Former Google Chairman and CEO Eric Schmidt warns that the artificial intelligence (AI) of the future will be much more powerful than today, and will have the ability to manipulate people in ways that some might not have imagined.
“And that’s not good,” Schmidt said, speaking on CNBC's 'Squawk Box,' where he discussed the future of artificial intelligence.
The problem is that these systems will begin to spend more time affecting the way humans look at the world. The systems are dynamic, and are still learning. They cannot explain themselves, and have behavior that humans do not expect.
Schmidt co-wrote a book titled The Age Of A.I. And Our Human Future with Henry A. Kissinger, and Daniel Huttenlocher. It’s about what happens when humans interact with that type of intelligence.
“What we say in the book is that this new generation of AI over the next decade or two will be an epical change in human experience, because we’ve never had another intelligence that’s human like, that’s not human,” he said. “We’ll be working with it, but we’ll also get made about it. We’ll be worried about it … it will change the way a child grows up. I will change military strategies.”
Schmidt said no one is asking these questions to solve them, and he hopes the book will prompt discussions.
Twenty years ago while working on AI, he said, the group did not understand the implications of the technology and what it would do to society. That’s the topic of the book.
“Some of it’s very good, but some of which is very worrisome,” he said.
The industry now knows how to approach these challenges, and that AI optimizes against a specific objective function. The technology works for whatever it is trained to do.
In social media, companies like Facebook try to maximize revenue. The best way to generate revenue is to increase engagement. More engagement can turn into increased outrage. The AI that the three authors are forecasting will become much more powerful than what is seen today in social media, which could increase manipulation.
Ten years ago the industry didn’t realize that AI would be so good that it could target the bias of certain individuals and be used to create misinformation, he said.
In the book we call for a renaissance of thinking,” he said. “We have to get the technical people talking with the philosophers, and the people who understand human psychology. … Imagine a baby growing up, and their best friend is not another human, but a computer. How do you feel about the shaping of that young mind?”