Reid Hoffman

And a regulatory agency backed up by fearful public opinion, unknowledgeable, ignorant public opinion, but fearful too, hasn't done anything about this. And then you say, well, what's the best way we worried about running out of electricity We're worried especially about green electricity going and doing this.

And so I myself have been investing in both fission and fusion over the last five to 10 years entirely philanthropically. I have no idea how to invest in this. When I do these investments, I write them to zero in my book. I have no idea. what the economic outcomes are of this, unlike like internet investing or software investing or what's the rational way of doing it.

But creating the technology for the benefit of society and humanity is extremely important. I've been putting lots of money into this out of a desire for putting America in a much better position, helping the world with this new technology, etc.

And given that it is an inevitable, it's a physics process. I worry about that. That's one of the, I think, likely outcomes of the next administration that will be a serious impact. Now, that being said, like you said, okay, bull bear on the next administration, they'll go, oh no, climate change is a fiction and it's a liberal conspiracy and you know, dah, dah, dah.

And you're like, why would the liberals bother with a climate change conspiracy? I mean, it's just like the logic of this is not great. But what they might do, which I would hope, again, be hopeful for, is they say, look, what we really need to be doing is doing the Apollo projects on the right kinds of clean energy.

And we're going to now go 10x down onto that, including all versions of nuclear, as well as everything else. And that actually might have a very positive impact over time in what we're doing with climate change.

Well, it's a frequent misconception because there's obviously a growth of a electricity curve that AI has got climate impact. One, it's something like data centers, I think, or something like it's some single digit percentage of all power consumption, like three something. So it's not very big. And then AI within data centers is like a single digit percentage of that.

So currently, AI is not accounting for any of the electricity that makes any of the difference. No, but it's growing. And you go, well, what happens if it continues to grow in a compounding way? And so and so everyone goes, ah, I can be I can be green and climate by posturing in the media, by saying AI is bad for climate. Pay attention to me, please. Lay, pay attention to me, please.

This is a little bit like the people who are being anti-nuclear, which then caused a whole bunch of coal plants to be built, which caused a whole bunch of environmental degradation as a consequence of being anti-nuclear. It's a similar thing because AI has several points of being massively positive on climate. Here's a simple actual data point.

So Google's engineers, some of the best in the world at running data centers, thought we have run our data centers as efficiently as possible. It was, I think, five, seven, eight years ago, DeepMind ran their algorithms on Google's data centers and figured out how to save 15%. of the electricity. So think about you're applying AI and intelligence to go, how do we run our grids more efficiently?

How do we apportion costs? How do we do all this stuff? How do we add intelligence to all this? That's one. Two, maybe even more significant, which by the way, that means AI technology can actually be a net energy saver as we apply it to the inference cost to all the electricity we're doing in the future. Number two is, well, what's happening as a boon for green energy right now

All of the major hyperscalers are doing multi-billion dollar commitments to green energy companies saying, if you can deliver the energy, we will give you the initial customer capital to build that. And by the way, we will subsidize because energy is always expensive when it's small and cheaper as it gets bigger.

We will pay the expense at the beginning to start building your nuclear, to build your geothermal, to build your solar. in order to do that. So they are functioning as the R&D venture capital for creating green energy for the entire society in building their new data centers. So that's another thing where AI is coming out to building green energy.

Look, the most far-fetched and unclear is it's also possible, but I think this is more of a science fiction thing is, well, kid, the patterns of intelligence and AI, and people are doing this work, like applying it to, can we get it to make fusion to work?

Can it solve certain problems that then make sustainable green energy possible because of the intelligence that AI can bring in its massive computation to solving these problems? And people are working on that on fusion containment, right? Because if you can make fusion work, you can solve climate change. Can you just unpack that for me, sorry? How does fusion completely solve climate change?

So fusion is the most scalable green energy. You know, the sun is powered by fusion. And obviously, there's various ways to try to use solar that I don't think we've tapped out on and everything from satellites with solar energy collection to various things on the ground. So I think there's a bunch of stuff there.

That's actually, in a sense, an indirect fusion because you have this massive fusion reactor that's known as the sun. But microfusion is both super safe and can produce enormous amounts of power very cheaply. And once you have enormous amounts of power, you can solve significant world crises. So for example, how do you do decarbonization?

How do you pull carbon out of the atmosphere in order to reduce global temperature back to that great band that humanity prospers in? And the answer is it takes a lot of energy.

Well, if your energy is carbon producing as you're doing it and you're not net doing it, so you have to be carbon neutral to negative on this, then if you can produce it at massive scale, you can pull carbon out of the energy, let alone power all of the things that human beings want, houses and transport and all the rest. So it's more than the silver bullet.

Well, I think it's a significant step forward. I don't know exactly where we will be in scale quantum computing. Usually, again, of course, the discourse with new technology always starts with the fears, which is a mistake that I'm always trying to correct. because it doesn't mean you shouldn't have concerns and anxieties and navigate.