The main thesis seems to be that the only ways that AI can go wrong is that they stop doing what we want, just like humans might do. Views attributed to Eliezer and the AI alignment community as a whole are strawmen of the most flagrant variety.
This article is trash.
The main thesis seems to be that the only ways that AI can go wrong is that they stop doing what we want, just like humans might do. Views attributed to Eliezer and the AI alignment community as a whole are strawmen of the most flagrant variety.
This article is trash.