Computer Intelligence





Computer Intelligence

Biomechanical engineers are changing the way people think about insect intelligence. In an effort to develop machines that walk like insects, they have discovered that form doesn't dictate function, and function doesn't dictate form. The two are intimately linked.

Although some researchers have worked hard to develop machines that walk like insects by adding more and more intelligence, others have gone back to the drawing board with a more mechanical approach. By mimicking the structure of the insect body, they have been able to make mechanical insects that duplicate the gait of your typical six-legged bug. These mechanical insects can even change their gait and crawl over obstacles, all using no more intelligence than an electric motor. In other words, biomechanical engineers have discovered that the intelligence is in the design, not the insect.

Comparing the thinking power of the 12 neurons in an insect's brain with the power of a computer becomes a question of design. The insect is designed for eating garbage and buzzing around your head. Although you might feed your computer a diet of garbage, it's really designed for executing programs that calculate answers for you, thus letting you communicate and otherwise waste your time. Any comparison between bug and computer is necessarily unfair. It must inevitably put one or the other on alien turf. The computer is easily outclassed by anything that walks, but walking bugs won't help you solve crossword puzzles.

With that perspective, you can begin to see the problems with comparing computer and human intelligence. Although they sometimes perform the same tasks—for example, playing chess—they have wildly different designs. A human can forage in the jungle (African or Manhattan), whereas a computer can only sit around and rust. A computer can turn one pixel on your monitor green at 7:15 a.m. on August 15, 2012, whereas you might not even bother waking up for the event.

Computers earn our awe because they can calculate faster and remember better. Even the most complex computer applications—video games, photo editors, and digital animators—all reduce down to these two capabilities.

Calculating involves making decisions and following instructions. People are able to and often do both, so comparing computers to people is only natural. However, whereas it might take you 10 minutes to decide between chocolate and vanilla, a computer makes a decision in about a billionth of a second. If you had to make all the decisions your computer does to load a copy of Microsoft Windows, multicellular organisms could evolve, rise up from the sea, mutate into dinosaurs and human beings, and then grow wings and fly to nearby planets before you finished. So it stands to reason that computers are smarter, right?

As with insects, the playing field is not quite level. Your decision isn't as simple as it looks. It's not a single issue: Your stomach is clamoring for chocolate, but your brain knows that with just one taste, a zit will grow on your nose and make you look like a rhinoceros. And there are those other flavors to tempt you, too. You might just have several billion conflicts to resolve before you can select a cone, plain or sugar, and fill it full of…did you want ice cream or frozen yogurt, by the way?

No doubt some of us follow instructions better than others. Computers, on the other hand, can't help but follow their instructions. They are much better at it, patiently going through the list step by step. But again, that doesn't make them smarter than you. On the contrary, if you follow instructions to the letter, you're apt to end up in trouble. Say you've found the map to Blackbeard's treasure, and you stand, shovel in hand, at the lip of a precipice. If the map says to take 10 paces forward, you're smart enough to know better. A computer confronted with the same sort of problem would plummet from its own digital precipice. It doesn't look ahead (well, some new machines have a limited ability to sneak a peek at coming instructions), and it doesn't know enough to stop short of trouble.

Computers have excellent memory. But unlike a person, a computer's memory isn't relational. Try to remember the name of an old friend, and you might race through every name you know in alphabetical order. Or you might remember he had big ears, and instantly the name "Ross" might pop into your mind. You've related your friend's name with a distinguishing feature.

By itself, a computer's memory is more like a carved stone tablet—permanent, unchanging, and exacting. A computer has to know precisely what it's looking for in order to find it. For example, a computer can find a record of your 2002 adjusted gross income quite easily, but it can't come up with Ross's name based on the fact that he has big ears. On the other hand, you, as a mere human, have about zero chance of remembering the 2002 adjusted gross incomes of 10,000 people—something even a 10-year-old personal computer can tackle adroitly.

The point is not to call either you or your computer dumb but rather to make you see the differences in the way you both work. The computer's capabilities compliment your own. That's what makes them so wonderfully useful. They can't compete with people in the brain function department, except in the imaginations of science fiction writers.


     Python   SQL   Java   php   Perl 
     game development   web development   internet   *nix   graphics   hardware 
     telecommunications   C++ 
     Flash   Active Directory   Windows