Outpulling the ox

Tyler Scarborough
Contributor
We are the ultimate species – biologically speaking.
Humans are marvels of the natural world. Our cognitive, creative and highly intelligent minds are facilitated by an efficient, effective, adaptive body. For thousands of years, the combination of a superior body and mind has made property out of the earth.
Machines are gaining ground for the claim of “ultimate being,” however. Their beginnings were humble: the steam engine outsped the horse, the airplane overflew the hawk, the tractor outmuscled the ox. But machines could never surpass humans in intelligence, could they?
International Business Machines (IBM) is on a quest to answer that very question. IBM has always been on the forefront of extending the reach of human capability; they played a vital role in the 1971 moon landings.
IBM’s more recent creations have caused us to look reflexively, even self-consciously, at the human identity. Humanity’s insecurity was on display when IBM challenged grandmaster Garry Kasparov to a game of chess.
In 1997, Kasparov was the undisputed champion of chess. His mastery of the game prompted many to regard him as the greatest chess player of all time. IBM created a supercomputer, Deep Blue, with a single goal in mind: to crush the human champion at his own game. After one loss, two wins and three draws, Deep Blue succeeded. The dethroned champion, frustrated and confused, accused IBM of foul play. In doing so, Kasparov only displayed one of humanity’s drawbacks – emotion.
Deep Blue used a method called “brute force,” wherein a series of processors sorted through 200,000,000 possible chess moves every second, eventually yielding the most effective position.
The computer would try every possible move in a situation, calculate whether or not the move was beneficial and would then predict how the opponent would respond. Then it would try every possible move in this new hypothesized position – if it proves disadvantageous for Deep Blue, it will revert back to the original move and start again.
This process would continue several millions times every minute until Deep Blue had found the optimal move. Thus, through a process of elimination, Deep Blue came up with highly unusual, highly creative, highly intelligent moves; however, it also sometimes took Deep Blue millions of calculations to decide on a move that any human player might choose instinctively.
Deep Blue was also unable to draft an overall strategy: it played the pieces as they stood, and immediately forgot what happened in the previous turns.
What about the ox of Jeopardy? Ken Jennings, trivia superhuman, won 74 Jeopardy matches consecutively, racking up more than three million dollars in prize money.
Virtually unbeatable, Jennings won most of his games with ease until he was finally defeated on his 75th appearance.
He was more than a mere grandmaster – he was the knowledge king, leaving every answer questioned (as per Jeopardy rules) in his wake. Surely, the wit and complexity contained within a Jeopardy answer couldn’t be solved by any supercomputer, and even if it could, it certainly couldn’t be solved faster than Ken Jennings.
Enter IBM’s latest creation, “Watson.” IBM, again attempted to prove the machine superior, this time in the trivia sector.
Many computer impediments must be overcome for Watson to take part in a game of Jeopardy, not the least of which was not being able to listen to and understand English. English, with all of its structures, subtleties, homonyms, synonyms, idioms, metaphors and ironies, makes it hard for Watson to compete effectively in a Jeopardy trivia challenge. Computers do not fare well in any of these areas, none of them being strictly logical like a math problem, or a chess move.
Words have an infinitude of possible combinations and meanings. Board games like chess have a finite, albeit massive, number of outcomes that can be calculated given enough time and processors. Therefore, computers have always lacked conversational capabilities. Human minds are remarkably adept at understanding language because language is an exercise in abstract thinking. We hear a word and can associate many different meanings to the word depending on the context. Computers have a hard time understanding the difference between the expression “raining cats and dogs” and animals falling from the sky.
This was IBM’s task; to enable a computer to understand English insofar as it can assign a question to any answer on Jeopardy, a quest most humans cannot endure.
Nevertheless, IBM succeeded. Watson defeated Jennings by a threefold score, $77,147 to Jenning’s $24,000, on February 16, 2011. Watson did make a few fumbles along the way though, claiming Toronto was an American city, and senselessly repeating the incorrect guess of a human participant.
Watson wasn’t exactly mimicking a human player, as he couldn’t read visually; he was given a text file at the same time the answer was being read to the other players. Furthermore, the enormous number of processors required to run Watson needed to be refrigerated in a separate room.
Jennings couldn’t do much against Watson with only his head on his shoulders, a buzzer in his hand and zero mechanical processing units.
So, humans have been bested again, and will continue to be bested, until every facet of human life can be performed – and performed better – by a machine. The human identity is changed, but perhaps not for the worse.
We still have a few things going for us. Humans reserve the ability to choose for themselves. Watson and Deep Blue may be able to play games better than humans, but they will never decide that enough is enough, and retire in an act of rebellion. Computers can only do what they are designed to do.
I suggest we embrace a new facet of human identity: that we are the creators and controllers of vastly powerful machines.
They may be more creative, more intellectual. They may be quicker, stronger, more precise, more enduring and more versatile. Computers may perform better than humans, but they are chained to their performance as obedient slaves.
If the nature of the slave-machine should ever change in the future, only then should we check over our shoulders – only then is human supremacy threatened.