00:00I think your investments are so interesting. Yes, you do have some of the models like perplexity,
00:04but I feel like yours is an industry VC in growth that is done really well because of like
00:09an anthropic investment or open AI. Why go into kind of the more, you know, the different sort of
00:18ecosystem and not necessarily put all your eggs behind like the foundational LLMs like
00:21a lot of your peers have tried to do? I think the foundational LLMs are really important. We're an
00:26investor as an example in reflection AI, which is an open source LLM that really competes with some of
00:32the foreign open source LLMs because we think it's really important to have open source. But one of
00:38the reasons that we think open source is important is because of the enterprise. We think enterprise AI
00:43that large enterprises want to be able to have access to not just the closed models, but to open
00:49models. They're more cost effective. They can be customized. And that's why we've really focused on
00:54things that drive AI at the enterprise level. See, that's really interesting to me, this idea of
00:58cost, because it has been this bubbling up problem that people have started to talk about.
01:02The information had a story earlier this week, just talking about like the skyrocketing use
01:06of open source models. Do you think that that will only continue or do you think
01:11anthropic open AI Gemini will find ways to like bend space and time and make everything cheaper for
01:15their users? Listen, I think that, you know, anthropic open AI Gemini are amazing companies,
01:21but I also believe that there's space for these open source models, which are there are a fraction
01:25of the cost that tokens are much cheaper. And frankly, you could just really customize them.
01:30And ultimately, probably the most important perspective we have at B Capital is the models
01:34keep getting better and better, but it's the contextualization of the model that matters.
01:39And so we need to spend more time molding AI into the enterprise and not just building better
01:44and better models toward AGI. What does that mean of how you're viewing software and software
01:48investing right now? I was, I was down in Berlin with Orlando Bravo, who was like the
01:51SaaSpocalypse, which is a horrible term to begin with. It's over. It's dead. It's not happening
01:55anymore. I mean, can you invest in software for the enterprise right now?
01:59Listen, I think the use case of software is fundamentally not going away. And there's a
02:04little bit of a belief that we're all going to build our own software and therefore we won't
02:07need enterprise software. I'm a bit skeptical of that. I think enterprise software, the use case
02:12will remain. Yes, there will be some cases where you will build your own software,
02:16software. But really what we need to see is that those enterprise software companies become AI
02:20first. A big example of this is if there's a three-day training to learn how to use an enterprise
02:26software, that software is not going to work. You need to be intuitive. The way that chat is intuitive,
02:31you have to be able to sit down and use enterprise software with no training. And that's where I think
02:35AI first companies have done incredibly well. By the way, when we teased you before the commercial
02:40break, we basically said that B Capital has this view that AI is underinvested. How? I mean,
02:46so much money is being thrown at this thing. Where is AI being underinvested?
02:50I think it's in traditional companies. As an example, you look at places like financial services
02:54and financial services is an area where they are adopting AI. But I think the scale with which
03:00financial services will get transformed with AI, we're just starting to understand it. There's
03:06trillions of dollars of assets that, you know, frankly, are still being managed in a very old
03:11fashioned way. And where the investment needs to happen is I think you'll see venture and private
03:15equity firms really investing and purchasing traditional companies and focus on the AI
03:21transformation of those companies. Because we all learned in the 90s and early 2000s that, you know,
03:27fintech didn't mean that banks went away. It just meant that banks learned how to use fintech better.
03:33And I think this time around, you'll see financial service companies that adopt AI can actually be
03:38the winners. I mean, everyone is talking about this idea of trying to transform companies. How
03:42many people in your industry do you think like actually get it and are going to be successful
03:46at that? It's really tough because the private equity firms have been doing transformation for a
03:50long time. I'm not sure if they have the AI chops and the venture capital firms are really good at
03:56AI,
03:56but owning and managing companies is much more than the AI. There's the people aspect. There's the
04:01governance aspect. There's just other aspects of growth. So I think you'll see the lines between
04:06venture capital and private equity blur. And I think it's an exciting time to be where we are
04:11today. By the way, what do you think happens? Because we're about to get more IPOs. We already
04:15had SpaceX, but we're going to have more AI IPOs with OpenAI and Anthropoc. What do you think
04:19happens to VC that gets those proceeds back? How do you think it will change the ecosystem?
04:23Well, you know, it takes six to 12 to 14 months to get those proceeds back. That's incredibly
04:29healthy. We've had a bunch of IPOs. We were an investor in SpaceX as an example. But it takes a
04:35while for that money to come back around. And what we're hearing from our LPs, from the
04:39institutional asset owners, is that they're really appreciative that those proceeds will
04:43come back because that's what makes the venture capital and private equity ecosystem work. And for the last
04:48four or five years, that that circular ecosystem hasn't worked as well.
Comments