The AI Revolution At Our Doorstep

In an episode of "Real Time with Bill Maher" last month, a panel discussion on rapidly advancing artificial intelligence technology provoked Dutch historian and author Rutger Bregman to say "we're like in February 2020 just before the pandemic hit."  This was a sobering analogy, and fitting given the degree to which basically everybody, aside from tech company executives, is whistling past graveyards about the comprehensive societal disruption that's likely to be right around the corner.

The equivalent of an atomic bomb is about to be detonated upon our species and the only political or media figure talking about it is Bill Maher.  I guess the good news is that those who are currently saddled down with jobs are about to have a lot more time on their hands to obsess about the Epstein File!

To be fair, it's too early to know if the AI revolution lives up to the hype, and both the media and elected officials could be proven wise to hold their fire and not incite a panic.  After all, a decade ago we were told that half the vehicles on the highway would be self-driving by the mid-2020s.  As of 2025, it's still well below 1% of all traffic.  More recently, we were assured that the long-promised electric car takeover that I've been hearing about since I was in elementary school was absolutely, unequivocally, we-really-mean-it-this-time gonna happen, and happen very quickly, in the 2020s.  Several years later, electric car sales' trajectory remains at a crawl, constituting only 7% of vehicle sales in 2024, vastly below expectations from the beginning of the decade.

Is it possible the big talk about AI will fail to materialize as well?  Or at least move at a much slower pace than currently projected?

It's absolutely possible, but I wouldn't count on it.  Not with the global arms race ensuing to perfect the technology and worry about the risks later.  The only thing I can safely project is that the inevitable disruption will differ in a variety of ways than what's currently forecast.  

The early conventional wisdom that the college boys are all gonna be thrown out on their asses from their cushy desk jobs while the plumbers and HVAC technicians will be laughing all the way to the bank is too rudimentary.  The economy doesn't exist in a bubble, so if the college boys that constitute the core of the American middle class experience a deluge of unemployment, blue-collar people will be affected downstream.  After all, if office cube farms are replaced with robots, there won't be much demand for plumbers or HVAC technicians in the human-free commercial property complexes, will there?  Furthermore, it's not as if the ongoing trend of conventional automation impacting blue-collar fields will stop or slow with the advent of AI.  If anything, it will likely accelerate.

Pontificating on further hypotheticals about the magnitude of AI disruption in the job market is ultimately too speculative, at least for an amateur like myself.  All one can do when they hear that more than half of work currently done by doctors and teachers will be handled by AI in 10 years is....to hope they're as wrong as they were about electric cars dominating the highways by 2025.  

The problem is that our elected officials are likely taking the same approach as they not only give the tech barons carte blanche to move forward with this dangerous technology as they see fit, but they borrowed $3 trillion to give them a fresh round of tax cuts as they orchestrate this disruption upon us.  And if the disruption of AI is even a fraction as substantial as these very tech barons are warning, the consequences for public policy will be even more pressing than those for the employment market.

Casual commentators on AI disruption, when faced with the follow-up question of what human workers are gonna do as this revolution plays out, will often tell us the government will have to offer some form of universal basic income to offset loss of employment and income.  There are a lot of problems with that supposition, and the biggest is simple arithmetic.  Even Andrew Yang's 2020 UBI proposal--offering a mere $1,000 per month in benefits--came with a price tag of $3 trillion per year.  The government collects about $5.2 trillion per year in revenues.  

In other words, even a UBI program that backfills a mere $12,000 per year in income, which is well below the poverty line, would consume nearly 60% of the federal budget.  This is the same federal budget currently running deficits in the trillions of dollars per year just trying to keep up with Social Security, Medicare, Medicaid, the military, and interest on the existing national debt, all programs whose tab will only exponentially increase during the same period of time that AI is expected to displace tens of millions of middle-class jobs.

There's always the possibility that AI will trigger a surge of economic growth so blistering that it will double federal revenues and UBI and everything else our hearts desire will suddenly become affordable.  That seems unlikely, and even if some iteration of that scenario played out, the tech oligarchs will run the country and the same guys sitting in the front row of Trump's inauguration greasing the skids for their budget-busting tax cuts in 2025 will be sure than any surge of federal revenues gets drained into their bank accounts in the form of future tax cuts.

Furthermore, whether AI generates an era of budgetary abundance or scarcity, you can be sure American politics will still operate under its long-standing zero-sum rules, spearheaded by a Republican Party that will effectively divide the working class against itself as the Republicans have done successfully since the Nixon era.  Those predicting the disruption of AI will necessitate an embrace of socialism and/or an expanded safety net are likely to be sorely disappointed.  AI will almost assuredly be rocket fuel to already scandalous levels of inequality, but the story probably won't be "winners vs. losers" so much as "major losers vs. minor losers".  And that's where the political story will get complicated.

I can already hear the Republican campaign ads targeting the "minor losers" of the AI disruption, warning them that the "lowlife, parasitic rabble and festering subhuman filth are sponging off of hardworking folk like themselves and eating Cheetos all day.  Let them get a damn job!!!"  And it will work!  Guaranteed!  Maybe an expanded safety net message will work in one election for the center-left.  Possibly even twice.  But the life cycle of safety net expansion, no matter how desperately needed, will almost assuredly run out of gas long before the economic disruption connected to AI will.  Anybody who's studied the recent history of American politics even a little should be able to see that coming a mile away.  

The bottom line is that neither our gluttonous private sector, willing to torch the country to ashes in pursuit of a strong quarterly earnings report to greet their shareholders with, nor our public sector, beholden to an easily distracted electorate licking their chops to villainize their struggling neighbor rather than the oligarchs bleeding them dry, will be the least bit prepared or capable of managing the decline in what could really live up to the typically overhyped term "existential crisis".

But maybe our new machine overlords will make it easy on us....and simply exterminate our entire species before we get a chance to wring our hands about it merely taking our jobs away.  The fact that virtually nobody in the AI development realm is prepared to debunk that possibility should alarm us that digital genocide of the human race is a very real possibility and possibly even the most likely outcome. And yet, everybody at every level overseeing the introduction and expansion of this technology is ready, willing, and able to take that risk.  

The highest-stakes gamble in human history awaits us in the immediate aftermath of a narcissistic sociopath becoming the most powerful man in the world.  Oh well.  At least we'll have high-pressure shower heads to enjoy before we go!