Peter Thiel
speaker
1,088 appearances
12 recordings
11 series
first heard Jul 2024
last heard 15 Jun
Peter Thiel’s voice in public audio — every appearance, attributed to the second.
Trend
recordings per month · last 12 monthsRecordings per month over the last 12 months — 4 in all, peaking in Jun 2026 with 2.
Appearances
intelligent thing, way, way God-like intelligence, way smarter than any human being. And then there was sort of the, I don't know, the CCP Chinese Communist Rebuttal, the Kai-Fu Lee book from 2018, AI Superpowers. I think the subtitle was something like The Race for AI Between Silicon Valley and China or something like this. And it was sort of it defined AI as it was fairly low tech.
It was just surveillance, facial recognition technology. We would just have the sort of totalitarian Stalinist monitoring. It didn't require very much innovation. It just required that you apply things. And basically the subtext was China is going to win because we have no ethical qualms in China about applying this sort of basic machine learning to sort of measuring or controlling the population.
And those were sort of like, say, two extreme competing visions of what AI would mean in the 2010s and that sort of maybe were sort of the anchors of the AI debate. And then, you know, what happened?
in some sense with ChatGPT in late 22, early 23, was that the achievement you got, you did not get superintelligence, it was not just surveillance tech, but you actually got to the holy grail of what people would have defined AI as from 1950 to 2000. 2010, for the previous 60 years before the 2010s, people have always said AI, the definition of AI is passing the Turing test.
And the Turing test, it basically means that the computer can fool you into thinking that it's a human being. And it's a somewhat fuzzy test because obviously you have an expert on the computer, a non-expert. Does it fool you all the time or some of the time? How good is it? But to first approximation, the Turing test, we weren't even close to passing it in 2021.
And then ChatGPT basically passes the Turing test, at least for like, let's say an IQ 100 average person. It's passed the Turing test. And that was the holy grail. That was the holy grail of AI research for the previous 60 years.
And so there's probably some psychological or sociological history where you can say that this weird debate between Bostrom about superintelligence and Kai-Fu Lee about surveillance tech was like this almost like psychological suppression people had, where they were not thinking, they lost track of the Turing Test, of the Holy Grail, because it was about to happen.
And it was such a significant, such an important thing that you didn't even want to think about. So I'm tempted to give almost a psychological repression theory of the 2010 debates, but Be that as it may, the Turing test gets passed and that's an extraordinary achievement. Then where does it go from here? There probably are ways you can refine these.
It's still going to be a long time to apply it. There is a question. There's this AGI discussion. Will we get artificial general intelligence, which is a hopelessly vague concept, which general intelligence could be just a generally smart human being. So is that just a person with an IQ of 130? Or is it superintelligence? Is it godlike intelligence? So it's sort of an ambiguous thing.
But I keep thinking that maybe the AGI question is less important than passing the Turing test. If we got AGI, if we got, let's say, superintelligence, that would be interesting to Mr. God because you'd have competition for being God. But surely the Turing test is more important for us humans. Because it's either a compliment or a substitute to humans.
And so it's going to rearrange the economic, cultural, political structure of our society in extremely dramatic ways. And I think maybe what's already happened is much more important than anything else that's going to be done. And then it's just going to be a long ways in applying it. One last thought. You know, the...
The analogy I'm always tempted to go to, and these historical analogies are never perfect, but it's that maybe... AI in 2023-24 is like the Internet in 1999, where on one level, it's clear the Internet's going to be big and get a lot bigger, and it's going to dominate the economy. It's going to rearrange the society in the 21st century. And then at the same time, it was a complete bubble.
And people had no idea how the business models worked. Almost everything blew up. It didn't take that long in the scheme of things. It took 15, 20 years for it to become super dominant. But it didn't happen sort of in 18 months as people fantasized in 1999. And maybe what we have in AI is something like this.
Figuring out how to actually apply it in sort of all these different ways is going to take something like two decades. But that doesn't distract from it being a really big deal.
Yeah. I've often, probably for 15 years or so, often been on the side that there isn't that much progress in science or tech or not as much as Silicon Valley likes to claim. And even on the AI level, I think it's a massive technical achievement. It's still an open question. Is it actually going to lead to much higher living standards for everybody? The internet was a massive achievement.
How much did it raise people's living standards? much trickier question. But in this world where not much has happened, one of the paradoxes of an era of relative tech stagnation is that when something does happen, we don't even know how to process it. So I think Bitcoin was a It was a big invention, whether it was good or bad, but it was a pretty big deal.
And it was systematically underestimated for at least the first 10, 11 years. You could trade it. It went up smoothly for 10, 11 years. It didn't get repriced all at once because – we're in a world where nothing big ever happens. And so we have no way of processing it when something pretty big happens. The internet was pretty big in 99. Bitcoin was moderately big. The internet was really big.
Bitcoin was moderately big. And I'd say passing the Turing test is really big. It's on the same scale as the internet. And because our lived experiences that so little has felt like it's been changing for the last few decades. We're probably underestimating it.
It can be a big change culturally or politically. But, yeah, the kinds of questions I would ask is how do you measure it economically? How much does it change GDP? How much does it change productivity? And – And certainly the story I would generally tell for the last 50 years, since the 1970s, early 70s, is that we've been not absolute stagnation.
We're in an era of relative stagnation where there has been stagnation. very limited progress in the world of atoms, the world of physical things. And there has been a lot of progress in the world of bits, information, computers, internet, mobile internet, now AI.
Showing 721–740 of 1,088 · page 37 of 55
← Previous
Next →