Doubling Down on AI-skepticism
This oped is a second salvo fired at the prevailing AI mania. In the first I had attacked its magical great leap forward thinking, calling LLMs mendacious talking horses.
 |
Artificial intelligence is already a mature enterprise, Amar Bhidé writes.
|
This oped, published in Barrons digs into the statistical models that make them so.Writing a book that Oxford will publish in December has contributed to my long standing skepticism about backward looking models that ignore change and context.
Trying to use LLMs to research, edit, and illustrate the book was a source of unending frustration (though “earlier” AI was invaluable).
I also learned that AI grew out of a “fork” in the cognitive revolution of the 1950s and 1960s which conceived of the mind as a computer. A second fork treated the mind as a “meaning constructor” where meaning was highly contextual, historical and cultural.
Both forks have value. But reducing all thought and speech to a mindless statistical model is absurd.
(ps. Email me if you’d like to look at the nearly final proofs of the book and don’t hesitate to pre-order on Amazon!)