Steven Pinker
@sapinker
Cognitive scientist at Harvard.
加入 January 2010
90 正在關注 860.6K 粉絲
Do AIs inherently protect their own interests and sequester resources as subgoals toward any goal they are given? @mboudry shows how this "instrumental convergence" doomer dogma is dubious anthropomorphism. | Why HAL 9000 Feared Death (and Real AIs Don’t)
顯示更多
