How Can There Be Artificial Intelligence When We Still Don’t Understand Human Intelligence?
Today I went to hear a talk by NY Times science writer John Markoff about his forthcoming book, Machines of Loving Grace: The Quest for Common Ground Between Humans and Robots as a kick-off event for UC Berkeley’s new series on People and Robots hosted by its CITRIS department.
Markoff said he framed the history of robotics, which he recounts in his book, as a dichotomy between those who want to augment human intelligence and those who think we can replace it. The latter group believes in the singularity, in which robots supercede humans.
This simplistic framework enables the author to classify inventors as either AI people or augmentation people. So Doug Engelbart, who invented the mouse, for example, is an augmentation guy, whereas John McCarthy, who first coined the term artificial intelligence to differentiate it from Norbert Wiener’s cybernetics, predicted in the early 60s that true AI was only a decade a way.
Almost twenty years later, in 1980, I interviewed McCarthy for Science Editor, a radio program I co-produced for the University of California and which was distributed by CBS Radio, and even then he was claiming that AI would become a reality very soon.
Markoff says these two communities don’t talk to each other, but he wrote the book in the hope there might be synergy. He also said that since “singularity hinges on scaling, I would argue it’s over,” because we’ve reached a limit as to how many functions a piece of silicon can perform at once — a limit called dark silicon.
As a Silicon Valley native — Markoff was born and bred in the Bay Area — the reporter thinks robotics will improve our lives rather than take away our jobs. In fact, he thinks robots could create millions of new jobs and cited the example of how the Internet has created new jobs.
No one asked about inequality, but several people mentioned the lack of gender and ethnic diversity in robotics and AI. Markoff said that more women might get interested in the field if the focus were on designing human interactions. The only woman he mentioned was someone who created My Perfect Girlfriend, a bot that “talked” to people online in such a natural manner, most thought she was real.
In the last few moments of his talk, Markoff admitted that despite his enthusiasm for his newly purchased, automatically braking car, “It’s about autonomy, not intelligence.”