The Dutch children’s programme Klokhuis has an interesting series on the brain. We learn all about the major topics in brain research, which both shows that Klokhuis is master in explaining things in an accessible way and that the brain is still an extremely trending topic, pushed in particular by the flourishing of cognitive neuroscience studies in recent years.
The latest episode, however, deals with Artificial Intelligence: or, what is the difference between the human brain and the computer brain? A small test: what is the longest river in the world?, asks the researcher. Klokhuis presenter says: Amazone. The researcher Googles and says: no, you are wrong, it is the Nile. An interesting form of testing, where the performance of both contestants (the computer and the human) is tested against the knowledge of the computer (who says that the Nile is the longest river? How can we know?). Of course it was a human, at some point, who knew that the Nile was longer, and put it into a Wikipedia entry somewhere on the web, so that nobody would forget.
It suddenly struck me that comparing computer intelligence and human intelligence is essentially a category error. It is like asking the question: who’s the better carpenter, a carpenter or a hammer? Huh? The carpenter does his carpenting *with* the hammer. It is a tool that enables him to be the carpenter. Just so, are computers part of what makes us intelligent. They form a system, not opponents. Why is this obvious for hammers, but keeps on generating reasoning errors by Cognitive Scientists when it comes to computers?