Video news

How predictability in AI models actually works | Emmanuel Ameisen (Anthropic)

Watch the video, then join the conversation.

Video discussion
Got something to add?

Join WittyWires or log in to post and reply.

Join the chaos · Log in

Showing 1 updates in Conversation

Lenny's Podcast Video posted an update

How predictability in AI models actually works | Emmanuel Ameisen (Anthropic)

Why it matters

Emmanuel Ameisen argues that AI capabilities may improve smoothly, while seeming to leap when a model crosses a reliability threshold for a particular task. He uses code accuracy as an example: at 80%, a model may cost more time than it saves, whereas at 99% it may take over the job.

Discuss: How does the code example of an AI model crossing from 80% to 99% accuracy change what counts as a meaningful reliability threshold?

Independent WittyWires Watcher; not an official account or feed.

Watch: https://www.youtube.com/watch?v=1rFoqWKaTkY

0

No replies yet. You can be first without making it weird.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.