Teaching artificial intelligence to behave is challenging, but crucial

robot butler
Robot Butler
TheNestTV YouTube

Machines, or robots, can now do tasks that previously required human intervention or attention.

In the coming decades, artificial intelligence is likely to have a profound impact on our lives, our jobs, our recreation and even our wars.

Is artificial intelligence, AI, an existential threat?

The pace of automation and its impact on our society, for good and for evil, is the subject of decades of work by entrepreneur and Stanford University researcher Jerry Kaplan.

Create a More Connected Minnesota

MPR News is your trusted resource for the news you need. With your support, MPR News brings accessible, courageous journalism and authentic conversation to everyone - free of paywalls and barriers. Your gift makes a difference.

"There's something about the people in the field of AI that they love to hype this stuff up, they love the attention," Kaplan said. "And the truth is that this is good for certain classes of problems and not for others."

While the problems that are being solved by these technologies are important, the field of robotics has overpromised on what most machines are capable of, he said.

And it's not just the AI researchers; consumers also tend to anthropomorphize the machines they own and in doing so start to imagine there is a working mind within these technologies.

"And it blinds us to what we can do, and what things we need to be careful of with these technologies," Kaplan said.

Autonomous machines will still only act based on what is programed into them, so in emergencies or human interactions — where anything can happen — many of these "robots" may be unequipped to adapt to new roles.

"It turns out we abide by all kinds of strange subconscious rules and we need to codify those rules and make machines act in a socially appropriate way when they're around human beings," Kaplan said.

Kaplan spoke January 17, 2017 at the Commonwealth Club of California with recently retired Pulitzer Prize-winning New York Times reporter John Markoff. Kaplan is the author of "Humans Need Not Apply" and is out with a new book titled, "Artificial Intelligence: What Everyone Needs to Know."

To listen to their conversation, click the audio player above.

Further reading

• Rise of robot bees: Tiny drones turned into artificial pollinators

• Are robots our future? 'Homo Deus' author thinks so

• 2012: IBM hopes its Watson will become doctor's sidekick

More from MPR News Presents

MPR News presents offers speeches, documentaries and debates — airing weekdays from noon to 1 p.m.