Raw LLM Responses

You're saying they can't know that will work, which is correct. You're also saying limiting computer models computer power won't slow them down, which is incorrect. The correct thing to say is "we don't know how much it will slow them down. I.e. how much more efficient the models will become and at what rate, therefore we can't conclude that will be sufficient protection." I would also like to point out that raw compute power is literally the driver behind all of our machine learning/AI progress so far. It stands to reason that the biggest knob we can turn here is compute power.