How AI Is Rewriting Seed Stage Investing with Kevin Hartz & Bennett Siegel | Ep. 49
The state of venture capital firms in the current AI cycle, what it means for seed specialists, trends with great founders, and what they're seeing in AI.
Kevin Hartz and Bennett Siegel are co-founders and GPs at A*, a five year old early-stage venture capital firm with $1B in AUM. A* has invested in companies like Notion, Cape, Whop, Paraform, Simile, Krea, Mercor, Watney Robotics, Andera and others.
Kevin is also the co-founder of Eventbrite (NYSE: EB) and co-founder and board member of Xoom, an online money transfer service that IPO’d in 2013 and later acquired by PayPal for $1.1B. Notable investments, primarily at the seed/early stages, include PayPal, Airbnb, Pinterest, Reddit, Anduril, and Palantir among others. Bennett was previously a partner at Coatue building out their venture capital business where he invested in earliest financing rounds for Ramp and Decagon, among other investments.
We discussed how AI is reshaping venture capital, software, and startup building – from the rise of younger founders and AI researcher-led companies to the growing pressure on traditional software businesses. We also covered the changing economics of seed investing, the influx of mega funds into early-stage venture, AI rollups, robotics, and why this may become the biggest technology boom yet.
Timestamps:
(0:00) Intro
(0:25) The A* Capital story
(1:16) Why big funds went into seed
(7:50) The mother of all bubbles
(10:46) Why founders are getting younger
(13:00) Mapping talent, not markets
(16:31) The rise of AI researcher founders
(19:16) Why seed investing is so hard
(22:54) Concentration and venture returns
(27:34) The AI rollup craze
(31:15) AI vs traditional software
(33:15) Robotics and the future of AI
(35:39) What’s next for A* Capital
Links:
https://x.com/kevinhartz
https://x.com/BennettSiegel
https://x.com/jaltma
Watch on YouTube; Listen on Apple Podcasts or Spotify
Social clips
A new crop of founders: researchers
One of the big changes in venture that Bennett and Kevin discussed is needing to understand a new type of founder -- researchers, which historically didn't always make for great founders but are obviously a huge part of the AI era.
"You have to listen so carefully...every word from a great founder will have so much meaning and intention about what they're gonna build."
Transcript
Disclaimer: Transcript generated with AI assistance and lightly edited for clarity and accuracy.
Inside A*
Jack Altman
I’m really excited to be here with Kevin and Bennett from A*. Thanks for doing this with me.
Bennett Siegel
Thanks for having us.
Kevin Hartz
Thank you.
Jack Altman
I want to start with just the quick history on A*, how the firm came to be, how you guys started working together. Then I want to go into seed investing. Can you give us the quick backstory on A*?
Kevin Hartz
We first met through Ramp, and that was at the seed stage. Eric and the team were raising and I participated as an angel. This is pre-A*. Bennett was at Coatue. That’s where we first interacted.
Bennett Siegel
We started A* five years ago now with our third partner, Gautam. The idea was to build a new kind of venture capital firm that could focus on partnership with founders at the earliest stages and work closely with them as they continue to grow and mature. Five years in now, we just closed our third fund. We’re at a billion of AUM, and the idea is to bring something different to the early stage market.
Venture Arrogance and Big Funds at Seed
Jack Altman
A year ago, I started this podcast, and one of my first guests was Josh Kopelman, founder of First Round, an amazing seed stage firm. One of the things he was talking about was the venture arrogance score, that really stuck. What he was saying was, bigger funds are super arrogant to think that they can have big returns, directionally speaking, because you own X percent of a company. How big can the companies be? What return are you getting?
That argument really resonated. It’s funny now to me, looking back 10, 11 months later. A lot has happened. There are more of these five to $15 billion funds than ever. The rounds are more expensive than ever. There are these three-plus huge looming IPOs, which kind of gives credence to the big fund. Some of these big funds are doing really well actually.
So there’s this whole situation. I’d be curious to hear your guys’ take on the landscape and being a smallish seed-focused firm relative to this market moment.
Kevin Hartz
Look, you always have to play the hand you’re dealt, and we live in this world of giants. It’s true about Anthropic, SpaceX, and OpenAI and these massive returning large-capital-needing companies. But at the same time, big funds, big fees for those partnerships. We stayed small so we can be agile. It’s like the equivalent of when you’re the CEO of a new startup and you don’t take a salary.
Jack Altman
Do you think if the fees were capped—instead of 2 and 20, if after like a billion there were no more fees—VCs would behave way differently?
Kevin Hartz
Absolutely. Without a doubt. I’ve seen funds go from small to large and the behavior absolutely changes. No names, no name drops here, but it happens that way. There’s so much in this really exciting upswing of the market. LPs want to get in, and they’re not able to negotiate those fees down.
Bennett Siegel
If you think about it, the fees haven’t changed as the asset class has changed. Look at long-only oriented funds in the public markets. They charge 1%, they charge 10% carry over a hurdle. At this point, these growth funds that are raising $5-9 billion, they’re investing in what would’ve been public companies a decade ago, and they’re essentially indexing the asset class. Why should you make 2 and 20 for that?
We built A* to be aligned with founders. We want to partner early, we want to take risks, we want to work hand in hand with them. It’s a different craft, but our incentives are aligned. We only win as they win. That’s different than I think the prevailing view in the bigger firms.
Jack Altman
You do need to do well enough as a big firm. You need to cross some threshold so you can keep raising your funds. But you think about a $5B or $10 billion fund, and you think about them playing at seed. I was joking with you before we started that your guys’ fund is like $450M or whatever. Some of these funds, if they raised $8.3B versus $8.8B, none of us would clock it. But that’s a whole extra A*.
They can use that going into seed rounds, and their incentives there are different. So I’m curious. What is their strategy? What is a $10 billion fund doing at seed, and why are they doing it?
Kevin Hartz
I like any unit of measurement that is an A* unit of measurement. Maybe it’d be adopted as a practice.
Jack Altman
We’ll all talk about A*s going forward.
Bennett Siegel
There are two facets. We used to joke at Coatue that there were two ways to make a billion dollars. One way was the Benchmark way. You find an incredible Series A, you lead it, you partner, you have a generational public outcome, and you earn your percentage of the company. That’s hard and takes a long time.
Kevin Hartz
Or the A* way, in addition to Benchmark.
Bennett Siegel
In addition to Benchmark. Or you put a billion dollars into a company and you double your money as it goes public in three years. That also earns you a billion dollars to distribute carry. One is a lot easier than the other and more repeatable?
So incentives have shifted in a way that reflects the market reality. Seed specifically—we see this with a lot of our friends at multi-stage firms—it’s option value. Let me build a basket of companies with a certain amount of ownership and see what pops. If it pops, I could double and triple and quadruple down. I don’t think founders want to be thought of as an option, as part of a basket.
Jack Altman
I obviously want to believe that too. But can I give you the counter there? I’m curious how you manage it in the market? The counter is basically, look, I also don’t want to be thought of as an option, but $10M at a $100M sounds a lot better than $3M at $30M. I have $7 million extra dollars to go hire a team, run my experiments, see where I’m at. That is the discrepancy that’s been created. So how do you navigate that as a seed specialist?
Kevin Hartz
We are a boutique firm, so we can’t be everywhere and we can’t cover every scenario. We find enough founders that are looking for a great partner to work with and to go through this battle. We’ve been through it before. I’ve been involved in founding a couple companies from inception all the way through IPO and beyond.
We can’t win them all. I would also say, the challenge of this era—and maybe I’m biased by this—is that the amount of capital out in the market does mean that you don’t have as much help in oversight. I’ve seen more rookie mistakes being made in this environment that just shouldn’t have happened.
Jack Altman
It’s funny, there’s a class of things that is very hard to sell in a fundraising process, in the quadrant of true/false, easy to sell/hard to sell. There is this set of things like what you’re talking about—help during the hard times—which I think is very hard. It’s real, but it’s very hard to communicate in a round process in a market moment like this. When markets are hard, maybe that resonates more. But when markets are really frothy, as it is in many ways right now, it’s just tough. It’s an interesting time as a seed specialist.
Bennett Siegel
We’ve seen this before. In 2021, that was the era where Tiger was offering every software company the lowest dilution, highest valuation offer. Solo GPs were promising very little help, but they were hitting the highest price bids at Series A. Most of those companies and most of those outcomes did not work.
Jack Altman
The pitch was, “We’ll be uninvolved.”
Bennett Siegel
Yes.
Jack Altman
The sell was, “We won’t talk to you.”
Bennett Siegel
A hundred percent. Building a company is a marathon, not a sprint. A low-dilution lead offer is relevant for founders. When we compete and when we win on deals, we’re not trying to get a value deal. We are also offering what we think is a low-dilution competitive offer, and support and partnership and things that founders should care about.
The Mother of All Bubbles
Jack Altman
I know you start at seed, but you do seed, A, and B. So you’re very in tune with all of those market stages. What have you seen over the last six or 12 months in terms of how these rounds are happening in succession, where valuations are going, how dilution is happening across them? What are the trends along those rounds?
Kevin Hartz
I’d like to think that we’re not top ticking, we’re not there yet, but we’re headed to the mother of all bubbles. A historic precedent. If you think of the eighties and the personal computer and what came out of that era, or the nineties and the internet, and the two thousands and mobile… AI and the platform and the boom that we’re seeing right now is immensely exciting. We have such a global audience to reach. I think we’re going to have a few years of this continuing to grow.
But unfortunately, as in capitalism, it always ends badly. I don’t mean end. We all know that after 2000, Amazon and Google and so on came out of the internet bubble. Now they’re multi-trillion dollar companies. Netflix and PayPal and so on. It’s an exciting time, but yes, capitalism is like a pendulum. It’s either too far one direction or too far the other.
Jack Altman
It’s funny on the bubble topic. I have a friend at a growth stage firm, and he and I keep track of all of the signs that we might be in a bubble. It’s things like how quickly the rounds are coming in rapid succession, what are the degenerate gambling behaviors we’re seeing from people in and out of venture, what’s happening with SPVs. You go through this list, and a lot of the signs are there.
The counter, which is a strong counter, is that there are a lot of extremely interesting companies. There are a lot of companies getting to $100 million of revenue with happy customers and crazy new technology, faster than ever. I’m guessing that you’re probably feeling a mix of, “Whoa, this looks like a bubble, but these companies look so interesting.”
Bennett Siegel
We’re optimists by nature. That’s why we’re venture capitalists. In any platform shift, you’re going to have some incredible companies that come out of this vintage. In fact, they’re probably going to be larger than anything we’ve ever seen. But you’re also going to have a lot of companies that aren’t going to go the distance. We all see it in our portfolios. Who’s building durable revenue streams? Who’s building moats around their business versus what the labs are offering?
If we’re doing our jobs correctly, we’re going to find these handful of companies that matter. But there’s going to be carnage in their wake, and not every company is going to make it. That’s true in every prior cycle.
The other thing I would say is, what’s changed since we started A*? Seeds used to be $20M to $30M posts. They’re now $40M to $50M. A great Series A used to be $100M post. Now they’re happening at $250M. Series Bs used to happen with real traction at a few-hundred-million-dollars valuation. They’re $500 million, $1 billion. Everything has changed, particularly as large funds have gone earlier. This will end absolutely fine for the best companies that will continue to grow and mature, but it will also be more challenging for founders that struggle to raise above their pref stack and don’t go the distance.
Founder Attributes in the AI Era
Jack Altman
So you guys are investing at the earliest stages where you’re basically just looking at teams. There’s an idea, but probably not much of a product, and the idea is probably squishy. You’ve got this backdrop where this is both the most exciting time ever and a scary time in certain ways as an investor. So what are the founder attributes that you look for right now? If there’s any contrast to what you looked at in the pre-AI cycle, I’d be interested in hearing that too.
Kevin Hartz
There’s a curious movement around age, in that founders are getting younger. I guess they’re not aging in reverse, but—
Jack Altman
No, it’s wild.
Kevin Hartz
They’re getting into—
Jack Altman
Just look at a YC batch. People in their teens.
Kevin Hartz
I bet they have great data on that.
Jack Altman
Yeah, it’s going young, for sure.
Kevin Hartz
It’s interesting. In the late seventies when Steve Jobs founded Apple and Bill Gates founded Microsoft, they were 19. Then you fast forward and there’s some other examples along the way. Then you had Zuck, and then you really had this accelerate. You had Peter Thiel and Luke Nosek devise Thiel Fellows in 2010. That was very prescient.
Jack Altman
It’s been incredibly successful.
Kevin Hartz
Yes, and it’s a nonprofit.
Jack Altman
It was unbelievable.
Kevin Hartz
It was very prescient of what was coming, which is this lowering of the age and dropping out.
Jack Altman
But it wasn’t quite like this right before AI.
Bennett Siegel
It was never like this before. There always were young founders building incredible businesses. But you and I have spoken about this. In this paradigm shift, who cares if you know how to build a SaaS product, or sell a SaaS product, or hire a traditional enterprise go-to-market team? This is new for everyone. We’re all rewriting the rules as we build these companies in real time. Young founders are situated very well because they’re the first adopters of this technology.
So it’s why you’re still seeing very large companies built by people under the age of 25 or 30 at an unprecedented rate. For us in practice, that means more and more of our founders are skewing younger and younger.
Kevin Hartz
Just the other day I was speaking to a founder going over some documents, and he said he would run things by his lawyer. I said, “Isn’t your lawyer your mother?” He’s a teenage founder. And he sheepishly said, “Yes, she’s a good lawyer by the way.”
Jack Altman
That’s good. What else besides age? We were talking about durability and the world is changing. What other attributes, aside from just young founders, have certain advantages now?
Mapping Talent
Bennett Siegel
We generally map talent, not markets. To be a seed-stage investor, you have to be founder-centric, but you also have to recognize what’s happening around you. The labs are not just infrastructure companies now. They are legitimate competitors at the application layer, which happens to be where most seed-stage companies are building. That’s where you raise an initial round and launch a business. Obviously there’s codegen, which is the central battleground, but it’s going to proliferate to every part of knowledge work and every part of the enterprise.
We do try to think about where there’s white space to build, or where you can have enough of a runway to finally build differentiation. Nobody is differentiated at seed. This whole idea of, what makes this different? Why couldn’t your competitor do this? Every seed-stage company could obviously be disrupted or built by an incumbent. We need to figure out where there’s at least enough white space, enough of a runway, that you can start to achieve liftoff.
Jack Altman
When you say you map talent, what does that look like in practice? What does a day look like? Is it just reaching out to interesting people in certain networks? Are you tracking people that you’ve already known? What does it actually tactically mean to map talent?
Kevin Hartz
It’s a lot like mining, where you’re in a vein of gold, you’ve found an area, but that exhausts itself and you’ve got to find new talent and new areas. I don’t know if that’s a good analogy, mining, but things are constantly switching. The founders are always the same. They’re always those that have had to overcome some kind of obstacles and still achieve at a very young age, and be able to recognize the signs.
We’ve gone through this big phase of IOI and all the coding competitions and things of that nature as a measurement for teens that are achieving on that side. So you’re looking for these types of attributes, but the reality is—
Jack Altman
Jane Street, I feel like, has been a really good one recently. There’s been more interest from ex-Jane Street people to be in startups. Obviously it’s a very talented group.
Bennett Siegel
This isn’t some unique insight, but high-quality people want to hang out with high-quality people. So when you go to the top universities, when you go to the accelerator programs, when you go to companies— the most talent-dense nodes work very closely together. As you start to work with founders in those nodes, you naturally meet their friends who go on to start great companies.
I also think, as you see people leave companies, you see certain patterns. Certain companies breed better founders than others. Airbnb, Stripe—so many of these companies are magical businesses. They worked, and not to say there wasn’t hardship in the early days, but when they took off, they really took off. So it doesn’t necessarily breed as many founders that run through walls to start companies.
Palantir is an example. Everyone’s a mini-CEO. Everyone had to build and launch a product and find product-market fit. It’s why I think Palantir has the highest per capita rate of unicorn founders of any company. Part of what we need to do is continue to find those areas where the next founders come out of.
Kevin Hartz
We love those Palantir founders.
Jack Altman
They’re fantastic.
Kevin Hartz
You can’t get enough. They’re so good. Whatever they feed them over there is really working.
Jack Altman
I also think probably with Palantir there’s a certain non-consensus mindset implied simply by wanting to work at Palantir. At least, that was true at a certain time. Now it’s maybe become a bit more mainstream. But I think for people who worked at Palantir and Anduril at a time before it was popular, on the contrary it was something that you had to defend. That takes a certain mindset that is related to being a founder.
Researcher-Founders
Bennett Siegel
One thing we didn’t talk about: there’s a whole new crop of founders that never existed, at least in my time of doing venture. It’s the researchers. It’s folks that are leaving labs or PhD programs that are raising massive quantums of capital.
Jack Altman
By the way, it used to be a big anti-pattern. What, like, researchers?
Kevin Hartz
Yeah. That’s hard. You’re totally right. Researchers tend to have high intellect, but—
Jack Altman
Academic.
Kevin Hartz
Not as commercial.
Jack Altman
But now obviously there’s all these counterexamples for maybe the first time. I’m sure there were examples in the past, but that’s a whole new thing that people have to grapple with. It’s very hard.
Kevin Hartz
It is. I think that’s where, when you’re spending time with the founder and getting to know them, if they can really articulate what they’re doing, you have to listen so carefully. A great founder will have every word very much have great meaning and intention about what they’re going to build.
Jack Altman
You’re saying there can be cues with the language someone’s using, of where their head’s at.
Kevin Hartz
Yes and how they prepare and how they’re thinking about the business. Sequoia had backed my previous two companies, and they were always sticklers on how you presented and how you showed up at the same time.
Meeting Founders: Cold vs. Prior Relationship
Jack Altman
Do you guys find that meeting founders in a round versus getting to know them before their fundraising is some huge difference? Do you have good examples of success with either? Do you prefer one over the other?
Bennett Siegel
The Decagon team’s a great example. Kevin, why don’t you talk about the history with Ashwin? But I think the prior history informed how we thought about this company.
Kevin Hartz
I had been an angel in a company that Ashwin had founded called Helia. Russell Kaplan and Daniel Barrios were in that too. Russell’s now the president of Cognition, and Daniel’s over at Meta. That company didn’t make it, but I stayed close and was always a big fan of that trio. When Ashwin went off and met up with Jesse to start Decagon, it was kind of a no-brainer to get involved in the seed round there.
Bennett Siegel
The funny thing about Decagon, there wasn’t really an idea. They had this insight.
Jack Altman
You knew the team was great.
Bennett Siegel
We knew the team was great. We knew there was white space in building on top of the foundational models in a moment in time when no one had really commercialized different enterprise applications. You had the history of the team to say, “Wait a second, we understand why their prior companies were a modest success, but why one plus one equaled more than two in this situation.” That’s an easier bet to take in some ways when you have that history.
The flip side is, there are founders we can meet during a process where we try and build as much rapport and relationship as we can. We try and understand them as people. We have to try and understand the opportunity, and those have worked for us well. But it’s a shotgun marriage. You need to move faster many times with less information.
Jack Altman
Especially in a market like right now. Sometimes you meet them, you have three days or something like that, or less.
Bennett Siegel
Three days. I would say at A*, more than half of our seeds have been pseudo-proprietary in nature. That doesn’t mean no other funds were around it. That doesn’t mean they’re not meeting other folks.
Jack Altman
It means you didn’t start cold during the process.
Bennett Siegel
Exactly. We maybe were first to reach out to them. We had some informed knowledge. We had a dinner with them in the past. There was some relationship of relevance and some insight we had going in around who they were as a founder, what they wanted to do.
Jack Altman
Is it obvious to you off the cuff which half of your investments are better?
Bennett Siegel
Yes. The half that we had some prior relationship with, that were somewhat proprietary, have disproportionately generated returns. Though we have done well with some of these shotgun marriages, I would say there’s higher volatility.
Jack Altman
One of the things I think is interesting in seed is that there aren’t that many firms that stick around seed for a long time and are very successful. I gave the example of First Round. You guys are obviously doing great. There’s a couple others that have been around for long periods of time.
But a lot of times people either leave seed, or they stop doing what they’re doing entirely, or something else happens. Why is it that it’s a less persistent part of the market than multi-stage, in your guys’ view?
Why Seed Firms Don’t Last
Kevin Hartz
Seed is just incredibly hard. It’s so challenging. You’re investing in people, no product, hardly even a roadmap or anything else. You have a great deal of uncertainty day in and day out. Then compound that with the current environment where the multi-stage firms have come in and want to get a toehold in seed, so they spray money in.
Jack Altman
When you say seed is hard, do you mean hard in the sense of high effort, or hard in the sense of difficult to get a good result?
Bennett Siegel
Both. You work incredibly hard to build relationships with founders and earn the right to invest $2-5 million in a company that by all odds will likely not work. Most companies do not become billion-dollar-plus companies, and if it’s not a billion-dollar-plus company it has almost no relevance to the performance of your fund.
It is genuinely hard to find these founders. It is genuinely a lot of work to convince them to partner with you, and it’s even more work afterwards to support them in building a company. So you have two facets of this. Historically, seed firms that have been successful maybe do not continue to adapt to the environment. They don’t refresh their networks, they don’t continue to be aggressive as new types of founders and new technologies emerge, and they tend to fall by the wayside. That’s one crop of seed funds.
Others that have been successful, that have found one or two or three amazing companies, they grow up, they graduate, they raise opportunity funds and growth funds. What’s left is a relatively small group of firms that still specialize at seed and look to work with the best founders, and don’t try to find value deals or diamonds in the rough, but compete to be the partner for the next generation of winners. If you can’t do that, seed is a bad asset class.
Kevin Hartz
Just to emphasize, venture capitalists should be much better managers of people, too. The generational change, handing things off, very few firms make it on from there. They usually die with their original GP.
Jack Altman
What do you think is the big source of difficulty in generational transition? Is it finding talent? Is it older partners letting go? Where do you think most people get it wrong?
Kevin Hartz
My perception has always been finding talent, but I’m sure maybe older GPs are holding on a bit long. You’ve got to find that needle in the haystack out there, somebody who’s going to be able to identify the next Google.
Concentration and Doubling Down
Jack Altman
It’s also interesting because as funds grow, you guys could very easily make a lot of investments in your seed companies that are performing well and start putting huge amounts of capital behind them. One day you could wake up and your growth fund is bigger than your seed fund, and now the incentives are to buy as many of your dollars as possible. I think that sort of explains where the focus goes.
Kevin Hartz
Are you inviting us to do Series A’s now?
Jack Altman
A hundred percent. A, B, C, whatever. Do it all.
Kevin Hartz
Do it all.
Bennett Siegel
We’re a little different from a traditional seed fund. We do have a reserve-heavy model. We do have more capital available for follow-on than seed. If all you’re writing is the first check in seed, it’s exceptionally hard to drive returns. For our best companies, we’ve piled in every single round, and out of a portfolio of 40 seed investments, we might have 3, 4, 5 companies that have the lion’s share of our capital. It’s concentration, but different than the way that a traditional growth fund would talk about it.
Jack Altman
You’re saying seeds hit so infrequently that if you don’t double down on the few that work, it’s very hard to get good returns as a basket.
Bennett Siegel
I think that’s absolutely true, and I don’t think your fund is going to return the multiple. We look at others of our peers who look at pro rata and say, “What’s peanut butter pro rata everywhere?” You always want to support your founders, but as you look at the Series A, the B, the C, the D, those follow-on decisions are actually in many cases more important than your initial investments.
Jack Altman
What’s interesting is I would argue that from a basket of seeds, it’s actually very difficult to tell by the Series A where you should be concentrating. But a lot of funds just do their pro rata in the next round, and that’s that. That really is just peanut butter across all of it. I would think it’s much easier to know where to put huge amounts of money a stage or two even later than that.
Kevin Hartz
Agree. In our case, we do agree to support our companies in the A. We’ll always do the pro rata in that round and then watch. But you’re right, like that inflection…
Jack Altman
There’s a different pool of concentration. Let’s say you had a fund that was a $100, and $40 of it’s going into initial seeds, and then there’s $20 for those pro ratas. I would think for the last chunk, done right, you’re probably doing 10% of the fund into a few companies each, that kind of thing. Is it that concentrated?
Kevin Hartz
This is interesting because Brian Singerman is creating this LP fund to find new managers to be able to concentrate and help them go all in. Since 2005, that’s what Founders Fund kind of pioneered, breaking the old venture rules and putting this inordinate amount of money in. Like when they sold all their Spotify off at a $9 billion valuation and rolled it into Airbnb, that was pretty monumental.
Jack Altman
Which is interesting because that’s more of a trader mindset, which you don’t normally see in venture. Spotify did well, but I guess that was a good trade because Airbnb was at what, $2B or something like that?
Kevin Hartz
Yes. $2.5B.
Jack Altman
So instead of getting a 5x, you got a 50x. Great.
Bennett Siegel
You need to know what great looks like to be successful. For us, we have a $450 million fund. More than 50% of the dollars come after seed. Most growth investors look at seed companies and they’re completely uninteresting: it’s a founder, it’s a product, it’s raw. Most seed investors look at growth companies and they say, “These are all amazing. Oh my God, they have product-market fit and revenue and scale, and the founder can actually string together two sentences.” That’s historically why people create separate early and growth stage teams.
To be really great as an early investor and to follow your founders as they mature, you need to know what great looks like at every level. To your point, the Series A may be a hard stage to do that. Obviously you as a practice now are picking that stage, so it should get easier over time. But as things become more consensus, they get more expensive. It’s harder to buy ownership. So it’s always that pendulum on where you want to be on the risk curve.
Kevin Hartz
The other side is, when you do back the truck up and really go for it, ensuring that it is the money round. In ’06 or ’07, we used to sit on the floor and project SpaceX launches. They’d launch them in the South Pacific somewhere and they would blow up every time. That was the sucker’s bet, to actually invest in those rounds as these things were blowing up. But as soon as they got one into orbit and that landed the contracts, that put forth the sequence of events. Then that’s the next phase of putting money in. And certainly Starlink was the big unlock more recently.
Rollups and the Venture Math
Jack Altman
Founders Fund’s SpaceX concentration is extremely impressive, to do it over and over into that company. I guess the key is, you’ve got to have a SpaceX in the portfolio that’s worth concentrating into for 15 years, but it’s still pretty impressive to do it. A lot of people just wouldn’t have the stomach for it.
I’m curious about some of your guys’ opinions on AI. Obviously I know you spend most of your time trafficking in founders and talent, but I also know you guys have strong opinions about the market and where it’s going. I’d be curious to hear some of those.
Bennett, maybe starting with you. We had talked a little bit about rollups, buying businesses that are older, established companies and applying AI to them. I know you’ve got a background doing growth as well, so you have some more familiarity with this. What’s your view here?
Bennett Siegel
There are a few things that are going to be worse for incumbent businesses and the venture capitalists that are participating in these, than the work of buying and transforming and dealing with the culture and the process changes. Everyone thinks it’s as easy as buying a business, deploying AI, improving free cash flow margins, and earning a return.
Private equity firms have been doing rollups forever. It’s incredibly challenging to turn around an existing business. I think this is actually a great business for founders. It used to be that you had to raise capital, you had to buy an asset, you had to improve the asset, and you would keep 20% of the profits. Now a founder can go and raise venture capital dollars, dilute 20%, so keep 80% of the profits and go buy a business. It is phenomenal for them. I think it’s an exceptionally tough asset class that allows venture capitalists to raise lots of money, but I think it’ll be very hard to achieve real profits doing it.
Jack Altman
By the way, on that point, as the venture capitalist, don’t you need the thing to appreciate a lot just to get back to even?
Bennett Siegel
Yes.
Jack Altman
If you gave somebody $1 billion and then they went and bought an asset for $1 billion and you only own 20% of that thing, don’t you now have $200 million?
Bennett Siegel
The ironic element of this is that the founder can’t lose. The founder has literally bought into a business with embedded asset value that they now own a large percentage of. But unless these companies appreciate dramatically in value, the venture capitalist owns a relatively small percentage. Now people are doing convoluted structures where there’s a holdco and an opco and there’s carry and there’s waterfalls. But no one has actually proven a way that you can monetize this in any way that resembles venture-like returns.
Kevin Hartz
The other dimension of it, in this venture capital/private equity play, is simply that venture capitalists are very good at watching the top line. They want things to blast off, and they’re not so precise. This is all about the top line, but the bottom line more importantly. Venture capitalists don’t pay as much attention to that with their companies. So it just doesn’t feel like the right culture fit for this sort of stuff.
Jack Altman
Are there any forms of rollups or structures that you are bullish on? When you look around the different types that are happening, is there any form where you’re like, all right, if I had to buy one, that’s the one I’d buy?
Bennett Siegel
I think there is something to be said for looking at services-oriented professions that have largely recurring, if not truly recurring, revenue sources, and looking at how you could replace people with technology. Because if AI has been very good at one thing, it’s looking at traditional pockets of labor spend and automating it.
So there’s a reason why accounting rollups are interesting for people. There’s a reason why HOA rollups are interesting for people. I think ITSM as well, for that reason. It all deals with tickets and deflection and replacing call center people. That feels like a very interesting area, but I would argue 90% of the value created is going to be buying the asset for the right multiple, which once again, venture capitalists are historically very bad at.
Jack Altman
So you’re not going to be doing any rollups?
Bennett Siegel
We will not be doing any rollups today at A*.
Kevin Hartz
On the other side, there are always great teams, exceptional teams, in using technology. Bending Spoons comes to mind, as we just sold Eventbrite to Bending Spoons. That’s an incredible operation of how they integrate all these assets, and what they’re doing in Milan is remarkable.
AI Application Layer vs. the Labs
Jack Altman
What about within AI? What are you guys most excited about? Where are you anxious? Maybe frame up a common meme right now. Software’s extremely difficult because the labs are just expanding, and the workflows are being replaced by agents, and the workflows I need that are important, the labs are going to own.
Kevin Hartz
We can’t take that position. We want to back these independent companies. The labs need to have a very audacious appetite and keep the growth up. But we’ll still keep backing the application layer all around, where we see the right barriers, the right teams, and so on.
Bennett Siegel
There’s a distinction, obviously, between application-layer AI companies and more traditional incumbent software. I don’t know the last time we’ve heard a founder pitching a traditional software business. Now, part of that is the meme that everything’s an AI company. Everyone puts AI in their deck. I think systems of record are still very sticky for a reason. The issue, we all know, is if you can vibe-code an app in a matter of minutes, why do you need so many engineers? What’s the value of the workflows that you’ve built?
But what we’re seeing on the application layer is that people are going after systems of intelligence, going after systems of action. They’re new types of spend, and we’re going to continue to back those. In terms of the public markets, there is a reason software’s having significant drawdowns, but the best companies should be able to navigate this. Kevin mentioned early on that we were investors in Ramp. I would say they are a great example of an incumbent pre-ChatGPT company that has now become an AI business, and agents are driving top line and margins.
The best teams are going to adapt. So I think you’re fine if you’re in an incumbent software business, as long as you have a team that is going to rearchitect everything from the ground up.
Hardware, Robotics, and What’s Out of Reach
Jack Altman
What about hardware and hard tech? I know you’ve been incubating a company that is software-enabled hardware, I think. I assume you guys have been investing in it. How do you guys think about non-software, potentially even non-AI, but at least hard tech hardware companies?
Kevin Hartz
Again, we want to go where the talent is, and we’re being drawn to these various hardware companies because there are really impressive people. It’s happening across the board. There’s a revolution there. We certainly don’t want to get caught in the same traps that have been difficult for hardware in the past, but hardware puts you out in the real world. Before, in software, we were trapped like a brain inside of a jar. Hardware and robotics and so on is an exciting new world for us.
Jack Altman
There is some good defense if you have some sensors and things out in the world, at least. Someone can’t just come along and vibe-code that.
Bennett Siegel
I think that’s exactly true. We just haven’t hit the ChatGPT moment for robotics. AI for the physical world should be as large, if not larger, than what we’ve seen for the knowledge world. The challenge is, right now, we haven’t seen much in the way of commercial applications.
We don’t care, we’re seed investors. We’re backing robotics companies, we’re backing companies working on sensors and edge AI. But we do think about the number of robotics companies that have raised at huge valuations with very little commercial promise. So we think a lot of this is still to come.
Jack Altman
Are there types of companies that you are excited about but, given your fund size, you just can’t engage in? I’m thinking of neolabs, or heavy robotics projects, or things like that that don’t start with a $5-10 million seed round.
Kevin Hartz
Alas, those are tough to see and have them go by us. But with that said, we’ve been able to invest in a lot of scrappy, verticalized robots. We’re in a company, Watney Robotics. It’s data center cabling and robotics. That is a very specific task that’s very high value, and it differs from going after a full horizontal market.
Bennett Siegel
Or we backed a research team out of Stanford called Simile, which is focused on market research, which went on to raise a hundred million after the seed round. At the time they only needed a little bit of money to prove out and commercialize their research. Those are the rare exceptions as you know. The normal case is they skip that little seed round, despite how hard we try to convince them of the value of it, and they raise $10M, $50M, $100M dollars. By the way, that is where a lot of the multi-stage dollars are going at seed. We can’t really do those rounds.
What’s Next for A*
Jack Altman
Makes sense. So you guys have this new fund now. Any changes coming for A* overall, or is it just more of the same. Bigger, better, faster?
Bennett Siegel
I think it’s more of the same. As you know, there’s inflation in round size. There’s a larger volume of founders flocking to start companies than ever. I think we just need to be playing the game on the field. We need to be meeting founders where they are, moving quickly, and leading seed rounds, and hopefully we can be a bigger and better partner over time as our fund size has grown modestly.
We started with a $300 million fund five years ago. Our new fund is not that much larger. We look to a lot of our peers who’ve stayed focused, who’ve become experts at their craft. That’s what we look to emulate, and what we’ll continue to do.
Kevin Hartz
We can also extend out. We’ve shown to be very patient, and we’re not just trying to put capital to work. We wait patiently for those right companies, even if it takes a little longer to deploy.
Jack Altman
You guys are moving into this building, right?
Kevin Hartz
We are.
Bennett Siegel
We want our founders to keep moving up the floors when they raise their future rounds.
Jack Altman
Fund them, send ‘em over here. It’s beautiful.
Kevin Hartz
We could put in a fire pole.
Jack Altman
I love that. I think we dig a little hole somewhere. I don’t know if you’re above or below.
Kevin Hartz
Of course below. It’s a seed thing.
Jack Altman
Yeah, exactly. All right, thank you guys. This is super fun. I appreciate you both making time for it.
Kevin Hartz
Thank you, Jack.
Bennett Siegel
Thank you.