's probably not a great, durable company.

There's some exceptions to that, but by and large, you really want to be somewhat monopolistic in, I don't mean that in the antitrust sense, but monopolistic in the functionality is so much better that why would somebody take any other approach than yours? I think that's hard to spot in the AI era because there's things that seem pretty revolutionary and are for a time, the ability to generate ad copy or something that really aren't because everyone has access to those tools.

And so I think we're going to see a lot of head fakes, by which I mean companies that just seem so good, but their advantage is so thin relative to those other nine ways that they're not really companies, they're just kind of like tools.

Do you have examples of these?

I mean, I don't want to talk negatively about any company because predicting who's going to fail is always like, it's not what we do as venture capitalists, we kind of predict who's going to be successful. But I would just leave it at companies that are pretty, where the tech stack is, the technology is really built on someone else's technology and all you have is a kind of UI on top of that, that I would predict is going to be hard to stay ahead of the pace of just fundamentally foundational AI research right now.

And we've seen some companies that have kind of flatlined and are having trouble kind of hitting their second act because of that. What I'm really attracted to are the companies where they just move so fast ahead of, that they've built something that can't easily be replaced by a third-party tool or a third-party or the foundational premise on which they're built on. And that's pretty hard right now because there's an arms race at the foundational level that there really wasn't with cloud computing where it was like Amazon Web Services was kind of the leader in that in the early days and was like sort of the only game in town initially.

Foundational AI models, there's a number of companies that are really competing really aggressively and building, productizing their foundational tech in kind of interesting ways. I think of like cloud code, which is like a productization element of cloud for developers that in some ways is going to be good enough for a lot of people. And OpenAI has operator and that's an area that there's been some really good startup activity in, but they're building, they're productizing their technology to be able to take actions and take agentic options on the internet and be able to do things for their users.

And that may compete with some tech that's built on top of their foundational models. So I think it's a hard era to be able to call the winners in because the technology at the research level is just moving so quickly.

On that, I think this pass two questions. One is what do you think is going to be a durable company considering that AI is eroding so much of the time advantage in most of the things. And then also was this, I think you mentioned going through a number of cycles before. Was this a case before too? Was there an era maybe that you could not foresee which company is going to be durable?

It feels different, but it's probably not. Like it feels different to me probably because I've experienced other eras, whereas before it was all new. And so to me, it feels very different. When the final tale is told, there's always companies that like jump out into the lead. Think pets.com and Webvan seemed like those are the ways the internet is going to be manifest itself in our daily lives. And they were right, just like not the right time.

And so I think in this era, the head fake is not that. The head fake is what you're enabling to happen is going to be enabled to happen, just not by you. It'll happen by tools from other, it'll happen because people will just build their own tooling or there'll be open-source projects that do it, or there'll be companies that do a lot of this stuff or whatever. So it feels a little bit disruptive, but there's always the companies that get out early in a market that don't fundamentally build enough foundational tech, don't understand their customers well enough and what their alternatives are that they stay ahead of where the market moves.

And that is just no different than the disruption story of every era. It goes a little bit back to why I think a lot of these companies will be built here, not all of them, because you need a certain fast-movingness and paranoia that you can't just invent one thing and say, okay, I invented one thing. Now let's go see...

You need to invent one thing and invent another thing and kind of just stay ahead of the other alternatives to what you're doing so that you maintain a kind of bond with your customers where they trust you and not something else. I look at what's going on in the data world and there's already this new data stack phenomenon where it seemed like the right way to build data teams at big companies was going to be to build a cloud data warehouse and have these connector companies that bring data in, transform it, and keep it there, and then be able to port it to other applications.

And that's happening, but there's already a disruption of that where you don't need to be as schema religious about how your data is imported anymore. AI enables you to dump a bunch of data in a lake and then figure out what you want to do with it later. That's a different stack that would enable a world like that. I predict something like that will exist and the former world is going to exist too, but there's already a little bit of headwinds there.

What companies are you really excited about these days and maybe, what are you looking for in the next five years?

It's more a style of leadership and a style of founder. I continue to be excited about companies that democratize access to technology for everyone. I used to say democratize access to tech in the enterprise, make salespeople more effective, product people more effective, DevOps people more effective. But now I think of it as the world because we're fundamentally disrupting the interfaces to tech because AI is more human-like.

I don't think that's a controversial statement, but it's going to get even more so. We could see profound effects that go beyond just business. I'm excited about anything that radically changes people's behaviors for the better across work and play and their daily lives.

Healthcare's an interesting one, your field, because healthcare wasn't an early or even current adopter of cloud in as big a way as some other industries, but it is adopting AI. It's almost skipping the transformative technology trend of the last decade, cloud computing, and moving directly into an AI native era where you have companies take notes automatically, transcribe things automatically, read radiology images automatically.

One investment we made is like a healthcare receptionist where it can talk to customers, answer their questions, schedule their appointments without a person necessarily being in the loop on most of those inquiries. That's acceptable now in a way that cloud computing was really tough for healthcare. So that's an area of interest of ours at Chemistry as well.

You mentioned the healthcare investment that you just did. Can we expand a bit on that?

Yeah, it's a company called Assort Health, a fast-moving team. One of the founders has a family healthcare background and at one point considered medicine, but instead wanted to solve this problem of bringing technology into healthcare. It uses multimodal AI models to provide the best receptionist operator experience for healthcare clinics and hospitals, where they can schedule your appointment, respond to basic queries, and things like that all in an automated way.

There's enough healthcare-specific know-how that if you were a clinician, you couldn't just plug yourself into the models and set up a phone operator for that. You need a pretty specific well-trained system to do that, that understands the way healthcare clinics operate. They're off to a good head start. That's our first investment in healthcare at Chemistry that I think has a bright future.

It's very interesting to see what's happening to the healthcare landscape and the hospitals. It sounds like we're going to have all of the scheduling being done with an AI. After you leave the hospital, you're going to have an AI telling you what to do and so on. There are so many things that you're basically removing from all the processes and operations of the hospital itself. Are there other things that you're seeing?

Yeah, it's interesting. It does feel like that's starting to happen and starting to happen pretty quickly. Healthcare skipped over the cloud computing era but is not skipping over the AI era. It's not going to be a revolution. It's happening incrementally in a way that maybe an AI is taking a first pass at reading x-rays, but a human's looking at them too and training those models and making sure they're not missing things.

But it feels like the accuracy rates are already there. Even vaccine development, with synthetic research, you can model out the outcomes even before you do the clinical studies, which is a big deal. I think it impacts every area of medicine. It's a little hard to predict, for me as a non-medical person, what are the chips that are going to fall first. Our investment in scheduling and being an operator, you're not touching a lot of healthcare data there.

But when you're looking at people's clinical records, you are. Let's take AI-assisted surgery: you're in somebody's body, so there's going to be a high bar for what technology gets deployed in those settings. But it does feel like that's going to happen. It might happen pretty quickly, just like you can take a car that's not driven by a human around now, and a couple of years ago, you would have thought that a little bit unbelievable.

Now, at least here in San Francisco and a few other cities, you stand outside, you'll see one drive by within a couple of minutes. Our level of societal acceptance for AI-guided processes that used to require a real expert human, like medicine, is going to just go down, or sorry, acceptance will go up, our hesitancy will go down.

We spoke at the beginning a bit about cults. What are the best teams that you've seen? How would you advise people in terms of company culture, team building, what are the best examples you've seen there?

I think the best examples are people that are very opinionated about it. They don't all have the same opinions, but they have an ethos to them, and they have an opinion not just on the way the world's going to work so they can build product for that world, but the way they are going to operate their system, the kinetic system of the company. I think about Owen at Intercom when I think about this question because he has a certain way he likes to operate, and it's not for everyone.

He would say there's a certain type of metabolism and product excellence that he's looking for, and a certain philosophical underpinning of that. He attracts people like that and probably repels people that are not like that, and that's healthy for an organization to know what its leaders value and the way they believe in operating the organization so that there can be alignment. I think about that even with our small team.

We have a team of nine at Chemistry. It's not a big team, but in every hire we make, we're pretty deliberate about ensuring that there's cultural alignment with what we're trying to achieve at Chemistry and the way we're trying to achieve it, or we would feel really hamstrung if we had people that were not on the same page. We've been pretty deliberate about that, somewhat slow to hire as a result.

The best leaders have a sort of followership to them that is hard to fake. You have to really believe in them, or you're not going to do the extraordinary things that are required to make a startup successful. You need to keep pace, especially in the AI area, where you can get disrupted so fast because you fall behind.

If you, as we mentioned, you're 17 years in the Bay?

Yeah, 18 years in the Bay, 17 years as a venture capitalist.

If a new founder comes to the Bay for the first time, how would you advise them to set up their company? Who do they speak to? What should they do when they arrive here?

That's a really good question. And to seed stage startup, they haven't raised money. I think there's not one right path there. But what I love about San Francisco specifically is you can meet a bunch of people really quickly. There's 40 meetups every night about some topic that's probably of interest to almost anyone working in some AI native way that is not going to be hard to hitch on to those.

So I would advise a little bit of that to meet people and find people that see the world similarly and think similarly. Then there's a class of pre-seed and seed investors now that are pretty easy to interface with, even for someone just moving here, where I would be trying to stage it, try to get in front of them with your ideas. But I stage it because if you're just moving here, there may be a little bit of acclimation where you try a few and learn what seems to land and what seems to resonate.

Then I think there's a lot of programs, like YC and other accelerators and incubators, that are designed for people that have good technical approaches but are not necessarily connected, coming in, that obviate the need for connections out of the gate. I think those are really good democratizing forces for startup development. You might not start a company but for access to Y Combinator, HF0, or Heavybit, where you're going to learn something and be able to just be given the right context that you need to be successful and help you with a lot of the administrative sidings.

For my last question, I usually ask everyone, you look backwards from 100 years into the future. How does the work look like?

Going 100 years or working backwards? Going out 100 years, that's so hard. Can we do 50?

Let's try 50. Let's try 50. Because I think you've seen the last 20 years of technology evolution, so it would be interesting.

I think we're right on pace with technology being more and more integrated into our bodies every day. A lot of people think of their phone as part of them. They're always with it. Now it's going to be able to recall everything, or pretty soon it's going to be able to recall everything you've ever done, every memory you've ever had, every emotion you've ever felt. You can then very soon be able to query that in the way that now you can put your notes into chat TPT and ask a question, hey, when did I meet this person?

So 50 years into the future, the idea of, we're just going to have perfect recall of everything that's ever happened, not just to us, but to any piece of knowledge that we've ever been given access to, a book, a learning of a language. You'll be able to use an AI assistant, and it won't even be called that because it'll be part of us, just like you don't really talk with the telephone anymore. You just call someone, and obviously you're using a device to do that, to be able to access all the pieces of knowledge that you might want.

That's part of the future. I think we'll have much more seamless infrastructure because we'll be able to predict much more effectively how we need to use our physical assets. I don't know exactly what the world looks like, but I think a lot of the anguish we have around making sure we have the right things in place in the right cities, housing, corporate office space, outdoor space, I think we'll have much more multimodal spaces, buildings, physical plant so that our cities will just be far more adaptable, which I think is a constraint on development now.

Virtual entertainment and virtual experiences is another one that your mind goes obviously to sort of like ready player one type stuff where you can live in a virtual world. I don't know whether that's the way the world goes. I have made some investments in VR, and I'm still on the fence about do we sort of live in a virtual environment at all times. I don't think that's what happens, but I do think we'll be able to have an augmented reality that's much more seamless with the physical world that can allow us to access spaces that we're not in if we go out 50, 100 years, where you could certainly be able to play soccer with someone that's not with you.

We'll figure that out. We'll figure out how to play games and converse and kind of be in a space together with people that aren't around. Then I think we're already seeing this with the proliferation of creative tools that are aided by AI. But the creativity that's going to be unleashed by giving everyone democratizing access to creativity is going to be world changing. I think they'll still be creative storytellers that are so much better, like Steven Spielberg and stuff like that.

They'll have to harness those tools. But just like we talked at the beginning about democratizing access to opinions through social media, we'll democratize access to creative pursuit so that people can build creative experiences, movies, film, books, art, that aren't necessarily vetted at the highest levels by society. There'll be some wild outliers that come out of the gates with the best painting or the best comic ever that just get access to those tools because they have access to the technology that's all around us.

It's one of the best answers I've ever heard. I'm wondering, where does it stem from? Is this from your information asymmetry by looking at a lot of great companies very early on? Or how do you draw that opinion?

Well, my father's a futurist. He thinks a lot into the future about this kind of stuff. So a lot of it is probably just programmed into me from a young age because he's been doing this technology prediction trends for 60 years or something since he was an MIT student. I like the non-consensus points of view. Part of that is stepping back from the day-to-day of investing and thinking about whether a particular idea that you're being pitched fits on a world arc.

I probably think more about the societal arc of where things are headed far more deeply sometimes than, OK, do the numbers of this investment kind of foot or not? In starting Chemistry, I've had even more time for that because just looking at early-stage ideas, just trying to seek out the heretical wild founders, once you get pitched a wild idea, even if it's wrong, it kind of stimulates your brain to think about what might be right or what might be right about it.

Or maybe it sounds wrong, but it's actually right. I try to make space and time for that, I think, is the fundamental answer. I just have the personality where I sort of assume my view is mostly wrong, but maybe a little bit right. If you assume that, and you're always trying to figure out what part is a little bit right, and if something's wrong, get me to the right answer. I just have that curiosity and hunger to always try to find the right answer.

I think you did Harvard and MIT, I believe, was it? College?

Oh, I went to Stanford. That was the first time I got out to the Bay Area. That was way before the era we've been talking about. Then I went to HBS, later for business school.

How durable do you think they are based on your experience with the universities and what you know from the new technologies that are coming?

Did you say how relevant?

How durable, how relevant.

Oh, how durable. That's a really good question. Despite what seems like a hard moment for universities, I think the elite universities are more durable than people suspect because there's a signaling aspect to education that no one's really been able to figure out a replacement for. The knowledge is definitely not that durable because you're going to be able to get all the knowledge that you learn in universities in other ways, quickly, if that's not already true.

But the question, I think that question comes down to when are we going to replace the signaling aspect of a Stanford degree? In startups, we've replaced it, right? If somebody pitched me and they said I was a dropout from high school, we had somebody at a Chemistry dinner who, British guy from London, who didn't quite finish secondary school, and I didn't ask him, but he's starting his company, and it's not like college was even in the equation at all.

I'm just as interested, if not more, in what he's going to do than someone that graduated from Stanford. So I think in the startup world, we've removed the stigma of not having a college degree, but in the rest of society, that may take a long time if it ever happens. So I think there's going to be much more of an emphasis on what practical skills do you get that you can't get in other ways. But the signal might stay relevant for a long, long time.

I have kids that are not even quite in high school, but I suspect that their desire to go to a good university is probably similar to how it is today, even five, 10 years into the future. I kind of hope I'm wrong about that, actually, because I think technology can be a democratizing force in access to who has the credibility, the signal that's bestowed upon you by graduating from a lead institution, but I'm not really that hopeful.

Super. How can founders find you?

Ping me on Twitter, X, cold email, find one of our colleagues around. I'm not that hard to find. I'm out there. I think my DMs in X are open, so easy to ping me.

Super. Ethan, it's been an awesome conversation.

Kyriakos, thank you. Thanks for inviting me on. This was very fun.

Super, awesome.