First of all, it’s hard to imagine anyone slowing down AI research. Even a ban by n large governments would only encourage research in places beyond government control.
Second, there is quite a lot of uncertainty about the difficulty of these tasks. Both human-comparable software and AI value alignment probably involve multiple difficult subproblems that have barely been researched so far.
First of all, it’s hard to imagine anyone slowing down AI research. Even a ban by n large governments would only encourage research in places beyond government control.
Second, there is quite a lot of uncertainty about the difficulty of these tasks. Both human-comparable software and AI value alignment probably involve multiple difficult subproblems that have barely been researched so far.
One could slow it down by convincing people who would otherwise speed it up.