blog

von Neumann in the Chinese Room

I recently taught a course on the philosophy of artificial intelligence. For better or for worse, it's hard to do that without discussing Searle's Chinese Room argument. This argument, published in 1980, has enjoyed a tremendous amount of influence in conversations about the possibility of machine intelligence. The idea that computers are constitutively incapable of thought beacuse they "merely" push symbols around traces directly to Searle. And despite decades of vigorous critique by philosophers and cognitive scientists, embers of the argument smolder on. Foolishly, then, I throw my own hat in the ring with a new objection.

The Chinese Room argument is deceptively simple. We are asked to imagine Searle-an anglophone monoglot-locked in a room with a rule book and several stacks of Chinese symbols.