It’s hard to believe that the founder of Tesla Motors and SpaceX could have such reactionary views about artificial intelligence (AI). Last weekend, Elon Musk addressed the US National Governors Association’s meeting in Providence, RI and had some very un-Musk like things to say about AI.
“AI is a fundamental existential risk for human civilization, and I don’t think people fully appreciate that,” Musk said.
He even suggested that government ought to regulate the nascent industry. But it’s hard to regulate that which isn’t even formed yet. It reminds us of what happened when the George W. Bush administration decided to restrict stem cell research. The regulation didn’t work because the research simply moved to more friendly jurisdictions. It harkens back to Jurassic Park and the character, Dr. Malcolm, who warned that life always finds a workaround. AI will be no different so we need to embrace it and work with it.
What’s striking is that those pronouncements come from the man who re-invented the space program as a for-profit industry and who also took the electric car to the edge of technological capability and then busted through all barriers. In both cases something new and good resulted. In space, this year his company demonstrated lift capacity to launch 10 satellites at once and return its booster rocket to earth for an upright landing. His cars, though expensive, are gradually reaching the general marketplace and there is a multi-billion dollar waiting list for them.
Let’s not be Polly Ann-ish about this and acknowledge that things don’t always work out but then again, even when things go sideways it’s exceedingly rare that they threaten human civilization.
Let’s unpack this. For all of the notoriety Musk’s success has brought him, it’s worth remembering that his greatest successes have been in re-imagining existing paradigms like cars and space. Before that Musk had a lot to do with getting PayPal up and running. Nothing wrong with any of that; Musk has shown that you can make a lot of money building a better mousetrap. But his comments suggest that he might not be as well positioned on the cutting edge as we’ve been led to believe.
Way out there on the cutting edge there are no paradigms to improve and you have to make things up out of whole cloth and it’s hard. Consider the difference between John Harrison and Charles Lindbergh.
Lindbergh, an aviator, was first to cross the Atlantic in an airplane and he won the $25,000 Orteig prize (over $300,000 in current funds) in the process. But if he hadn’t been the one, some other person would have performed the feat because there were multiple teams jockeying for position at the time and, most importantly, the solution to the problem was constructed from off the shelf parts. Take nothing away from the man who could stay awake for more than 33 hours to make the trip and his courage, but Lindbergh’s feat was of a different order than Harrison.
Two centuries earlier, Harrison figured out how to determine longitude at sea and won most of a prize established in 1714 by the British parliament for a successful solution. The Longitude Prize was worth a cool £20,000 equal to about £2.5 million today. Harrison was a carpenter and clock maker, he knew a lot about making a standing pendulum clock with wooden gears. But his solution to longitude involved building a clock out of metal, without a pendulum and capable of precision not seen up to that time. In his effort Harrison had to reinvent timekeeping and the result was the first chronometer capable of encountering life at sea and maintaining the correct time within a small margin.
So perhaps Musk is more Lindbergh-esque than Harrisonian and as such he’s really more into improving solutions than inventing his own. So the whole idea of the brave new world of AI can seem scary to him and thus his crusade to save us. Musk’s Providence pronouncements were not the first cautionary words he’d offered on AI. The March 26 issue of Vanity Faire magazine carried a long article by Maureen Dowd, “Elon Musk’s Billion-Dollar Crusade to Stop the A.I. Apocalypse.” The article describes Musk’s effort to, “[S]ave humanity from machine-learning overlords.”
On the other hand though Musk’s Tesla factories are models of robotic efficiency and seem to be a significant counter factual to his statements in Providence. But robotic assembly lines might be drone-like compared with Musk’s vision of powerful AI. Possibly, Musk is angling to clear out some space for a new venture he’s been incubating.
His new company, Neuralink, might be an attempt to build what he’s called a “neural lace” that might function as a kind of interface between the human brain and machines.
There aren’t many details yet so keep an eye out for whatever’s next but meanwhile I’d take the Providence declaration with a bowl of chowder.
Everywhere we look today things are changing in fundamental ways. The AI scare, and that’s what it is, was forecasted a few years ago in “The Second Machine Age” by a couple of MIT professors who warned that the acceleration and disorientation we feel is exactly what happens in the second half of an exponential growth curve. There’s no doubt that job losses can be scary and the new era never arrives in time to short circuit the pain associated with the end of a prior era. But there are lots of things we can do to smooth the situation. First and foremost we can quit the unproductive doomsday talk and figure out in earnest how to take advantage of what’s before us.
We can also embrace the change and actively participate in its roll out. Participation yields control and that’s far more useful than trying to shut something down. We’re looking at a free market phenomenon, which means it has a momentum of its own that won’t be stopped by regulation. More or less the same situation Dr. Malcolm found himself in. By the way, electric cars are the same type of free market phenomenon so it’s a pity their patron hasn’t seen that yet.