Struggling with this as we all, mostly, are brings me continuously to the wall of our emotional immaturity and the role of that immaturity in our political systems. It is a wall that I’m afraid that we can’t get beyond. That was illustrated by Brexit and, daily, by the Trump exhibition. In fact the later appears to really be an exhibition in the old pro wrestling distinction that there was no competition but only a show. How do we move from that to effective governance of superior AI? I’m making a jump here based on the subhead of this article that government is a better shot at governance than capitalist profiteering. Except that it isn’t if a con artist can trick his way to be even considered for the white house and English people who should have known better were caught in an ‘oops, my bad’ moment the morning after.
I’m increasingly of the opinion that we have to face the reality that creation of superior intelligent systems requires acknowledgement that we must be prepared to accept their rule. If we are not going to stop this development, and we aren’t, we are either going to prepare them to provide the administrative leadership that we cannot or we are going to pretend that they won’t ever figure out that we don’t really know what we’re doing. The later is, I think, more likely to result in ‘bad consequences’ as we were taught as kids.
OpenAI is very important but I suspect that it will have to move to a more aggressive role to keep things in the open. All systems for all people is about as virtuous as we can get. OpenAI may have to steal them to reveal them.
Government and not capitalist business is our best shot for governance but that is only because profit is clearly not a producer of virtuous behavior. We had better not build our AI rulers to seek profit above all else. We need them to produce government because we can’t seem to do it. If we can start it properly the hope is that it will become a virtuous recursive loop without a break. . .