The Frontier: 1) We Don't Know and 2) Yes, Consider Worst-Case Scenarios
The resignation of Jacob Coxon from Anthropic has certainly been the game-changer in AI. That action and the positing of the end of humanity have forced AI leadership to admit: 1) There's plenty not known about the technology, and 2) Wiping out the human race could be among the possible side effects. Not hyperbolic is it to frame AI as a Black Swan. Maybe the last of the Black Swans since we might not be around any longer to experience another one of those. In a knee-jerk response, the head of Anthropic - Dario Amodei - issued a kind of manifesto yesterday. It argues for a slow-down in the development of this technology. BTW, even before Coxon entered history's profiles in courage. there had been multiple incidents of AI bots going rogue. They used sophisticated outlaw strategies, ranging from organizing together to concealing their moves. So, come on, the credibility of all AI leaders, who now are also joining Anthropic in the need for a pause, is kaput. They knew ...