Artificial intelligence has a lot of prominent people shaken up. Elon Musk, Sam Altman and others  worry AI programs and AI-enabled robots might replace humans in their jobs before the economy can adapt, or worse run amok in Terminator-like apocalyptic scenarios.

Entrepreneur and Singularity University founder Peter Diamandis is worried about the opposite situation. He fears the field of artificial intelligence could be stifled by rules the way stem cell research was under Republican President George W. Bush, who in 2001 announced a block on federal funding for new stem lines. "It had the experience of really putting the kibosh on that kind of work," Diamandis tells Inc.

"One of of the things I think is very true and important for people to realize is that you can't regulate against technologies. If an individual is working in AI or biotechnology or whatever the case might be, and you say 'that's way too dangerous, we need to slow this down, we're going to put hurdles and regulations in front of it here in the United States'... All that means is that technologies leave the U.S.," Diamandis said in a phone interview.

The ultimate effect is not that research in the area ends, but that it ends where the regulations exist and are followed, "and so all you're doing is starting to put barriers up against yourself."

The view isn't surprising coming from Diamandis, whose organization Singularity University (a for-profit corporation that does not confer degrees) is optimistic about technological progress even by the Panglossian standards of Silicon Valley. The premise of the university's programs is that technology can lead the world to a place where all humans have access to first-world standards of living, as described in Diamandis's book Abundance: The Future is Better Than You Think.

But Diamandis's confidence in the merits of AI doesn't mean the AI community shouldn't come up with its own guidelines to prevent technologies in the field from taking a turn for the destructive, he says. He defers to his colleague Neil Jacobstein, chair of AI and robotics at Singularity, to explain how AI can be directed along a productive path.

Speaking to a room of about 50 business people during a quarterly Executive Program on the Singularity campus at NASA in Santa Clara Monday, Jacobstein said many of the dangers of AI boil down to trust issues between people.

He cited as an example Microsoft's chatbot Tay, which started spewing anti-Semitic, racist and misogynist tweets when she was released into the wild of the internet. The mishap with Tay was partially a result of Microsoft's naivety about how AI can be abused by humans. It was "our fellow human beings," he said, "who hijacked AI and made it very prejudiced."

Jacobstein clarified that he thought AI could pose dangers absent human interference--an AI could be very good at a certain task that, when taken to an extreme, results in great harm. Because of AI's propensity for performing tasks that might have unforeseen consequences, and because of AI's susceptibility to human manipulation, one responsibility of companies working with AI is to release products that are hard to tamper with.

Jacobstein made the last point when an audience member asked why Apple's Siri and Microsoft's Cortana often fail to meet expectations. You ask for directions to a restaurant and Siri starts calling a random person in your phone's list of contacts. That kind of misfire is enough to cause many to doubt AI is as impressive as tech pundits present it to be.

In a defense of sorts, Jacobstein said Siri and Cortana are limited in their capabilities on purpose. Apple and Microsoft have the capability to put out something better, but it's prudent for the companies to hold back until proper safeguards are in place.

"You're going to see the products in the world trail the power of the technology," he said.

Editor's note: This story has been corrected to reflect that Peter Diamandis is founder of Singularity University, not CEO. Singularity has provided further clarification that the organization is specifically classified as a benefit corporation, a type of for-profit entity that has legally defined goals of both making profit and benefiting society. 

Published on: May 3, 2016