16 September 1999, 794 words
If you follow science news, you might be forgiven for thinking that the human race will soon be ruled either by genetically-engineered supermen, or intelligent computers, or by some cyborg combination of both. One recent headline trumpets that scientists have created super-intelligent mice. Another proclaims that computers are more creative than people. They're building an artificial lifeform; they're making computer chips that can be implanted in the brain; they can clone human beings.
News reports about science tend to be generally accurate, but you often have to suss out opinion from fact. Scientists are cautious people, whose favourite words are "probably", "perhaps" and "maybe". But the public likes certainties and, even when the reporter includes the scientists' hedging, readers often edit out it out of their own minds. And the most egregious science reportage almost always involves issues connected with intelligence.
Take the smart mice story. The Reuters report begins: "Scientists in the United States said they had created a strain of smart mice by inserting a gene into the rodents' brains, proving it possible to improve the intelligence of mammals, including humans." I'm willing to lay odds that no scientist ever said anything about improving human intelligence. What they may have said was that the gene insertion technique may be used for treatment of certain forms of mental illness, such as Alzheimer's or dementia.
The article went on to say, "The study revealed that a common biochemical function was at the centre of all learning." Now even a merely intelligent layperson will immediately start smelling rotten carite. The procedure described involves the insertion of a single gene called, in that quirky way scientists have, NR2B. Once inserted in the mice, the NR2B gene keeps open a brain receptor longer, allowing a memory connection to be formed more quickly. So the mice learned to go through the maze faster than their non-altered brethren.
To which I say: Big hairy deal. Learning to go through a maze, while no doubt good clean fun, will not change the world. And common sense alone tells us that a complex activity like learning requires more than just a good memory. The absent-minded professor, like all stereotypes, has some basis in reality, and everyone has read stories about idiot savants, like the autistic character played by Dustin Hoffman in the movie Rainman, who can remember all sorts of trivial facts but are quite retarded.
If, like me, you're a layperson who reads a fair amount about the mind, the story becomes even more dotish. The brain has no single memory system, but several types of memory agencies that work in different ways to suit particular purposes. Brain damage provides evidence of this. Injury to one part of the brain causes loss of names; another kind of injury results in an incapacity to recognize faces; there are even cases of people forgetting past tenses. But, in any case, memory is no great feat compared to other kinds of intelligent behaviour. Even prodigies don't construct long-term memories more than two to three times faster than the average person (who generally takes about half-hour to do so).
As for computers becoming smart, that's even more absurd. When the Deep Blue computer beat world chess champion Garry Kasparov in 1997, the media hype almost suggested that the End of Mankind was just around the corner. The scientists knew better, of course. "Chess is a computational game governed by specific rules on a finite board," wrote mathematician Roger Penrose. "Computational power is not the same thing as intelligence...Tasks that a child can perform effortlessly, such as reacting to an emotion or understanding the spoken word, are outside the ability of current computers...There are fundamental reasons from our knowledge of mathematics to believe that the quality of human understanding lies beyond the scope of any computer...computers will never achieve pre-eminence in areas such as politics or the arts, where human sensitivity have fundamental importance."
But we needn't even get so complicated. Artificial Intelligence researchers have a basic axiom: The hard things are easy and the easy things are hard. It's easy to program a computer to calculate the vector of a descending Jumbo jet; it's hard to teach it that when a woman comes home from work, she brings her head with her.
So we needn't worry about cyborgs taking over the planet. Computers are still dumb and, while the NR2B gene may help cure brain diseases, it's hardly likely to let us bio-engineer geniuses. It mightn't even prove that more efficacious than natural methods - there's evidence that some brain diseases may be prevented if we live mentally active lives. On the other hand, as Bertrand Russell once wrote, "Many people would die sooner than think; in fact, they do."
Copyright ©1999 Kevin Baldeosingh