I used to look forward to his takes. Some of the past posts were genuinely insightful, but now all I hear is the cliched difficult road leading to an AGI whose consequences always seem utterly dire for anyone involved, perhaps except OpenAI.