We are still not seeing a prospect for AI to have a serious negative impact on overall employment
Yeah, we are.
There are over 1.3 million truck drivers, 1.5 million 'rideshare' (uber, lyft, etc.) drivers, and about 300,000 taxi drivers in the US; Almost all of them will be put out of work in a very brief period when autonomous vehicles reach the tipping point into commercial viability, which will likely happen within a decade, perhaps much sooner.
Autonomy is a big deal in moving vehicles, and the debate over self-flying aircraft has been going on even longer than it has for self-driving ground vehicles. As a former Boeing employee, I have feelings about it, but my feelings are perhaps not as intense as those who feel that their livelihoods can be wiped out by AI. You present a nightmare scenario in which one can imagine millions of people thrown out of work rather suddenly, and you present a timeline for it that makes this very threatening to workers in the service industry you are personally involved with. So I'm not going to tell you that you have nothing to worry about. There are a lot of people who run public transportation companies that salivate over the idea of not having to pay all of those salaries. So they'll spend a lot of money on driverless vehicle projects that may turn out to be more of a gamble than promoters of the technology would like them to believe. I simply don't believe that your timeline for the driver apocalypse is realistic, and I do think that there will always be human beings on conveyances with automatic guidance systems. It is too dangerous to entrust the lives of the public to fully automated vehicles without the ability of a human being to take over control of the vehicle.
When I visited airbus once, a tour guide on one of their newer planes, while showing us the pilot cabin, quipped that their dream vision for the cabin was to have a single button to cut off the autopilot and a guard dog to prevent a human pilot from pushing it. The chief test pilot at Boeing once gave us all a nice lecture on the controversy between Boeing and Airbus in which he compared the experience of flying more pilot-friendly Boeing planes vs. pilot-averse Airbus planes. What the Boeing pilot pointed out to us was that most reported incidents tended to be caused by pilots not understanding what the automation was doing. That was certainly a factor in recent times with the disasters which befell the the Supermax 737s. Boeing had tried to sell those aircraft as if they needed only minimal pilot retraining and didn't need more expensive safety features installed, which they were trying to sell as extras. The chief test pilot at Boeing once gave us all a nice lecture on the controversy between Boeing and Airbus in which he compared the experience of flying more pilot-friendly Boeing planes vs. pilot-averse Airbus planes. In fact, those aircraft had entirely different automated piloting that had been installed in 737 airframes--not a great engineering decision, but one intended to save the company a lot of redesign and retooling in the manufacturing process.
That's three million people in one skill area alone, out of a us workforce of 165 million; A two percent increase in unemployment would be very noticeable, even if it were not massively biased towards the "flyover" states where other employment has already largely disappeared. The proportion of midwestern workers who are drivers, particularly truck drivers, is far greater than in the population as a whole.
It certainly isn't profitable to employ humans to drive vehicles, if your competitors are doing the same work without them.
And autonomous vehicles are very close to reaching that tipping point. They're not actually intelligent (but then, in my experience, nor are a significant number of human drivers). But they don't need wages, or rest breaks, or vacation time.
Oh, I agree with you that they're not actually intelligent, but autonomous vehicles are moving robots, so there is a need for much greater intelligence than there is for chatbots trying to trick people into thinking that they can think. Right now, we have a lot of pilot projects going on and some limited deployment of these highly experimental systems. I think that the rush to deploy them is way premature, but that does reflect this rather stupid mentality that machines are somehow smarter than humans and can actually replace them. My view on automation in vehicles is that it has a bright future when seen as a way of augmenting humans so that they convey vehicles more accurately and safely. Machines don't draw salaries, but the people who create and maintain them do. Anyway, it will be a long way off before those driverless vehicle projects throw millions of people like yourself out of work. All it will take is a lot of people being killed in a few newsworthy accidents involving one of those vehicles.
This isn't a new or unique phenomenon, of course. Automation has been making various professions obsolete for centuries. But it's still going to be a rude shock to a system that's configured with the expectation that people can and should get jobs to earn a living. Increasingly that's not true. And the rate of increase is accelerating rapidly.
If you pay attention to the news, it is happening rapidly. It is all the rage among company executives. They view labor costs as the single most controllable expense that eats into profits. I've been in company meetings where we were told that management saw labor costs as something like a dial on a machine that could be dialed down if we couldn't come up with ideas for cutting costs. During financial crises, they didn't have many other dials to fiddle with. However, it turns out that labor is still indispensable, and firing workers is an effective means of losing experience and expertise actually needed to make a profit.
