TechCrunch’s StrictlyVC evening in Los Angeles late last week brought together two straight-talking investors currently working in AI. Carter Reum is the founder of M13, a startup firm with $2.5 billion in assets under management that has been a seed or Series A investor in 17 unicorns, he said. Chang Xu is a partner at Basis Set Ventures, which launched in 2017 as one of the first AI-focused funds and is now investing in its fourth fund, with nearly $1 billion in assets under management.
On stage, in a room full of sun in El Segundo, the two were as entertaining as they lit up, covering how to make prices in a market that has never moved quickly, how to find companies that will not get steamrolled by a hyperscaler, and what the SpaceX IPO will do LA The interview has been shortened and edited for clarity.
Is there an AI infrastructure bubble?
Chang Xu: There is both a bubble and not a bubble. It’s not a bubble because we’ve never seen this kind of growth curve before. ChatGPT went from $1 to $40 billion in six months in revenue – that’s unprecedented growth at the time. We have a portfolio company, Open Art, that went from $1 million to $10 million of ARR in the first year, and $10 million to $70 million in the second year, [and it was] the maximum income of that time is only 20 people. The bar for good growth has completely changed. When you have this opportunity to compound rapid growth, the valuation doesn’t seem so crazy because you’re calling that the final price. On the other hand, if you put the value of each currency in that equation, there is no way that it will work well with the portfolio. So it’s a weird time.
Carter Reum: I always laugh because we act like we’re the first in the world of real estate, but we’ve seen this before – with the cloud, and the iPhone, and the car in the 1920s, when people were worried about losing their jobs, and they lost, and life went on. This is stronger and faster, but has the same power. What’s different about this cycle is that previous cycles had founders versus innovators – Zuck versus Evan, Travis versus John Zimmer. In this cycle you have founders competing against founders, competing against the greatest, most funded founders the planet has ever seen, again competing with the ten largest technology companies in the world. And I would argue that for the first time in history, the incumbents actually have an advantage – technology, capital, data, talent. So as quickly as some of these companies rise, they may fall. I actually find it difficult to invest in a market like this. But if you get it right, you look like a smart person.
How do you price deals when startups are making money faster than ever but it’s not clear how sustainable they are?
Rheum: We always do cocktail napkin counts. We were looking at a business the other day – AI software for brands. I asked: how many were the winners of the last round? Will there be more products in the world? Are they willing to pay double or triple for the software this cycle? We ended up not investing because we couldn’t do the math.
Xu: We are always very close to what is technically defensible, because that boundary changes every quarter, maybe every month, sometimes every week. The framework we are thinking about is investing under AI and over AI. Under AI, you have all this infrastructure that’s being rethought — databases, version control, deployment tools — because they’re all built for people. Now you have agents using all this infrastructure, and agents need very different things. Last year I never thought you would need a new GitHub. This year I can count on two hands how many really strong teams are going after GitHub for agents. On top of AI, when things get too crowded, we always go back to: what is secure, and what has long-term separation?
How do invest in companies that won’t be divided by OpenAI or Anthropic or Google?
Rheum: We always try to imagine where they start and where they will end. It was obvious that they would follow the marketing and obvious places. So we have a thesis around the conflict as a moat – we like regulated industries. We had an exit just shy of a billion for a company disrupting 911 call centers with AI. Hyperscalers may get there eventually, but as a result of several billion dollars, they won’t get there anytime soon. Health care – they will go there, but there are many regulations that limit them.
What keeps us all up at night is that it can change on a dime. You looked to see them coming through the rearview mirror. I tell every inventor: you need a microscope in one eye and a telescope in the other. The microscope is day by day – what should I do this week, take it out. But you better get your binoculars out, because the world is changing very quickly. You have to be a domino player and a chess player, because your board is constantly changing.
Xu: The framework we use is: is this a deep market or a speed market? In fast-paced markets, fast-trackers are faster than ever — it’s all about speed of execution. In deep markets, hard things are still hard. We actually have a portfolio company that uses transgenic chickens as an alternative to drug production, because it’s more expensive to make complex proteins. Cheaper, obviously, if you have chickens do it. Chickens still take so long to hatch – for today [laughs]. Those are deep markets, and we invest accordingly.
Chickens anyway, do you see novel ideas at the moment, or more versions of old companies?
Xu: Both. Consistency categories – agents used in finance, agents applying to health care – you see a lot of really strong innovators going after them, and many of them are going to win. But the most interesting ideas are the ones where you think, ‘Wow, I don’t know if that could be a business.’ OpenArt, when we first started supporting it – right after that, Dall-E came out, Stable Diffusion came out, they started a discovery page of commands that you could write to get certain types of graphics to produce. How is that business? Absolutely no sense. They went from $1 million to $70 million in two years and have been growing rapidly ever since. There’s a lot of depth to that market that we can’t sort out. But from the beginning these were young inventors who tried something at the time that they found interesting, and they kept iterating until they found a business. If they had started a year later, they would have missed the window.
The thing about VC is that it’s always a case of bad ideas becoming good again. Four or five years ago you would have said it was a bad idea to invest in anything sold to Hollywood. And then we did a bunch of deals in creative AI, generative AI, which led to what companies today are doing incredibly well – first generating images, then videos, now world models. That world has become much bigger than we could have ever imagined when we looked at the previous generation of software that was sold to Hollywood. Then you have Cursor, which everyone says is an AI wrapper. An outflow of $60 billion. And researchers – when my husband did his PhD at MIT, his salary was above the poverty line. Now researchers are all Twitter followers.
Rheum: I think we are still in the first innings. The first wave of any technology cycle, even a steep and fast one, is often very visible – more competition, more congestion. The second and third ripples are where it gets interesting. Think back to when you were a child: if you take a heavy rock and throw it as hard as you can and make it bounce across the water, the heavier the rock and the faster you throw it, the longer the ripples. That’s what we’re going to have here. I get excited two, three, four years from now, because there will be business models and companies that we can’t imagine today. As a VC, those second and third bets are the hardest to get right – but if you do, few people think about it, you pay reasonable rates, and the ROIs are often much better.
The SpaceX IPO will put a lot of money in the hands of the people who live here in LA – workers in particular. What does that mean for this ecosystem?
Rheum: When Anthropic and OpenAI finally IPO, there will be a lot of VC and institutional investors. Never has this amount of money been returned and distributed as widely as SpaceX will. If he is there [in this room] you have a house for sale, a boat, an airplane – definitely use that ride. But more importantly, every major liquidity event produces a second wave. The previous LA cycle produced the likes of Riot Games, Tinder, Snap. This is a different order of magnitude.
Three years ago everyone said San Francisco was dead. Turns out it’s a little deadlier than people expected. I think it would be the same with anyone who writes LA There are so many very smart people here – technically, but also people who understand product, content, creators, influence. This first wave is a technology wave, and the technology talent is concentrated elsewhere. But what comes after the waves of technology? New business models, creative thinking, cognitive culture. That’s going to be the next wave, and I think it’s more likely to be centered in LA
Xu: The interesting thing is that the next frontier in AI is not so convincing – taste. Making films, making videos, making emotional things, making things connected to certain cultures. San Francisco has amazing tech talent, and that’s exactly what models get so good at doing and speeding up. LA has taste in spades.
If you shop through links in our articles, we may earn a small commission. This does not affect our editorial independence.
