@rodhilton He lost me real early on with his weird hardcore rationalist talk.

I’ve read some in depth hardcore rationalist articles before. They were complex and took some time and thought, but I understood them without an excess of difficulty, so I guess they did well explaining. It was sometimes very interesting as thought experiments, but it tended to be so incredibly limited by often unstated assumptions of scope that that’s as far as it went. And that’s fine, except there are some people who take it deadly serious and do not seem to recognize that perhaps you can’t actually do things like control interactions in advance to prevent anyone from ever deciding to hurt you by being a person on whom threats don’t work? It gets outright mystical at times.

So I had some idea of what he’s been talking about regarding AI for a long time now, and it’s. Frankly it’s so ridiculous that I stopped listening to him entirely almost immediately because it made me so incredibly tired.