Episode Transcript
[00:00:00] Okay, everyone, I want to give you my very early perspectives on the America's AI Action Plan. There's a 28 page document from the White House and then President gave a speech yesterday for about 30 minutes covering some of the topics, but he spent about three quarters of his time talking about other things. But a couple things that I think are just worth thinking about and just want to make you aware of. First of all, it's a very bullish position for the country trying to reduce regulation and increase investments in AI technology, AI data, AI and the military, and license our technology to other countries.
[00:00:39] So the idea that we want to hold AI as a proprietary advantage relative to our adversaries seems to have completely flipped in that they're encouraging open source AI, which is releasing the AI models into the outside world so the Russians and the Chinese can use them. And they are also a big fan of license technology to other countries, which I know Nvidia made a big push for, so that we could prevent the Chinese, for example, from building their own chips because they would be using ours. So we'll see how that plays out. It's a kind of a flip flop from where we were before. There's a lot of discussion in the AI act about supporting deregulation of energy data center development and infrastructure development, which is great.
[00:01:25] It's not clear how much money from the federal government's being allocated to this, but that comes in different forms.
[00:01:31] There's a significant section on skills and workers, which is terrific. Encouraging the development of AI literacy, creating a research center for the role of AI in jobs, even promoting the funding of direct subsidies to workers that have been displaced by AI and potentially tax benefits for companies to reskill people of education, which is great. A lot of that already exists. So that's all really incredibly good stuff. There was a couple of strange things in there. There is a section in the policy which explicitly states that we want to promote American values in AI and that the AI tools should not discuss diversity, equity and inclusion or global warming. That's a weird one. It's a little bit odd. I'm not even sure it's possible, to be honest, but it's in there, you know, because this is a political document. There's lots of push for investment in AI in the military, which is great. Obviously we need that. There's a lot of continued push for increased speed of permitting for semiconductor manufacturing. Although the problem in semiconductor manufacturing isn't just building the manufacturing process, it's hiring and finding the skills. And we're talking to a bunch of companies about that. There's a lot of discussion about that issue. I mean, I think the general theme of the 28 page document is that all over the government we're going to make investments, positive investments in AI. But you can tell that there was a lot of lobbying for small things in there from different parts of the industry. And I think they did a pretty good job of not forcing the government to make specific decisions on any particular technology in any way. There is a discussion of a chief AI officer in the federal government, which will be an interesting position, which is kind of almost an impossible job if you think about it, but it'll be interesting. And then there was a pretty interesting discussion that the President made about not regulating the fees or royalties of intellectual property, in a sense. He stated that we can't be paying for books that are consumed by AI. And I think that's a little bit of a problem for those of us that are in the intellectual property business. I'm not sure that it's really a legal document, but as a position it seems a little bit odd to. Because we actually probably want to build technologies that allow the AI engines to license content and not just steal it. And I think that's a technical problem, not a legal problem, because we already have copyright laws. But anyway, that's in there. So I don't know that there's anything particularly significant in the 28 page document other than a very strong commitment. And it's clear that the administration and David Sacks have spent a lot of time thinking about all the aspects of this, including the data, the skills, the infrastructure, the technology, the semiconductors, the issues with defense, the issues of licensing and so forth. So it's a good step forward in clarifying that we're all behind this. And I think the reason this action plan probably is going to be useful is the agencies that are embarking on various AI projects now have some guidelines on what to do about it. But let me sort of add some other things. You know, the AI industry is exceptionally interesting and exceptionally powerful and very young in terms of its value. Companies are barely beginning to adopt and understand how to implement various AI solutions. We talk to companies about this all virtually every day. And the use cases are growing and growing and growing, and the technologies are growing and growing. But you could also argue that we're a little bit ahead of our skis because the revenue from Google, from Microsoft, from OpenAI, from Anthropic, from Amazon are still a small fraction of the infrastructure that's been developed, including xai by the way. So we're in a stage of massive infrastructure building for the future in a horse race to make sure nobody loses their edge, their position. You know, Meta is trying to stake out the glasses market and the market for agents and animated animatrons, although I think Elon beat them to that with the new version of Grok. There's obviously infrastructure being developed by OpenAI to build agents and devices with. Jony, I've Amazon's got AI all over its business. I think Amazon's in the most practical mode and then the cloud licensing from Amazon, Google and Microsoft is, you know, basically the tool set that we're all using to build all the tools that we're building, including most of you. So it's a pretty exciting time to be a part of this because we're all part of this new era of technology and invention. It's almost as big as when the first PC hit the market in 1981. For those of you that were around, we had no idea the unlock and unleash that would take place with devices other than mainframe computers. And this is going to be very much like that. Deep fakes, automatic creation of music, video, audio, design tools, data from third parties being collected and maybe not paid for it. There's just a million issues here. Bias in recruiting, bias in promotion. Let me tell you where I'm at. We just spent a day with one of our largest partners yesterday. We're doing a whole bunch of things on Galileo that are going to come out. Some of it is going to be available in August. This stuff does a lot more than you realize. We started a Galileo Advocates program for those of you that are using Galileo and doing cool stuff. You'll be hearing from us and you can join that program to tell us what you're doing and we'll share it with others. And the use cases are really kind of astounding. Yesterday, for example, we had a call with two clients. One was one of the national labs in the United States government and they're using Galileo for all sorts of talent issues and analysis of skills and very technical things in the military community. The other one was a large media company with some people that do a lot of program analysis in hr and they are using Galileo to build evaluation models for their HR investments, which is a big problem. Of course, we never really know the value of some of these HR things we do. These are. These are some really interesting use cases, to say nothing of just skills assessment, building onboarding guides and various documents and policy documents inside of companies and then building learning programs. You know, Galileo Learn is on fire. And we're talking to lots and lots of big companies about it. And so there's massive benefits beginning to unfold here. And there will be a lot of job displacement, no question about it. But those of you that are willing to just strap in and learn new technologies and play with tools, you're going to be fine. You're going to find new things to do. We're not going to run out of jobs. Don't worry about that. And the government's now at least well educated. I think the, one of the best things that's happened in this administration is through David Sachs and others. There are people in Washington that get it. They understand the technology, they understand the vendors, they understand the markets. So we don't have this strange situation where the government or the Congress asks people silly questions and they don't understand what the technology is. The policy decisions are tricky, and there'll be many, many more yet to be made. But I feel good about where the country's going personally from that standpoint. More to come on all this. We will be telling you more about it as time progresses.
