Pinned post
The Frontier Red Team at @AnthropicAI just made the case that you should be able to check which AI model answered you, without ever seeing its weights. Their research names DeepProve “the current peer-reviewed state of the art.”
LAGRANGE
This account is verified because it's an official organization on X. Learn more@lagrangedev
DeepProve: a cryptographic receipt for every AI inference. now open source.
Halo: autonomous coordination engine for the toughest conditions.
@LagrangeFndn $LA
Halo: autonomous coordination engine for the toughest conditions.
@LagrangeFndn $LA
379 Following 94K Followers 3.8K Posts
Latest posts 5
It's fixable: a proof that comes with every answer shows exactly which model produced it.
Which AI tool at work would you want to check first?
Which AI tool at work would you want to check first?
notion’s ai chat transcripts got exposed through a provider most of its users never chose. i think every ai tool should have to tell you which company actually ran your prompt.View quoted post on X
Last week, we asked four AI apps which model they are. Right now you just have to take their word for it. Ask yours and post what it says.
If a routing service quietly swapped the model answering you for a cheaper one, nothing in the reply would tell you. Anthropic's red team used that exact example last week to make the case for ZK.
What @AnthropicAI built and where DeepProve fits ⬇️
What @AnthropicAI built and where DeepProve fits ⬇️
A player jammed our swarm 389 times over Los Angeles, for more than ten minutes. It lost no drones and finished all eight sites 30 seconds after the jamming stopped. How long can you hold it?
holdtheswarm.lagrange.dev
holdtheswarm.lagrange.dev
0:27