Tamsin Leake comments on Which investments for aligned-AI outcomes?

[Tamsin Leake]

I think it’s exceedingly unlikely (<1%) that we robustly prevent anyone from {making an AI that kills everyone} without an aligned sovereign.