Biotech is the biggest bull market in A.I., says Lux Capital's Josh Wolfe

Our next guest was a witness in the hearing, and staying ahead of China is what drives his investment theory. So joining us now is Josh Wolfe from Lux Capital. Josh, it's great to have you back on the show.

Great to see you, Morgan. How are you doing?

I'm doing great. I'm curious what your key messages and key takeaways were from your interaction with House lawmakers earlier this week.

Well, first of all, I think it's a great committee, the select committee on the ccp. It's important because they're making a really fine distinction with the difference between Chinese, Americans, Chinese people, China itself and the ccp. So that's something that's really important. We don't want to have sort of racist or jingoistic messaging. The key messages that I took away, number one, it's a bipartisan issue, and that is something that is really rare, seeing people on both sides of the aisle really agree that this is a threat, not just an emerging threat, but a present threat. The main messages that I had were really three or four things. Number one was that there's a moral dimension to us as investors when we're making investments to people that are developing technology. You really have to consider the moral dimension. When we ship components and technologies overseas to peer companies and peer countries. We are not just shipping a piece of technology, we're shipping American values. And similarly, if you are importing Chinese technology that is directed by the hand of the ccp, you are also not just importing Chinese technology, you are importing Chinese values. And there is an American dream and there's a Chinese dream, and they're very different. So that was one of the key things in the morale morality piece. Second piece was on technology sovereignty. And what that means is twofold technological sovereignty. For any of the portfolio companies that we're investing in, we want to make sure that the supply chain is robust, that they're not solely dependent. And you see, of course, with the dominance in rare earth minerals, in battery assembly and selling, really important that you have technological sovereignty, but also that independent countries throughout the world are going to make a choice. Are you going to be a free and sovereign country, or are you going to be dependent upon China? And there's one technology in particular people have not really focused on, and that is nuclear reactors. This is an area where I actually think that the government is behind in identifying. We used to have 104 domestic nuclear reactors. They're down to 93. China went from 2 to 25 to 55 on their way to 150. And they are starting to export that to the rest of the world. I'm not worried about nuclear proliferation. I'm worried about the proliferation of dependency on China for electricity and all that comes with it.

Is this an area that you're making investments in light of that? I mean, it's interesting. I would have expected you to bring up AI. We have these conversations about, you know, so called arms race, but the conversation around nuclear not really happening so much.

Yeah, we've had some great successes in nuclear, on cleaning up nuclear waste and the Fukushima disaster and it's something that we continue to look at. You know, it is interesting on AI, of course, it is the topic du jour. I am very bullish on American AI, if you want to call it that, for a very simple reason. A lot of the foundation models that we're all using and experimenting with are built on the open Internet and repositories of data that are very free and open. And for the most part that means that we will approach an asymptote of truth. Okay. When somebody queries something, you're more likely to get an accurate answer. If things are censored and you're really relying on a repository of stuff that the CCP is putting forth, you're going to approach a repository that approaches only the party line. So I think over time we're going to be advantaged because we're free and open and we're trend towards truth.

Interesting. I want to ask you about that. Josh, good to see you again. By the way, John Lux invested in Hugging Face four years ago, a startup that helps companies build and deploy AI models. And you know, this year it's getting name dropped like crazy. What is the next phase of your AI investment? Since you've been pretty visionary in that, Is it industry specific models and those that are able to best create those? Is it other layers of the AI stack like, like health care where you already have investment?

It's a great insight. In fact, yes, we always say that we're contrarian investors and we want people to actually agree with us just later. And there's the five year psychological bias. And as you noted four years ago with Hugging Face, people want to be invested today where they should have been. Or five years ago when Brandon Reeves, who's one of the partners at Lux, brought this into our partnership, I was like, Hugging Face, we're investing in emoji. But yes, it has become basically the GitHub, the repository for all the key AI and ML models. So that was the first thing we've Also made investments on alternative chips and infrastructure and architecture. I will say everybody believes that Nvidia is the way to go, that Nvidia has a monopoly on this. And I actually think that there is some weakness in the cuda, the language that a lot of people are able to program those chips on. There's things like Pytorch and others that an ecosystem is developing that might give advantage to folks like AMD and even to distributed compute. One of ours is a company called Together Compute that basically said, wait a second, do we have to go from 10 to 100 million to a billion to train these foundation models? Could you leverage the large compute that's already out there? Could you repurpose compute that was used in crypto and basically have distributed compute? To your point, it's a great insight. I actually do think that health care is going to be one. I think financial services is going to be another. In both of those cases, you have institutions with a real proprietary repository of data. We heard for 10 years, big data, Big data, Big data really amount to a lot. Big Data now is the reservoir that a lot of the algorithms can draw upon and build. The biggest area that I think, not just in health care, but in biotech, I truly believe that it's great that we're making images, it's great that we're making sounds and videos. Runway ML, another one of our companies going to create full cinematic features without humans or actors. And so all the strike stuff that's going on, you know, it's really interesting in that conversation. But biotech in particular, I think is going to be absolutely huge. People that are training foundation models for true scale in biology, from designing individual cells to ultimately cellular medicines, I think is going to change humanity. That, to me, is the biggest bull market within AI.