Dwarkesh Patel: evaluation

Transcript context

…You were going to run out of whales. I love that this professor at Yale, kind of a consultant, did this experiment. He needed some extra money and he did some studies that showed that actually this stuff called rock oil, you could turn it into a lighting fuel fluid. I love the risk taking of it. But it's hard to imagine… we wouldn't be where we are. We wouldn't have the world. Today, it wouldn't be a world of eight billion people were it not for it. Obviously, there's going to be change. I'd say right now, the incentives for innovation are there. That's why we may see a runway of what's going to come, but it may really come from the side. And something else that the kerosene is the fact that oil for the first 50 years is used for only lighting. Another thing that's interesting about that is people are asking now about these AI models. You can literally get a million tokens, like many books length of content, out of these models for 15 cents. This is one question people are asking. Let's say you did a hundred billion dollars worth of tokens, what does that look like? What does an industrial scale use of intelligence look like? With crude oil in the beginning you're producing a certain amount, but you had a glut because you're only using it for lighting. You then discover this industrial scale use of this technology, which is obviously motorized transportation. That’s one question you can have for AI. Currently what we're using these models for is research and chat and whatever. That’s like the kerosene. What would the equivalent of billions of vehicles look like for AI? That's a question that I'd like to ask you. It is a sense that we are at the beginning of something new. I remember a political leader in Central Asia saying, “AI is going to be the true source of power in the future.”…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.