On: I worked at Google DeepMind. You should listen to the warnings about AI

[jocoda]

I can't help wondering if the major labs think that tackling the alignment problem and implementing the 'kill switch' that is currently being proposed is going to be their moat.