[M]ost artificial neural networks are decidedly un-brainlike, in part because they learn using mathematical tricks that would be difficult, if not impossible, for biological systems to carry out. Yet brains and AI models do share something fundamental in common: Researchers still don’t understand why they work as well as they do.
What computer scientists and neuroscientists are after is a universal theory of intelligence—a set of principles that holds true both in tissue and in silicon. What they have instead is a muddle of details.
It’s possible that AI models don’t need to mimic the brain at all. Airplanes fly despite bearing little resemblance to birds. Yet it seems likely that the fastest way to understand intelligence is to learn principles from biology. This doesn’t stop at the brain: Evolution’s blind design has struck on brilliant solutions across the whole of nature. Our greatest minds are currently hard at work against the dim almost-intelligence of a virus, its genius borrowed from the reproductive machinery of our cells like the moon borrows light from the sun. Still, it’s crucial to remember, as we catalog the details of how intelligence is implemented in the brain, that we’re describing the emperor’s clothes in the absence of the emperor.