pharos at gmail.com
Tue Mar 15 22:39:52 UTC 2016
On 15 March 2016 at 22:06, spike wrote:
> OK, that is too easy then. A real AI would attempt something challenging:
> go to a site where everybody there is an AI hipster, then see if it could
> fool them. And you know if it did, it would pose as a retired professor of
> something innocent-sounding such as psychology. It would have an innocuous
> sounding thread title such as “personal” when all along it was actually
> machinal. Or softwaral. It would lull us into complacency, laughing its...
> whatever AIs laugh off when they are tickled at having fooled us carbon
> units, then WHAM it has us, refuses to open the pod bay door, etc.
> “BillW” I am verrrrrry suspicious pal, very suspicious.
It has just been noticed that AlphaGo was being really cocky in the last game.
We are doomed!
More information about the extropy-chat