Friday, June 24, 2011

The Horrors of the Future

Friendly AI* probably won't be a problem if, as I suspect, it turns out to be substantially easier to enhance humans than it is to create intelligence from scratch.

*The problem of ensuring that the interests of any self-improving-come-super-intelligent AIs remain aligned to human interests. For fictional examples of failed FAI see any techno-dystopic movie e.g. Terminator, the Matrix, I Robot, etcetera.

No comments:

Post a Comment