Evidence receipt / prediction
Published · transcript-backedIlya Sutskever: prediction
27 Mar 2023 Dwarkesh Podcast Ilya Sutskever (OpenAI Chief Scientist) — Why next-token prediction could surpass human intelligence
“Rather than achieving one mathematical definition, I think we will achieve multiple definitions that look at alignment from different aspects.”
Source trail
Everything needed to verify it.
- Speaker
- Ilya Sutskever
- Attribution
- Verified speaker
- Claim type
- prediction
- Recorded
- 27 Mar 2023
- Publisher
- Dwarkesh Podcast
Transcript context
…Let's talk about alignment. Do you think we'll ever have a mathematical definition of alignment? A mathematical definition is unlikely. Rather than achieving one mathematical definition, I think we will achieve multiple definitions that look at alignment from different aspects. And that this is how we will get the assurance that we want. By which I mean you can look at the behavior in various tests, congruence, in various adversarial stress situations, you can look at how the neural net operates from the inside. You have to look at several of these factors at the same time. And how sure do you have to be before you release a model in the wild? 100%? 95%?…
Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.