On AI, the Left needs a better answer

Take a breath, then focus

I’m not ashamed to admit that, to this day, my first association when someone utters the phrase “AI” is the movie Terminator 2. To use a term I hear much too often among my parenting cohort, that film marks a “core memory” from my childhood. I saw the movie with a group of kids at the theater in the relatively new Walden Galleria mall outside Buffalo, and it was easily the most dystopian hellscape I’d witnessed to that point. (I’d managed at that age never to have allowed myself to be trapped in the chips aisle of a Wegmans two hours before a Bills game.)

The wasteland depicted in the movie’s opening scenes, along with the storyline revealing that the whole catastrophe might have been avoided but for a few profit-obsessed corporations, was haunting. Frankly, it still is.

The notion that we might today be at that very juncture—that we may be unleashing some similar dystopia on future generations without really considering the implications—surely lurks behind a great deal of the progressive and left-leaning discomfort about the AI revolution that’s emerged since ChatGPT’s release just a few years ago.

Added to that mix is a worry, associated with nearly every technological breakthrough, that AI will render whole swaths of the workforce obsolete, leaving young and future breadwinners without purpose or opportunity. As new polling from Searchlight reveals, more Americans think AI will hurt the economy than help it (37% to 28%), decrease rather than increase wages (55% to 13%), and replace rather than supplement workers (51% to 34%).

Those are, to be fair, legitimate concerns. There’s no denying that every technological disruption in history has, in fact, displaced some segment of the workforce—and you needn’t look far into the recent past to see how even more concentrated disruptions have rippled out into much broader-based political consequences (I’m looking at you, Brexit).

But, by the same token, progressives need to put those concerns in context: the fact that there are so many worker shortages today across so many different industries suggests that, in the long run, the workforce is unlikely to be rendered permanently indolent by large language models. The wheel, after all, didn’t mean humanity had no use anymore for human labor. Nor did steam power, or the sewing machine.

With all that said, the temptations on the Left can be wildly contradictory. If this technology is going to render legions of workers unemployed and, worse, potentially unemployable in the field where they are expert, many would rightfully want to flip a kill-switch.

A movement determined to put the interests of the working class first could not abide a permanent dislocation so massive. But if, in the end, AI is a more modern equivalent of the wheel—if, in the end, it simply empowers workers doing more menial tasks the opportunity to use their minds for higher level work that then redounds to create a more productive society—then perhaps progressivism should be more bullish on AI.

The problem, of course, is that we can’t know for sure which of these two scenarios will play out, or if there’s a third on offer. And so, even before considering what’s politically wise, let’s just consider the limits of what so-called elites actually can claim to know.

First, for all the warnings coming from experts similarly scarred by their own memories of Terminator 2, convincing cases have nevertheless been made for the vast potential of AI for good. To cure cancer and other awful diseases. To unleash and spread new wealth. To free humanity from the drudgery of routinized work so that we can each be more creative and productive. To solve for the problems of what Ezra Klein and Derek Thompson have identified as self-imposed scarcity. To propel the sort of improvement in living conditions, particularly in impoverished parts of the world, akin to those that accompanied the agricultural and industrial revolutions. And who knows what else?

Well, among all the entities you might ask for ideas, the government almost certainly doesn’t know. To be fair, no one today can predict with absolute certainty how AI will evolve, just as none of the experts working on nuclear technology during the Second World War could have predicted during the 1940s how the Cold War might play out. But whatever faith we might have had in the Establishment’s expertise back then, we certainly don’t today.

And that’s for good reason: The American government has piled up a pretty poor record of connecting cause-and-effect in recent decades. And so public skepticism of expert opinion is well within the “fool me once, shame on me; fool me a whole bunch of additional times, and I’m not going back to the same well again” frame of thinking.

The examples are too voluminous to list in full. But most everyone agrees now that Washington got Iraq wrong. And the mortgage-backed securities. And pandemic preparedness. And the so-called science of reading. Why would anyone believe that, as of today, Washington is equipped to make the right call about whether AI should be blessed wholeheartedly or condemned forever? 48% of Americans think the U.S. Congress knows little to nothing about AI. Americans would be right to be skeptical of the government’s judgment.

That doesn’t mean liberals shouldn’t lead, or that they should blindly follow public opinion. President Bill Clinton often quips that voters hire politicians to see around corners—and that’s what liberals should try to do. But if the aggregate effects of AI are too far off to understand in full, what liberals can do is note how AI and other technologies are affecting progress today, and react along those lines.

And that requires a much more nuanced perspective that is neither bullish nor bearish on AI comprehensively, but rather connects to the ways ordinary people are experiencing AI now. Take a few examples.

We all live in fear of a tough conversation with a doctor, more likely about a loved one even than about ourselves. And you can only imagine the relief you’d feel if you learned that a new technology made it possible for doctors and scientists to address a condition that would, just a year earlier, have been an inevitable death sentence.

To the degree AI can be a source of new medicines and cures, the Left should be all for it. To the degree AI can help us detect vulnerabilities in our cyber defenses, we should be all for that as well. To the degree AI can make it easier and cheaper for us to live more comfortable lives, we should be all for that too. No one should have to vacuum their own floor if an AI vacuum cleaner can sweep up for you (without ruining the shag carpeting).

But when AI is clearly being used for nefarious purposes, liberals should be siccing government on the bad guys. Americans are far more worried about government doing too little to protect them from AI than they are about over-regulation, by a 67%-12% margin.

When AI is being used to create deepfakes designed to humiliate young women, for example—to create videos that purport to reveal them in the throes of perversity—we should hold perpetrators responsible. When platforms create mechanisms for children to have inappropriate conversations with chatbots, liberals should be at the forefront of efforts to protect the innocent. When AI is being used to trick the vulnerable into “donating” their life savings, liberals should make their priorities clear.

There are innumerable reasons for the Left to be skeptical of this new technology, just as there are always good reasons for those who care about working and vulnerable communities to look with some degree of cynicism toward changes and developments that stand initially to benefit those who are already doing well.

The fact that Donald Trump and a whole bunch of deep-pocketed “tech bros” who appear to have been born on third base (or, at least by now, are already well past home) are so enthusiastic just puts further stink on a shift that is sure to cause a great deal of upheaval.

But the Left needs to take a breath before directing a stink eye at the enterprise as a whole. We may well look back in a half century and remember this as the moment when we made the mistake of not taking more decisive action—but we don’t know that yet, and we might regret that we stood in the way of technological change that might otherwise have saved untold suffering decades earlier.

Government does not of late have a good record of making big bets, and the American people have earned the right to be cynical. To that end, liberals should aim to be good stewards of the public interest, and demonstrate in more measured ways that ordinary people can trust democratic government to make reasonable judgments.

The American people are telling leaders that they are concerned about job losses with the permeation of artificial intelligence. They’re practically begging leaders to lead on an issue where the government is uniquely situated to provide concrete solutions. Those leaders should listen. They should take the information they have and act on it. And when the future is uncertain, they should pause before rushing to conclusions.