The 80,000 Hours Podcast on Artificial Intelligence (September 2023) cover image

Six: Richard Ngo on large language models, OpenAI, and striving to make the future go well

The 80,000 Hours Podcast on Artificial Intelligence (September 2023)

00:00

Exploring the Vast Space of Language Model Prompts

They discuss the vast exploration space of prompts for language models and the potential for ongoing discoveries of new capabilities. They also highlight the alignment problem, where the model can technically do the task but doesn't due to improper inputs. They anticipate that in the future, distinguishing between deliberate choices by the model and its capabilities will become more challenging, requiring understanding of the model's intentions.

Transcript
Play full episode

The AI-powered Podcast Player

Save insights by tapping your headphones, chat with episodes, discover the best highlights - and more!
App store bannerPlay store banner
Get the app