If you were to stand on a street corner and ask 100 people to define Artificial Intelligence, odds are, you’d get 100 different answers. And, according to AI visionary and educational innovator Dr. Roger Schank, it’s likely every one of those answers would be wrong.
“Early attempts at AI were a lot about making computers smart by teaching them to play chess. It was a horrific mistake to define it that way,” he said. “What we got was artificial intelligence which is so confusing, people in the field are still confused as to what it might be about.”
Instead of making a computer look intelligent by teaching it to play chess, think 10,000 moves ahead and do things people could never do, Schank believes those computers should have been taught to play chess like a Grand Master chess player would play.
“I always saw AI as a field that could tell us more about people by getting us to figure out how to imitate people by doing the kinds of things people do,” Schank said. “Why does that matter? It matters because people talk to each other. That’s a sub-section of AI called ‘Natural Language Processing’ and it’s phenomenally hard.”
To illustrate his point, Schank cited the example of the typical intelligent, talking robot in the movies and pointed out that, while that robot may talk, it rarely asks questions.
“Does it have a point of view that’s new? Can it make an interesting argument with you? Does it have something it can teach you or you can teach it? These are the right questions for AI,” he explained. When a computer can stop and think, I’ll be very impressed and say, ‘Wow! We have AI!’”
For Schank, another concern is that many people see big data as representative of the growth of artificial intelligence. While he certainly appreciates the utility of big data, it’s not truly AI in his opinion.
“Big data is nice. It produces something that’s useful to humans. It’s doing something interesting, but it’s not intelligence,” he said. “When Google Search finds something for you, it has no idea what it just read. Does that make it intelligent? No, its useful.”
Schank has similar feelings toward IBM’s Watson cognitive computing system. In his eyes, Watson is merely a use of algorithms that work well but lack the element of human insight.
“AI is about human intelligence. Human intelligence is about getting smarter with every interaction,” Schank continued. “I’m interested in creating a machine that would actually be useful in the sense that it was really smart and could solve the Natural Language Processing problem and the memory problem. And its gotta’ get smarter through experience.”
While its present state is certainly diffuse, the truest form of artificial intelligence in the future will be found at the intersection of memory, learning and education, Schank said. And the computer of the future, he believes, will offer the value of all the wisdom of the world and the ability to deliver it just in time.
“A storytelling computer is really an important step. A step toward simulating the right things, not the wrong things,” he said. “But it’s not going to know who you are. It’s not going to adapt to you. And it’s not going to get smarter. That doesn’t mean the people at Google or Apple couldn’t do that. It’s very expensive and they just haven’t found the economic reason to try.”
Getting to a device that offers timely information that a user needs, as well as the ability to converse and to learn with each interaction, will require plenty of people experimenting to get it right, not to mention a large amount of money, Schank said. While we can replicate just-in-time teaching, ultimately it has to be a device that’s worth investing in.
“The ultimate device I think we should have is the intelligent teacher. The one who can help you when you’re in trouble and have a conversation with you, which a lot of people don’t have. We can do that in the next 10 or 20 years if the money is there.”
About the Author:
Dan Faggella is a graduate of UPENN’s Master of Applied Positive Psychology program, as well as a national martial arts champion. His work focuses heavily on emerging technology and startup businesses (TechEmergence.com), and the pressing issues and opportunities with augmenting consciousness. His articles and interviews with philosophers / experts can be found at SentientPotential.com