I wanted to know when a curve stops being a hockey stick and becomes a civilizational break. Marc’s answer starts with cheap energy: water, food, geopolitics, national security.
Then we get to AI. I asked whether this time is different because we are creating a species more intelligent than we are. Marc thinks it is — and that both abundance and extinction stay on the table.
Transcript
Marc: shallow tech is fine. It's not for you. It's not for me And I'll give you an example of it now One of the things I'm doing now is I know in the future if we have cheap energy it will change every percent Everybody will have sweet water at almost no cost. Therefore, you have health outcomes because you've got the water You have food, nutrition who cares about geopolitical tensions that are connected to energy. Those countries become obsolete in a very big way. National security. Imagine thinking about it as a country that has on their income statement an energy line GDP, and you squish that down to by 90 Imagine businesses that have an energy line
that you squish down to 90 Some businesses as ideas, entrepreneurial ideas, are sitting in the garbage can because they don't pencil because of the energy component, suddenly come to life Some businesses and ideas that are revolutionary that have never been conceived of because intuitively you can't do that, now suddenly come to life So an entire class of entrepreneurs pops up or even businesses reinventing themselves. Imagine households and their line of energy and you squish that by 90%. Now all of that squishing results in surplus It's like savings You're saving money on energy. Well, savings is an investment, equals investment in GDP in any economics. So now you have all this money to invest
in the future and that investment produces who knows what kinds of dividends for humanity So I'm saying that the civilizational level of discontinuity discontinuity is a curve is going along and suddenly something happens and it rips the curve It doesn't just accelerate like a hockey stick. Hockey sticks are not revolutionary. They're just like, wow No this is when you break the hockey stick You put one end of the stick up high and the other end down low and there's a dot dot connection. That's the discontinuity And I see energy as being that, and the only thing that can produce that is fusion And that's the only thing that can produce it. The rest is incremental
keeping up with global demand Which by the way, is doing pretty well with natural gas coming on stream, coal is still booming in the industry and even hydro and wind. solar and battery storage they're all doing just fine but they're incremental Fission is a bit of a jump, it's still incremental. Fusion that's it. So what I am dedicating some part of me in the next 10 years is to make that happen and it's in the form of the SPVs and the fund put together to invest in Oh wow okay. That's proactive Yeah, I want to do that Not because they need our money necessarily but because it puts me in the middle of my
friends in the middle of that cauldron we can figure stuff out and either go back companies, create companies, do something to accelerate that thing, that thing called fusion by some number of you Yeah, great. And if I can mobilize a billion dollars to do it, I will make a lot of bets, investor-style bets By the way, an investment is push from the present Investing is push from the present Yeah, because you're picking a company that you think can push through the challenges of supply chain and Why isn't investing pulling from the future Good investing is absolutely not Thank
Pablos: you. I'll take it. What
Marc: you need to do, this is why I said necessary and sufficient, is both. You pull from the future is the vision, the conviction of what society what humanity will look
Pablos: Which is fusion, so that's the pull And the push is this fusion or that one. Which supply
Marc: chain, which reactor, which The two come together But there's no connector between the present and the future
Pablos: For probably different reasons, I arrived at very similar conclusions. Energy is the single biggest thing that we've been investing in And after this, let's talk about fusion, because a couple of things might be surprised but do you think there's a you know it's probably worth digging into since it's what's under everyone's mind do you did is your feeling about AI that something fundamental changes about this process obviously you know if somebody were gonna think of breaking the hockey stick and having this kind of discontinuity. AI is one of these drivers
Marc: For sure. You're asking a more general question about human evolution
Pablos: is it different this time? I think that's really the
Marc: Let me tie off the fusion point on that theme and then get back to your bigger question The problem with fusion is plasma management For obvious reasons. It's a chaotic, broiling world that has to be in real time And the Sun does that through gravity because it's massive But we don't have that gravity effect inside a container chamber What that means is the following As the fusion is happening, for it to continue to be productive, you have to manage its dynamics. Its dynamics are all quantum Everything about fusion is quantum scale And what does that mean? That means that there are two other investments One for AI and one for quantum. I'm going to explain what it is
You need sensors that are sensing anomalies. Either the thing is collapsing or the thing is bursting. You can't collapse because then you quench the entire process It can't burst because then it shuts down the entire process to protect the contagion It has to be like that So you have to sense where is it, what's going on The sensing has to be done at quantum speed. These are quantum sensors It has to be analyzed away. That's AI What is this? at that speed same speed because this is happening very quickly. Once the AI processes the information coming from the sensor it has to direct it to lasers. Because in the future that I see, there are thousands of lasers
surrounding no one has done that, but there are thousands of lasers surrounding the chamber. Not like six
Pablos: but thousands. More than 192 huh 100,000 Oh, wow
Marc: the conversion one stops No single one is like gigantically connected Those lasers have to be pointed charged up not too much, not too little, Goldilocks lasers and shoot to make the entire roiling broiling thing be well-behaved then, sense again, have I been affected Ask the AI says, not exactly. Reshoot And do that until that particular anomaly is dealt with while we're sensing the rest of the… So it's quantum lasers. It's quantum sensors. It's quantum compute. And the only quantum compute that makes sense here is not a general purpose computer. It's a single thing that does one thing and one thing only is this solve for this equation And with one set of algos and
some machine Wow So there's quantum and AI. And when I said that I was invited to do a fund on quantum. I said, quantum has no piece This is a piece
Pablos: that's true. You've just completed the two hardest things into one
Marc: That means I can discard 98% of AI and 98% of quantum. Yeah, right Fair enough. Okay 95 because I want this one through line Yeah Does that quantum and does that AI contribute to So if someone is making a quantum chip it is an absolutely special purpose chip a good one would be a breaking RSA Special purpose chip That's all it does Special purpose algo That's what I'm interested in the quantum computing industry one of those companies will produce that, can produce that chip And with the algos attached
Pablos: to it. You think they can now
Marc: Because they don't know about it. No, they don't know about this application This is an ASIC. This is like a hyper-focused ASIC. Application-specific integrated circuit It's very application-specific. It does one thing The mechanism I just described All right, well that
Pablos: sounds like an ambitious way to solve fusion I hope it works I have no idea I
Marc: have no idea. At this time, I'm standing in that vision and seeing what happens 20, 30 other problems to get fusion to work. Yeah. Fuels and huge supply chain stuff. This is the one I think it's missing to invest into No one's doing it, which is a problem No one's doing it. Chances are I'm wrong
Pablos: Yeah, great investment thesis
Marc: Not a good one So I have to do a lot more digging into who is doing this. and how credible they are. Back to your bigger question Your much bigger question is AI, different in kind, that's so different that it breaks the hockey stick with discontinuity and actually enables a conversation, a real conversation about abundance and extinction Eutopian and dystopian. That's my super intelligence keynote okay it's online
Pablos: cool yeah we'll for sure show it yeah
Marc: it was actually where i met you no no i don't think i i didn't catch it were you in the we met at that summit
Pablos: event but i wasn't i was i didn't catch your keynote right i wish i had yeah i don't remember what happened there was some reason i remember now there's some reason why i couldn't be there because it but
Marc: it's not it's it's you can find it on YouTube or something. So you can go see it. That's the thesis okay. The thesis is that AI is instrumental as a novel completely novel experience in all of human evolution All right Because we are creating a species that's more intelligent than we are and that's never happened before. Humans have always been at the AI
Pablos: that or we killed them off, but yeah
Marc: I don't think there were any We don't know But certainly within the dynamics of deep history, if we saw some people that were smarter than us, we
Pablos: would definitely try to kill them. But
Marc: eventually they'd prevail because they're smarter. Maybe that's true You have better technology, better logic do think it is absolutely going to create