nafyi.

LA CONVERSACIÓN ORIGINAL

AI researchers debate how close we are to recursive self-improvement

Dwarkesh Podcast · · 1:37:01

John Schulman discusses uncertainty about generalization, improving experimental methods, and the role of realistic prompt distributions in model distillation.

John Schulman
Abrir Audio del editor (se abre en una pestaña nueva)
AI researchers debate how close we are to recursive self-improvement
Dwarkesh Podcast

De un vistazo

Momentos clave3

Fragmentos breves y atribuidos, con el contexto necesario para verificarlos. La conversación completa permanece con su editor.

9:42 (se abre en una pestaña nueva)
AI research

Generalization is difficult to predict

Extracto original

The whole field relies a lot on generalization and it’s very hard to predict when you’re going to get generalization, or when you’re going to get some kind of out-of-distribution generalization
John Schulman·

Este fragmento no incluye más contexto. Consulta la conversación original.

11:34 (se abre en una pestaña nueva)
AI research

Research methods may have substantial room to improve

Extracto original

If you think hard enough, you probably could have expected some of these things beforehand. There is probably some very clever way to do a small-scale experiment that’ll let you build the theory that then will generalize to the large-scale experiment. So I would expect that we’re nowhere near the ceiling of how well you can do research
John Schulman·

Este fragmento no incluye más contexto. Consulta la conversación original.

26:48 (se abre en una pestaña nueva)
AI research

Distillation depends on a realistic prompt distribution

Extracto original

If you have a really good realistic prompt distribution for distillation, you can match the big model really well. But if you only have this distribution of easily verifiable tasks, then you can match the big model on all the benchmarks, but you do worse on this broader distribution
John Schulman·

Este fragmento no incluye más contexto. Consulta la conversación original.

Fuente y metodología

Estas perspectivas enlazan a sus fuentes originales. Las paráfrasis están identificadas y no son citas textuales.

Abrir transcripción o material de origen (se abre en una pestaña nueva)Reportar un problema