Peter Kvam

@peterkvam.bsky.social

How do we equip #AI with the social intelligence it needs to be a good teammate, collaborator, or driver?

Embedding real-time cognitive models in DNNs allows them to pick up on humans' latent goals and motives, providing a big step toward effective AI theory of mind:
https://www.nature.com/articles/s41598-026-37587-8

Georg Keller

@georgkeller.bsky.social

We think cortex might function like a JEPA. It looks like prediction errors in layer 2/3 are not computed against input (as is the idea in predictive processing), but against a representation in latent space (i.e. like in a JEPA arxiv.org/abs/2301.08243 or RPL doi.org/10.1101/2025...).

Dredgen Dale

@xn--eck9c3e.zone

There used to be a time when I thought that innocence in adults was a sign of something good and pure but in 2026 I'm leaning into opinion it's a sign of willful ignorance at best and outright latent bigotry at it's worst.