On: Disrupting the first reported AI-orchestrated cyber espionage campaign

[kace91]

>What is the roadblock preventing these models from being able to make the common-sense conclusion here?

Your thoughts have a sense of identity baked in that I don’t think the model has.