opinion

COLUMN: The limits of artificial intelligence



Most of us probably don’t live with the constant fear of a robot uprising, but it’s still scary and exciting to know how far artificial intelligence has come.

Some of us might hope for a future with robot maids from “The Jetsons” and comedic Benders from “Futurama.” AI still has a long way to go before robots become more humanlike, and the recent breakthroughs in AI tell us about the nature of intelligence itself.

Two weeks ago, the Google software AlphaGo beat the current world champion at the complicated board game “Go.”

This is a huge step forward for the efficiency and productivity of computers and engineering. It’s difficult to say what it means for the future of AI, but that doesn’t mean it isn’t fun to speculate what it might 
look like.

In 15 years, the professional jobs for which we prepare students will be done by intelligent machines, wrote Eric Cooke of Times Higher Education. Devices will replace university faculty by that time, Stephen Joel Trachtenberg said.

Though robots have done amazing things — from beating “Jeopardy!” pros to speeding up business transactions — there are still things that separate robots from humans.

Our software still struggles to replicate feelings, emotions, creativity, fantasies or other humanistic characteristics.

It might be easy for a computer to perform mathematical calculations, but it’s difficult to imagine how a robot could compose a Mozart concerto or write a James Joyce novel.

David Gelernter, professor of computer science at Yale University, wrote in the Wall Street Journal, “AI could, in principle, build a simulated mind that reproduced all the nuances of human thought and dealt with the world in a thoroughly human way, despite being unconscious.”

While this could happen in principle, we’re still far from understanding the scientific basis of how our brains deal with nuances of thought. Recently, Microsoft’s chatbot Tay turned into a misogynistic racist Holocaust-denier due to the way it copied behaviors of Internet trolls.

Many writers have expressed concerns about the ways AI might be used for dangerous purposes. But this catastrophic failure is only the work of online jokesters, not anything significant nor anyone’s serious intentions.

After the supercomputer Deep Blue’s victory against the world chess champion Gerry Kasparov in 1996, Douglas Hofstadter, a professor of cognitive science at IU, said, “It was a watershed event, but it doesn’t have to do with computers becoming intelligent.”

Robots aren’t overtaking human beings in intelligence. They’re only beating us at things we used to think were incredibly difficult for a robot to do.

In reality, we’re understanding a newer meaning of our own intelligence. When we realize what robots are capable of, we understand more about what separates them from us.

And, through this, we can bridge the differences between robots and humans in order to develop AI capable of doing what isn’t currently possible, such as composing a symphony or writing a poem. As we continue to push the limits of AI, we will understand what makes us human.

When that happens, we can worry about robots taking over the world.

Like what you're reading? Support independent, award-winning college journalism on this site. Donate here.

More in Opinion



Comments powered by Disqus