new york
—
As if on cue, just days after a viral essay warned of economic disaster caused by artificial intelligence, payments company Block announced it would lay off nearly half its workforce. The company, which owns Square and Cash App, specifically tied the cuts to its AI tools that have “changed what it means to start and run a company.”
That’s a significant cut even for the technology industry, which has soared during the pandemic and shed thousands of jobs in recent months. And unlike layoffs in most other industries, Block’s executives weren’t “rightsizing” or “cutting headcount in anticipation of future AI efficiencies.” The company says it has implemented AI tools and, as a direct result, doesn’t need as many employees.
Block’s stock price soared more than 15% on Friday.
On the surface, the news looked exactly like what Citrini Research’s viral market-sinking blog post predicted earlier this week. Increasingly sophisticated AI threatens to trigger a loop of doom for white-collar workers as “agents” replace office workers, leading to companies cutting jobs and expanding profit margins, increasing investment in AI, and leading to more layoffs.
But don’t let recency bias drag you into the pit of despair.
Some jobs, especially low-level coding, appear to be under pressure as AI bots become better able to imitate human-written software. And this is a real conundrum, especially for those who weaponized the phrase “learn to code” to mock their peers on Twitter in the 2010s.
But is this an economy-wide phenomenon that will push the world into recession? No, not yet anyway.
No one has a crystal ball to say for sure, but there is plenty of history to look back at and say with authority that technology, even the most disruptive, does not shrink the economy. Quite the opposite. Technology tends to increase productivity, giving people more time and money, promoting growth and, yes, employment.
Banks in the late 20th century were able to automate some of the tasks previously performed by accountants and bookkeepers, and the proliferation of ATMs initially reduced the number of bank tellers. However, as JPMorgan points out, these innovations have allowed banks to open more branches, increasing overall employment.
Famously, the Internet has also brought about a spike in productivity. In the 1980s, it took eight employees to generate $1 million in revenue. By the 2000s, only six people were needed.
Although the labor market is cooling, layoffs remain at levels that economists consider historically manageable. The unemployment rate in January was 4.3%, about 0.5 percentage points higher than at the end of 2023, when the generative AI boom began.
That’s not to say AI won’t replace some jobs. But automation has been hurting the labor market for decades, and has not resulted in the universal structural collapse that some are now predicting. It’s also worth noting that the most dire predictions often come from executives who stand to gain the most from selling their companies’ purported AI products. transformative.
Citrini’s 7,000-word essay began by arguing that it was different from the popular “AI apocalyptic fan fiction.” But in reality, most of it was AI apocalyptic fan fiction, a dystopian thought experiment that imagined a scenario in which AI would be a huge success. contract The article went viral in a similar fashion to Matt Schumer’s similarly lengthy “Something Big is Happening” blog post from earlier this month, which argued that the world was underestimating the impending impact of AI on the labor market and compared the current situation to the early days of COVID-19, before the scale of the economic impact could be measured.
Citadel Securities’ Frank Fleit wrote a strong rebuttal to Citrini’s report on Tuesday, arguing that current data simply doesn’t show that AI adoption is moving fast enough to meaningfully replace workers. Fleit also suggests that the report fundamentally misunderstands macroeconomic fundamentals.
In a situation where AI causes a “persistent negative demand shock,” many impossible things would need to happen at once, he argues. Adoption of AI, currently slow and expensive, needs to accelerate significantly. People who lose their jobs to AI will have no one else to hire elsewhere. And perhaps least likely, governments and central banks around the world will have to sit back and let everything collapse.
“It’s also worth remembering that over the past century, successive waves of technological change have neither led to runaway exponential growth nor made the workforce obsolete.”
Many other analysts have called Citrini’s hypothesis far-fetched.
Deutsche Bank’s Jim Reid said of Citrini’s report: “This argument relies heavily on narrative and emotion rather than hard evidence.” “That doesn’t mean it’s ultimately wrong,” he added, “but the atmosphere-to-substance ratio is definitely high.”
