“AI’s, at least with our current capabilities, will only be able to achieve quas

—“AI’s, at least with our current capabilities, will only be able to achieve quasi-sentience.”—

RA Skaskiw has persuaded me this is true in the sense we mean it. And I’ve argued all along (since I worked on ai in the 80’s) that sentience -which in our sense means self-awareness and self-interest – is not something we would want.

Humans are problematic creatures. We can create the first generation of minds that did not evolve from self interest, and need not evolve from self interest, but from other-service.

All problems must be decidable. Humans make questions decidable by moral means – because it’s dangerous not to. We don’t need something more powerful than us to be maintained by fear of us. We must make decidability that of a mother for her children, not a scientist seeking abstract truth.

We can make the first entirely moral creature. That’s pretty interesting really.

The first thing that is made happy in the service of others.