For four of the 20 years David Ferrucci worked for IBM, he led a team of more than 40 researchers and software engineers to create Watson, a computer with artificial intelligence that beat high-ranking Jeopardy! champions on national television in February 2011.
The machine used to run Watson was the size of 10 refrigerators, with a memory containing two million books, all to take on the brain power of two competitors, Ferrucci said.
About 100 people, including young students and older instructors, some who brought their school-age kids, filled a classroom in SPSU’s School of Computing and Software Engineering for Ferrucci’s presentation, “In and Out of Jeopardy.”
Jasmine Watkins, 38, a business student who will graduate in a year, said although SPSU is a small, family-oriented school, the college is succeeding at staying relevant with new technology.
Watkins, who is a full-time administrative assistant in SPSU’s information technology department, said she came to the presentation “to see what all the fuss is about.”
Ferrucci said artificial intelligence is created by engineering a computer system to perform tasks that require interpretation and reasoning.
For instance, Watson not only had to understand what a Jeopardy! question was asking and give the correct answer, but also select categories, choose if to buzz in, and make the best wager, based on probabilities, for the Daily Double.
Josh Skelton, 31, who is studying computer game design and has one semester left before graduating, hopes to attend the University of Georgia to earn a master’s degree in artificial intelligence.
Skelton said the point of artificial intelligence is to create a program that is not just a database of information, but will mimic human functions.
These human elements in gaming often involve the automated system making mistakes or being unpredictable.
One of the largest hurdles for the Watson team was mastering the English language, which is a large index of words that are open to interpretation, especially with phrases and slang, Ferrucci said.
Science fiction or real-life problem solver?
Ferrucci led off the presentation by asking the room, “How many people think computers will take over the world?” Then added, “if you do, will those robots be benevolent?”
While talking about artificial intelligence, Skelton described machines that without being controlled by humans are able to track, target and fire weapons on each other. And camera systems can be trained to know the difference between an apple and an orange.
But Skelton chooses not to see a possible dystopian future that is frightening, like the movie the Terminator.
Skelton believes artificial intelligence will be used to make society better, and said computer games will have a greater possibility to be learning tools.
Ferrucci said he predicts that artificial intelligence will prove to be beneficial.
“It is only technology that advances the quality of life,” Ferrucci said.
Ferrucci said the positive impact would be felt worldwide, especially when patterns based on factual statistics are used in economic predictions.
“Economics affect everyone, everywhere,” Ferrucci said. “Computers are not swayed by politics, only on the impact the policy would have.”
But, Ferrucci said the benefits will only happen with human and machine collaboration, with the computer solving elaborate mathematic problems.
Ferrucci brushes off the fact that millions of factory workers have had their jobs eliminated, replaced by robots.
“If we get the computer to do more of the grunt work, we can be more open to be creative,” Ferrucci said.
Watkins said artificial intelligence has a large potential to solve transportation issues, especially with cars that self-correct if a driver falls asleep or is about to drive into a road hazard.
Ferrucci agreed to this benefit, joking during the presentation that “we are learning the weakest link in driving is the human.”