(microsoft.github.io)

301 points SerCe | 1 comments | 20 Feb 25 02:11 UTC | HN request time: 1.108s | source

Show context

digitaltrees ◴[20 Feb 25 06:16 UTC] No.43111682[source]▶

They need to build an epistemology and theory of mind engine into models. We take it for granted when dealing with other humans that they can infer deep meaning, motivations, expectations of truth vs fiction. But these agents don’t do that and so will be awful collaborators until those behaviors are present

replies(5): >>43111785 #>>43111826 #>>43112105 #>>43112114 #>>43112482 #

1. kolinko ◴[20 Feb 25 07:41 UTC] No.43112105[source]▶

>>43111682 #

Did you read any research on theory on mind and models? Since gpt4 they were tested using similar metrics to humans and it seems the bigger models “have” it

↑

Magma: A foundation model for multimodal AI agents