Reminder that not all "experts" agree with the AI doom thesis. Yann LeCun thinks it's nonsense.
I don't think anyone really has "expertise" to perform AI doom prognostication, regardless of whether they created dropout or ReLU or the Transformer architecture or are a professor at Oxbridge/MIT or whatever else people take as an impressive credential. But even if we accept the logic of those who want to argue from the authority of expertise, we don't have expert consensus.
I don't think anyone really has "expertise" to perform AI doom prognostication, regardless of whether they created dropout or ReLU or the Transformer architecture or are a professor at Oxbridge/MIT or whatever else people take as an impressive credential. But even if we accept the logic of those who want to argue from the authority of expertise, we don't have expert consensus.