Google DeepMind: The Podcast cover image

Speaking of intelligence

Google DeepMind: The Podcast

00:00

What's My Worst Model?

Geoffrey irving is a safety researcher at deep mind. He spends his time typing messages to language models and being flattered by its responses. In order to understand just how they work, let's consider how they work on the level of a single sentence. For example, say a language model is trying to complete the sentence, the ice cream made me blank. What might that missing word be? Happy? Angry, zilophone?"

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