Dwarkesh Podcast / episode intelligence
Speakers in the public record
Claim mix
Evidence policy
Every row below preserves an exact excerpt. Identified speakers are linked; unresolved voices are labeled and excluded from people counts.
Claim ledger
22 published records
01 / evaluation
“Correct me if this is wrong. It seems like you’re implying that right now we have models that are on a per token basis pretty smart.”
02 / belief
“I'm not sure how you would maintain this equilibrium for a long period of time, but I think if we got to that point we would be in an okay position.”
03 / prediction
“I expect that models will be able to use websites that are designed for humans just by using vision, after the vision capabilities get a bit better.”
04 / belief
“Some of the people are very talented, and we even find that they're at least as good as us, the researchers, at doing these tasks and they're much more careful than us. I would say the people we have now are quite skilled and conscientious.”
05 / belief
“I agree that you'd probably also want to supplement that with some kind of fine-tuning.”
06 / belief
“There are a lot of different, separate axes for improvement. We think about data quality, data quantity.”
07 / belief
“I think people will still have different interests and ideas for what kind of interesting pursuits they want to direct their AIs at.”
08 / belief
“Obviously, the user might ask the model to do something that we think is actively harmful to other people.”
09 / belief
“On the other hand, I would also expect to gain a lot from doing practice at training time. So I think that you’d get the best results by combining these two things.”
10 / belief
“I believe the actual edge cases were explicitly stated rather than the kinds of things where are obvious.”
11 / belief
“Stepping back from 3.5, I think I heard you say somewhere that you were super impressed with GPT-2.”
12 / uncertainty
“I don’t know if it’s a phase transition but I might expect the same out of models where there might be some capabilities that work at multiple scales.”
13 / recommendation
“I want to go back to the point you made earlier about how this process could be more sample efficient because it could generalize from its pre-training experiences of how to get unstuck in different scenarios.”
14 / prediction
“I expect the mental model of a helpful assistant or helpful colleague to become more real.”
15 / preference
“On the other hand, I think I heard you make the point that a lot of our preferences and values are very subtle, so they might be best represented through pairwise preferences.”
16 / evaluation
“We had browsing as another feature in it although we ended up deemphasizing that later on because the model's internal knowledge was so good.”
17 / evaluation
“I would say there's a decent amount of room for variation in exactly how you do the training process.”
18 / prediction
“Because there doesn't seem to be a model since GPT-4 that seems to be significantly better, there's a hypothesis that we might be hitting some sort of plateau.”
19 / evaluation
“There are some algorithms that work this way, like mixture models or multiplicative weight update algorithms, where you have—I don’t want to say mixture of experts because it means something different—basically a weighted combination of experts with some learned gating.”
20 / evaluation
“We found that a tiny amount of data did the trick, even when you mixed it together with everything else.”
21 / evaluation
“To some extent, putting a lot of stuff into context will take you pretty far because we have really long context now.”
22 / prediction
“I realized a lot of the things that people thought were flaws in language models, like blatant hallucination, could be not completely fixed but things that you could make a lot of progress on with pretty straightforward methods.”