@VraserX: I worry that the most extreme AI-doom rhetoric within effective altruism could help radicalize future terrorists. Convi…

I worry that the most extreme AI-doom rhetoric within effective altruism could help radicalize future terrorists. Convince someone that AI will kill everyone, that its developers are knowingly causing it, and that governments won’t intervene, and violence can start to look like self-defense. A radicalized minority could see attacking data centers or targeting AI CEOs and researchers as saving humanity. That possibility worries me even though it doesn’t describe the intentions of the wider movement. You don’t have to explicitly call for an attack to help someone rationalize one. Anyone making these arguments should be unequivocal that their fears do not justify violence.