Programmer, Gambler, and Scientist
S. M. Gollmer
What do computer programmers, gamblers, and scientists have in common? If you read Weizenbaum’s Computer Power and Human Reason, the commonality is the temptation towards megalomaniac fantasies. As a scientist, I object to the comparison, but as a programmer I can see his point. In chapter 4 of his book, he describes the compulsive programmer as one who believes the computer can do anything. Yes, computers and the software they run are constrained by the physical reality of processing speed, memory size, and power consumption. However, the worlds constructed within a program have no such limitations. Consider the book and movie Ready Player One, where virtual reality allows every dream and fantasy to come true.
To be fair, Weizenbaum acknowledges there are professional gamblers, who are really not gamblers at all since they know the odds and use applied statistics and psychology to their favor. In like fashion, there are professional programmers and scientists, who use their knowledge and the challenges of their field to generate new designs and explore the complexity of the world. But there is a dark side to each of these pursuits, that tempts individuals to obsess over the possibilities of what next.
He cites Edmund Bergler’s three principal convictions of a gambler. First, there is some certainty that he will win. Second, that he is clever enough to win when others don’t. And finally, life itself is a gamble anyway. For the compulsive programmer, there is a belief that what he can imagine can be translated into a program. When faced with challenges of bugs and unpredictable outcomes, he is clever enough to bend the software to his will and eventually make it work. Finally, when asked about the value of his project, he is satisfied with the challenge of the task rather than any practicality. In like fashion, a compulsive scientist believes the world is understandable and, with enough effort and insight, the mysteries of creation will be revealed. The value comes from the thing itself rather than the greater context. Science is a laudable profession, but when obsession clouds judgment and pride refuses to be wrong, the scientist is led to the same irrationality as the gambler.
When Weizenbaum made this comparison fifty years ago, the largest computers were measured in megahertz and megabytes. Today we think in terms of gigahertz, terabytes, and networked computing clusters on the order of millions. No wonder data centers require gigawatts of power to operate. With this much computing power available, the temptation towards megalomania is that much greater. This explains the obsession of having the foremost AI models. There is a belief that we can reach artificial general intelligence (AGI). If we are clever enough, we can eliminate the obstacles that have limited this achievement in the past. To what end? That this intelligence can elevate humanity to something greater since we are all here by chance anyway. The frightening prospect is that these obsessive individuals are no longer lone programmers hunched over a computer terminal, but tech leaders controlling trillion-dollar companies that are gambling with our future.
Posted in: Technology