Thursday, March 24, 2016

The Thao Worra Test on artificial intelligence

In light of the recent incident with the Tay A.I. chatbot that Microsoft tried to deploy that went pro-Hitler in less than 24 hours, I hereby proclaim this the Thao Worra Test. How long can your AI interact with humans and NOT turn "evil". I know. Some might say this isn't as important as the Turing Test or the Voight-Kampff test, but indulge me.

No comments: