Those of us who are old enough (and geeky enough) remember Julia and other early 'chatbots', computer programs that act like people in chat rooms. Or try to act like people, at least. The next generation of chatbots is here, powered by neural networks and given the task of protecting the unwary from the unsafe. Judging from the transcript in this message, they will be very difficult to detect (though if they're online 24-7, that would be a dead giveaway). One wonders, will the chatbots ever get good enough to teach? More. Via Dave Green's NTK.
Today: 4 Total: 85 [Direct link] [Share]
