Artificial Intelligence and the End of Jobs


            New Orleans         This whole artificial intelligence thing is sort of an academic exercise for me.  I haven’t tried the various Chat-things.  It’s on my list, but I haven’t been in a hurry.  The more people talk about it as a bit whacky and not ready for prime time, the more relaxed I am about wasting any time on a deeper dive, just happy to be on the trailing edge of new technology once again.  That doesn’t mean that I haven’t read about it or think it’s nothing to follow and worry about.

For example, reading a forever long interview in Wired with the folks behind OpenAI, which has been one of the leaders and darlings of investors in the feeding frenzy of Wall Streeters to get in on this surefire, certain next tech boom-bubble-bust, a comment made by CEO Sam Altman caught my eye.  He was asked about jobs, particularly well-paid jobs for what former Clinton Labor Secretary Robert Reich used to call “knowledge workers.”

I wanted to know how today’s workers … [how] would [we] fare if we were suddenly surrounded by AGIs [artificial general intelligence].  Would they be miracle assistants or our replacements? [He answered] ‘A lot of people working on AI pretend that it’s only going to be good; it’s only going to be a supplement; no one is ever going to be replaced,’ he said. ‘Jobs are definitely going to go away, full stop.’

Now that’s interesting.  Small comfort, he wasn’t necessarily talking about you and me.  There will still need to be people to fix the machines, clean the floors, and make sure there’s coffee, assuming they don’t end the world as we know it.  A professor at Princeton, Ed Felen, studied all of this and “mapped AI’s emerging abilities onto specific professions according to the human abilities they require, such as written comprehension, deductive reasoning, fluency of ideas, and perceptual speed.”  Hmm…sounds like some of us pretend we can operate with those tools on our best days, so maybe I should be lucky that I still know how-to knock-on doors.  What he really found is that “highly educated, white-collar workers” would be the first to go, fully stopped.  The list of the “most exposed” included “management analysts, lawyers, professors, teachers, judges, financial advisers, real-estate brokers, loan officers, psychologists, and human-resources and public-relations professionals.”  Truthfully, many of us would not miss a lot of those people, if we run into them at all.

Covering his tracks, they give money to UBI Charity, “a nonprofit that supports cash-payment pilot programs, untethered to employment, in cities across America – the largest universal basic income experiment in the world….”  I hate to paint this dude as crass, but they also are going hard with a for-profit company that could seamlessly and securely deliver cash payments from governments to all of these people they predict that artificial intelligence will leave jobless so that they can survive somehow.  Who guessed that AI could finally make welfare popular?

A lot of this sound creepy, especially when they demand regulation by governments, which seems as much a shirking of responsibility as a plea for help, and they talk about building an off switch and governor valves to keep these machines from going to the dark side.  The problem of machines with superintelligence and no concept of truth or embedded morality is of course frightening to contemplate.  Reading that this guy also has various places that he has used his millions to establish as hide-outs, if everything goes rouge also doesn’t inspire confidence.  I can sleep at night, because none of this is exactly happening tomorrow, but we must know by now that having confidence in the techies for our future is a prescription for disaster.