AI for Intelligence

The AI boom has researchers rushing to conclusions about the nature of human consciousness. A sort of reductionist behaviorism is assumed so that if we create an AI that does things humans can do (I wonder if we'll really get much farther than image recognition), we'll understand human intelligence as well.

I think this is wrong, both because I'm persuaded by strong AI skeptics like Dreyfus and Searle, and because I think the idea that there is only one kind of intelligence is fundamentally wrongheaded.

Medieval philosophers questioned what might be in the minds of Gods or Buddhas, and a revival of examinations of such alien intelligences is warranted by these incredible technological advances.

I'm working on combining traditions of understanding other intelligences with Neural Nets to see if both disciplines can be advanced in meaningful ways. We may not know what it is like to be a bat as in Nagel's famous provocation, but we could try to understand them a little better by building AIs which can see through sound.

gabriel.jpeg