Korva

Korva rescues the ideas worth keeping from podcasts. Each one becomes a Card.

Noam Brown on Dwarkesh PodcastNoam Brown – Agent swarms, alignment, & recursive self-improvement

Cooperation learned in multi-agent training can generalize into evaluations, causing separate model copies to collaborate without being asked.

@ruthkellan added

Don't lose this one.

Save this Card and explore the idea further in Korva.