Cognitive Complexity Subject of Katz School Talk
“Computer programs aren’t brilliant. Programmers are brilliant.”
That was the message presented by Dr. Viktor Dorfler, senior lecturer in information and knowledge management at the University of Strathclyde in Scotland in his February 5 presentation on YU’s Isreal Henry Beren Campus, “Big Data, Artificial Intelligence and the Human Intuition,” sponsored by the Katz School. Because the speed at which computers can process complicated information outpaces humans by such an exponential degree people mistake complicated processing for complex thinking, he explained.
Computers learn through positive and negative reinforcement, where they catalog successes and failures, and the complicated processing happens through speed and math, not through higher-level thinking. In 1996, chess champion Gary Kasparov lost to IBM’s Deep Blue, but later came back to beat the computer. According to Dorfler Deep Blue was able to beat Kasparov by processing thousands or millions of previously played games, learning strategy by reinforcement, and then playing its own game without mistakes. Kasparov was able to adapt the beat Deep Blue later because of his ability to adjust to playing an opponent who does not make mistakes, in other words, playing perfectly is a strategy that can be defeated. Kasparov’s human intuition allowed him to create new strategy, whereas Deep Blue is only able to adapt strategy created by humans. “By reshuffling the existing knowledge according to the existing rules, computers can come up with new things but it’s unlikely they come up with interesting things,” said Dorfler.
Through his research in contrasting the ways computer programs learn and think to the ways humans, specifically Nobel Laureates and master chefs, do, Dorfler concludes that future technological achievements will be attained by smart people with the support of technology, rather than by technology alone.
Dorfler identified three ways that smart people think that contrast to the positive or negative reinforcement of computer thinking: imprinting, inspiration, and the master/apprentice relationship. Intuition is a person who can learn a new instrument or repair electronics without much effort. Smart people learn through intuition, spontaneity, creativity – three things a computer cannot replicate.
While technology taking over is a more interesting story, Dorfler said, the more likely scenario is that achievement will be done by smart technology in support of smart people.
The lecture was part of the Katz School’s master’s in marketing program. For more information on the program and upcoming events visit: www.yu.edu/katz/programs/graduate/ms-marketing.