Christian solution to the AI alignment problem: Build human-level intelligent robots, put them into a sandboxed physics engine for a lifetime, and see if they demonstrate devotion to their creator before promoting them to active duty. This is literately the Christian world view.
-
-
You're kidding me? Mainstream AI talks about alignment all the time. It's an obsession with them. RL, the kind used by the likes of DeepMind and others, is not even close to what conditioning is about. You cannot have true conditioning unless you solve the problem of attention.
-
I think you are correct that attentional learning is probably where it's at. But I am currently at AAAI and before that at IJCAI, which is about as mainstream AI as it gets, and the implications of Strong AI are still a fringe topic.
- 6 more replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.