The Guardian recently published an editorial of mine advocating against yet more public investment in AI:
The AI revolution is here and, as its proponents remind us daily, it will remake our world, making every company and government agency more efficient and less error-prone while helping us unlock hitherto unheard of advances in science and technology. Not only this, but if we play our cards right, big tech’s latest explosion could yield unprecedented economic growth.
Though, we might ask, growth for whom?
Consider OpenAI, the technology giant behind ChatGPT. In a promo video announcing the latest upgrade for their flagship software, its CEO, Sam Altman, bragged: “It can write an entire computer program from scratch.” Three days later, the New York Times reported that computer science grads “are facing some of the highest unemployment rates” among their peers. And it’s not just coders and engineers. AI-powered automation promises to swallow up jobs at the low end of the labor market too, with McDonald’s, Walmart and Amazon all clamoring to integrate AI tools to automate everything from service interactions to warehouse picking and sorting.
But more than this, I made the case against the so-called “knowledge economy” as such:
In the late 1990s, the dawning of the knowledge economy was heralded as the solution to many economic woes. As the economy of brains replaced the economy of brawn, Americans were promised new heights of greatness. Sure, factories would close and with them millions of high-wage, union jobs would disappear, but the new jobs at Google would be so much better. As a generation of workers was laid off, their children were encouraged to “upskill”, go to college, and learn to code for the jobs of the future. How ironic, then, that AI, the zenith of knowledge work, is resulting in the abolition of knowledge jobs. Karl Marx once wrote that the bourgeoisie created its own gravedigger in the immiserated proletariat. Today’s tech elite seems intent on realizing that prophecy.
It’s not only that the information age supercharged a new class of oligarchs, from Bill Gates and Jeff Bezos to Elon Musk, who now command unfathomable sums of wealth. It’s also that further down the income ladder, wide class cleavages have opened up along educational lines. As computer-based work became prized, wage inequality between college-educated and non-college-educated workers created a widening social gulf.
A few points that didn’t make it into the article are worth mentioning.
First, despite the seemingly new insistence that ‘we’ better not ‘fall behind’ in the AI race (where is the finish line exactly?), injunctions to innovate have been a part of our civic religion for quite some time. An eon ago, Ronald Reagan argued that “the breakthroughs in superconductivity bring us to the threshold of a new age.” And it's “our task,” he continued, “to herald in that new age with a rush.”
After him, Bill Clinton prided himself on making “unprecedented” investments in science and technology in order to “unleash” the New Economy. Not to be outdone, every president since then has boasted of record-breaking tech investments. Unprecedented investment is now the precedent.
And what of the effects of all this tech boosterism? Are we now living in the techno-utopia Reagan predicted? Millions of lost factory jobs, skyrocketing inequality, stagnant wages, and bedeviling inflation seem to suggest otherwise. Even the ‘winners’ of the Knowledge Economy—college-educated workers—seem to be losing.
We built an entire economic strategy around the promise that a college education would be a ticket to the good life. Clinton insisted that “what you earn depends on what you learn.” Obama implored us to “learn to code.” Yet the wage advantage of a college education has crumbled. Not surprisingly, the number of people who think getting a college degree is important recently hit a new low—just 35%. And some 37% of college graduates are pursuing careers in skilled trades that require no such degree.
Meanwhile, further down the income ladder, we see a worrisome rise in the number of non-college educated workers dropping out of the labor force altogether. By one estimate 10.5% of American working-age men have stopped looking for work altogether. This is what the indignity of obsolescence looks like.
The idea that we could replace place-bound blue-collar jobs with global laptop gigs was always a fantasy. So, too, was the idea that education could act as a substitute for industrial and social policy. In some ways these notions have actually exacerbated our social and political crisis. By insisting (and then ensuring) that obtaining a college degree is necessary for success, we’ve effectively robbed working-class neighborhoods, small-towns, associations, and labor unions of an entire generation of leadership. If you were a smart and ambitious blue-collar kid, you weren’t expected to succeed your parents in place and profession. Nor to represent your class interests in politics. Instead, you were likely plucked from your social world and set down on the college conveyor belt. You were expected to supersede your folks, to transcend your class, and win a white-collar career. As a result, the losers in the Knowledge Economy lost some of their sharpest young prospects to the laptop class.
But even on its own terms, the tech-first approach to economic development appears to be a failure. The entire period of the Information Age has yielded uneven and unimpressive GDP growth. Especially when compared to the gains of the much ruder industrial epoch. Even when the numbers are strong, GDP growth today has a weak relationship to wage increases. In an industrial economy productivity gains are tightly related to wage increases, but in our Knowledge Economy they have decoupled entirely. Whatever economic gains are reaped from increases in productivity they accrue exclusively at the top. Wages for the vast majority remain flat.
So warped is our Information Economy, that growth itself now depends largely on the consumption habits of the rich. According to one Moody’s analyst interviewed by the Washington Post “everything rests on what that top 10 percent decides to do or not to do.” The wealthy now account for fully half of all consumption. In such a lopsided economy there is a real danger that wage growth for the rest of us won’t even keep pace enough for system-wide stability. That danger becomes yet more acute as Silicon Valley seems intent on automating high-wage jobs like trucking. A society where the rich consume most of the product and hoard most of the money, while the rest of us are made increasingly redundant, or else trapped in low-wage dead-end jobs cannot work. A rebuild is in order.
Fixing this requires a renewed commitment to investing in the place-based—and decidedly low-tech—economy. A recommitment to investments in infrastructure, industry, and public service:
Most of what we actually need to achieve some measure of the common good requires common labor. To rebuild our crumbling infrastructure and even to upgrade our electrical grid, we need electricians, steelworkers and bricklayers – not gargantuan data centers. To clean city streets, we need more, and better-paid, sanitation workers – not “smart” trash compactors. To handle problems of crime and social order we need more police officers on patrol – not a fleet of robot crime dogs. To improve transportation, we don’t need self-driving cars, we need buses and trains with people who drive them. In other words, there is plenty of meaningful work to be done, if only we, as a society, invested in the low-tech economy. Not to mention that all the essential stuff of life – love, family, friendship, community – are still best left in analog.
The good news is that new breakthroughs in technology should make such an economic strategy more, not less, viable. Advances in robotics and AI promise extraordinary gains in productivity that, if employed as part of a broader industrial strategy, can and should result in shorter workweeks and faster build times for new bridges, warehouses, roads, power plants, and factories.
Ironically, in order to reap those rewards we’ll need to leave the Knowledge Economy behind.