Today's Reading
Indeed, the discrepancies between artificial intelligence and human intelligence are nothing short of perplexing. Why is it that AI can crush any human on earth in a game of chess but can't load a dishwasher better than a six-year-old?
We struggle to answer these questions because we don't yet understand the thing we are trying to re-create. All of these questions are, in essence, not questions about AI, but about the nature of human intelligence itself—how it works, why it works the way it does, and as we will soon see, most importantly, how it came to be.
Nature's Hint
When humanity wanted to understand flight, we garnered our first inspiration from birds; when George de Mestral invented Velcro, he got the idea from burdock fruits; when Benjamin Franklin sought to explore electricity, his first sparks of understanding came from lightning. Nature has, throughout the history of human innovation, long been a wondrous guide.
Nature also offers us clues as to how intelligence works—the clearest locus of which is, of course, the human brain. But in this way, AI is unlike these other technological innovations; the brain has proven to be more unwieldy and harder to decipher than either wings or lightning. Scientists have been investigating how the brain works for millennia, and while we have made progress, we do not yet have satisfying answers.
The problem is complexity.
The human brain contains eighty-six billion neurons and over a hundred trillion connections. Each of those connections is so minuscule—less than thirty nanometers wide—that they can barely be seen under even the most powerful microscopes. These connections are bunched together in a tangled mess—within a single cubic millimeter (the width of a single 'letter' on a penny), there are over 'one billion' connections.
But the sheer number of connections is only one aspect of what makes the brain complex; even if we mapped the wiring of each neuron we would still be far from understanding how the brain works. Unlike the electrical connections in your computer, where wires all communicate using the same signal—electrons—across each of these neural connections, hundreds of different chemicals are passed, each with completely different effects. The simple fact that two neurons connect to each other tells us little about what they are communicating. And worst of all, these connections themselves are in a constant state of change, with some neurons branching out and forming new connections, while others are retracting and removing old ones. Altogether, this makes reverse engineering how the brain works an ungodly task.
Studying the brain is both tantalizing and infuriating. One inch behind your eyes is the most awe-inspiring marvel of the universe. It houses the secrets to the nature of intelligence, to building humanlike artificial intelligence, to why we humans think and behave the way we do. It is right there, reconstructed millions of times per year with every newly born human. We can touch it, hold it, dissect it, we are 'literally made of it', and yet its secrets remain out of reach, hidden in plain sight.
If we want to reverse-engineer how the brain works, if we want to build Rosey, if we want to uncover the hidden nature of human intelligence, perhaps the human brain is not nature's best clue. While the most intuitive place to look to understand the human brain is, naturally, inside the human brain itself, counterintuitively, this may be the 'last' place to look. The best place to start may be in dusty fossils deep in the Earth's crust, in microscopic genes tucked away inside cells throughout the animal kingdom, and in the brains of the many 'other' animals that populate our planet.
In other words, the answer might not be in the present, but in the hidden remnants of a long time past.
The Missing Museum of Brains
'I have always been convinced that the only way to get artificial intelligence to work is to do the computation in a way similar to the human brain.' —GEOFFREY HINTON (PROFESSOR AT UNIVERSITY OF TORONTO, CONSIDERED ONE OF THE "GODFATHERS OF AI")
Humans fly spaceships, split atoms, and edit genes. No other animal has even invented the wheel.
Because of humanity's larger résumé of inventions, you might think that we would have little to learn from the brains of other animals. You might think that the human brain would be entirely unique and nothing like the brains of other animals, that some special brain structure would be the secret to our cleverness. But this is not what we see.
What is most striking when we examine the brains of other animals is how remarkably 'similar' their brains are to our own. The difference between our brain and a chimpanzee's brain, besides size, is barely anything. The difference between our brain and a rat's brain is only a handful of brain modifications. The brain of a fish has almost all the same structures as our brain.
These similarities in brains across the animal kingdom mean something important. They are clues. Clues about the nature of intelligence. Clues about ourselves. Clues about our past.
Although today brains are complex, they were not always so. The brain emerged from the unthinking chaotic process of evolution; small random variations in traits were selected for or pruned away depending on whether they supported the further reproduction of the life-form.
...
