Yes, they discussed the possibility of a superintelligence that could somehow escape human control, and at the end of the month, the conference organizers unveiled a set of guidelines, signed by attendees and other AI luminaries, that aim to prevent this possible dystopia. But the researchers at Asilomar were also concerned with more immediate matters: the effect of AI on the economy.
“One of the reasons I don’t like the discussions about superintelligence is that they’re a distraction from what’s real,” says Oren Etzioni, CEO of the Allen Institute for Artificial Intelligence, who attended the conference. “As the poet said, have fewer imaginary problems and more real ones.”
At a time when the Trump administration is promising to make America great again by restoring old-school manufacturing jobs, AI researchers aren’t taking him too seriously. They know that these jobs are never coming back, thanks in no small part to their own research, which will eliminate so many other kinds of jobs in the years to come, as well. At Asilomar, they looked at the real US economy, the real reasons for the “hollowing out” of the middle class. The problem isn’t immigration—far from it. The problem isn’t offshoring or taxes or regulation. It’s technology.•
Almost developing a driverless car isn’t nearly the same thing as perfecting a fully driverless one, that last two or three percent to be worked out making all the difference. Getting most of the way there is useful but not transformational. When autonomous has truly arrived it will impact environment, safety, economics, law and urban, suburban and rural life in myriad ways.
Mary Barra just announced GM is deploying a new fleet of (almost) driverless vehicles for testing. Despite the bold headlines, that’s no so different than what other traditional automakers and Silicon Valley startups are doing, though the company is stressing that it’s uniquely positioned to mass-produce the cars once autonomous is a going concern–whenever that is. If money and talent are mainly what’s required, the industry has those factors covered. GM alone is spending $600 million annually on their division and is in the process of recruiting nearly 1,200 additional engineers.
While those are solid, well-paying positions, the lucky new hires endeavoring to remove human hands from the wheel will also, if successful, be disappearing millions of blue-collar jobs. That will make us richer in the aggregate but put undue pressure on segments of society, though as Nicholas Carr recently wrote, the promised AI-induced jobspocalypse has yet to materialize despite all the bold predictions. Has our “death” been greatly exaggerated or just deferred?
My best guess is that new tools, once envisioned, often take longer to perfect than we hope (or fear)–remember that Lillian Ross reported on VCRs and a Netflix-like service in 1970! The process, however, may speed ahead faster now than in the past because tools today are cheaper and more powerful. It’s probably more a question of whether we’ll produce an adequate array of new positions to replace the old ones and enable workers to educate and re-educate themselves to continually cope with shifting landscapes.
Two excerpts follow, the first about GM’s announcement, and the second concerning AI’s possible impact on the middle class.•
LAKE ORION, Mich. — General Motors said Tuesday it has finished making 130 self-driving Chevrolet Bolt test vehicles, an achievement that the automaker says will help put it at the forefront of the race to develop and deploy autonomous cars.
CEO and Chairman Mary Barra said GM is the only automaker currently capable of mass-producing self-driving vehicles.
“The autonomous vehicles you see here today are purpose-built, self-driving test vehicles,” Barra said before several hundred employees gathered at the plant in Lake Orion, Mich., Tuesday. “The level of integration in these vehicles is on par with any of our production vehicles, and that is a great advantage. In fact, no other company today has the unique and necessary combination of technology, engineering and manufacturing ability to build autonomous vehicles at scale.”
The self-driving version of the Chevrolet Bolt is the second generation of vehicles capable of handling nearly all road situations on their own without driver intervention. They are equipped with the latest array of equipment, including cameras, radar, sensors and other hardware designed and built by GM and its suppliers.
The new version of the self-driving Bolts must still be driven with a person behind the wheel who is alert and ready to take control if necessary.•
From Cade Metz at Wired:
IN FEBRUARY 1975, a group of geneticists gathered in a tiny town on the central coast of California to decide if their work would bring about the end of the world. These researchers were just beginning to explore the science of genetic engineering, manipulating DNA to create organisms that didn’t exist in nature, and they were unsure how these techniques would affect the health of the planet and its people. So, they descended on a coastal retreat called Asilomar, a name that became synonymous with the guidelines they laid down at this meeting—a strict ethical framework meant to ensure that biotechnology didn’t unleash the apocalypse.
Forty-two years on, another group of scientists gathered at Asilomar to consider a similar problem. But this time, the threat wasn’t biological. It was digital. In January, the world’s top artificial intelligence researchers walked down the same beachside paths as they discussed their rapidly accelerating field and the role it will play in the fate of humanity. It was a private conference—the enormity of the subject deserves some privacy—but in recent days, organizers released several videos from the conference talks, and some participants have been willing to discuss their experience, shedding some light on the way AI researchers view the threat of their own field.