How I Invest · 5 min · Oct 28, 2026

Early AI markets can produce more than one leading company

GEGavin Ezekowitz · Co-Founder & Chief Investment Officer, BFA Global Investors

He bought into Anthropic after a forced FTX share sale, while seeing room for multiple AI leaders.

In this episode

9:07 Underwrite the team and milestones, not a distant market forecast6 min
29:19 Venture returns require convexity, pacing and position sizingPro pick 6 min
48:09 Avoid letting macro timing override your investment mandatePro pick 5 min
1:49Early AI markets can produce more than one leading company5 min
26:59 SpaceX's compute buildout shows the value of anticipating shortages3 min
37:28 Small funds can turn a focused founder network into an edge5 min
43:18 Small initial stakes can dilute a fund's biggest winner3 min
1:06:49 A manager's response to a mistake can end an LP relationship5 min

“what I missed, it's so clear in that company now, is that that was a company that iterated multiple times on their journey.”

GEGavin Ezekowitzat 5:31 · How I Invest

Takeaways

FTX's forced share sale created an entry at a $17 billion valuation.
He expected multiple AI leaders, not one inevitable winner.
He missed Cursor's ability to iterate and disrupt its own product.

About this segment

He bought into Anthropic after a forced FTX share sale, while seeing room for multiple AI leaders. Missing Cursor taught him to look for teams willing to disrupt their own products.

GE

Gavin Ezekowitz

Gavin Ezekowitz is co-founder and chief investment officer of BFA Global Investors.

Built a career across capital markets and investment banking, then became a venture and growth investor.

David Weisburd, Host

From “Gavin Ezekowitz on AI Valuations, Asymmetric Returns, and Why Great Investors Fail”, Oct 28, 2026. Watch the full episode

Transcript

From the transcript How I Invest publishes with the episode.

2:01
They're all GP relationships. I'm curious because you mentioned iAnthropic. You invested originally at a $17,000,000,000 valuation. What was your thesis there, and how's how's that evolved? Well, my my thesis there was very, very simple. I'd observed Anthropics appreciation up to... It was valued at 30,000,000,000 in the in the the round that had just occurred. But, of course, the the FTX had this bankruptcy, and the bankruptcy had to sell their shares of Anthropic.
2:36
And to me, it just seemed obvious that this was an amazing opportunity. We fortunately had formed a relationship with a guy who we've done a lot with. They were bidding for, I think at that time, dollars 10,000,000 or $15,000,000 worth. We were there in size. We had to really convince people because it was not... The company didn't look anything like it does today. It was really pre Claude. And it looked like OpenAI was just going to win everything, if you think about that period.
3:10
And I think that there were a couple of views that we held at the time. One was, and it's a view I hold today, was that when you're very early in a long cycle, people try to predict who the eventual winner will be. And most often, it's not that company. It's not Yahoo or whatever it is. Multiple There winners, or they could be different leaders. And so it has turned out with Anthropic, And our view is quite simple.
3:42
They look like they've got, you know, a tremendous shot at being the other company or one of the other major companies. This feels like a really great price. Let's give it a go. Yeah. This first mover advantage, I've thought a lot about it in technology because very few of the companies were first movers. Google, I believe, was the eleventh search engine. Facebook was maybe even further along in the twelfth or thirteenth search engine.
4:08
But yet, this first mover advantage persists. And I think the reason for that is, first of all, the first mover advantage is quite persistent in other asset classes. So outside of venture capital, the first mover advantage tends to tends to hold much more strongly. And secondly, of course, people invest for two reasons, fear and greed. And I think fear more than greed, and people just get afraid of losing that that number one player.
4:36
And even if they passed on OpenAI for one round and two round, three round, some some people... At some point, those investors capitulate, and they they wanna go and go into these momentum investments. I think that's that's true. I think I don't know that it's different in in AI, but one of the observations that I would make is that it is very evident from the early days of the the AI phenomenon that this was there there was gonna be multiple disruptive events, whether you can recall even recently the deep sea.
5:11
You know, suddenly, we had these Chinese open source models, and you have all of these events that were gonna take place. And I think my thinking from the earliest days was this is neat, but it's very structure suggests lots of disruption and even lots of disruption of the disruptors. And, you know, you can observe a company I I missed. I had lots of I had lots of opportunity at ten billion on Cursor.
5:39
It was like going once, going twice. And they all like, Oh, no, I don't think so. And what I missed, it's so clear in that company now, is that that was a company that iterated multiple times on their journey. And in fact, even now, I was with the leadership team some time ago. They're iterating again within their current x AI environment. I miss the fact that some of those really special companies are able to innovate, disrupt themselves, watch the new, new thing, catch on.
6:15
They've got the quality of the team. They're not married to a singular thesis, that this is the only way, and we're just going to plunge ahead. They've got a very flexible mindset. And I think that if you look at that approach, you can see that, yeah, we may see vast differences in the kinds of companies that emerge out of businesses that start very differently.