In late 2022, when ChatGPT was first blowing everyone’s minds, Kenji didn’t really get it. Sure, it did some pretty cool things. But so had plenty of other generative AI models he’d tinkered with over the years. What so many outsiders thought was a world-changing breakthrough seemed to him, a machine learning engineer, much closer to an incremental improvement. All the fear that AI would swallow up everyone’s jobs felt overblown. He certainly didn’t think it could do his job anytime soon.
About a year ago, though, he began to feel less sure. The AI tools he was using in his work at Block, the payments processor formerly known as Square, were becoming so good that he found himself delegating more and more of his coding and analysis to them. “At some point you look around and say, ‘Gosh, I’m not doing that much of the work anymore, am I?'” he tells me. Still, he believed his human mind had an edge: He understood the broader context necessary to ask the right questions. Besides, he was in a role that was close to the money, and businesses rarely cut those kinds of positions. “It certainly dawned on me that I could be in line for redundancy,” he says. “I just didn’t think I was quite there yet.”
Last week, Kenji was among the more than 4,000 employees who were laid off at Block. And this wasn’t just your run-of-the-mill restructuring. Block was culling nearly half its workforce — an astonishing move for a company that isn’t on the verge of bankruptcy — and its CEO, Jack Dorsey, left no room for interpretation about why. “Intelligence tools have changed what it means to build and run a company,” Dorsey told investors. “A significantly smaller team, using the tools we’re building, can do more and do it better.”
The most ominous line in Dorsey’s letter came toward the end: “Within the next year,” he wrote, “I believe the majority of companies will reach the same conclusion and make similar structural changes.”
I do feel a little bit like a horse and buggy in the age of the automobile.
If Dorsey is right, white-collar professionals are facing a stark future — one that Kenji had the misfortune of entering first. What he sees now from the other side feels instructive. The day after his layoff, he spoke with me on the condition that I not use his real name to avoid jeopardizing his future job prospects. I expected him to be angry or at least bewildered. Instead, he was almost Zen-like in his reflections, clear-eyed and calm. “There was the first 30 seconds of holy shit,” he tells me. “But then, as I read the whole thing, I was like, ‘Yeah, I get it.'”
One reason Kenji may have been so quick to understand is that he’s had a front-row seat to tech’s cycle of automation for years. As a machine learning engineer at Block, Kenji built systems to detect fraud automatically — part of a broader shift in finance that has reduced the need for humans to review transactions for signs of suspicious activity. Block employs a team that reviews a small sample of transactions to make sure the models are flagging the right cases. Kenji understood that the better those models got, the fewer people would be needed to check their work.
There was something unsettling about the reality that someone whose work helped automate other people’s jobs was now being automated himself. For years, the standard advice was simple: Become the automator, lest you become automated. Five years ago, machine learning engineering was the future-proof career — the apex predator of all occupations — until, suddenly, it wasn’t. Large language models became the hot new thing, relegating traditional machine learning to yesterday’s breakthrough. “I do feel a little bit like a horse and buggy in the age of the automobile,” Kenji says.
In the face of disruption, Kenji did what any good technologist would do: He embraced the new tech. For more than a year, Dorsey had urged employees to incorporate AI into their work, and Kenji became a power user of Block’s homegrown AI tools, goose and g2, as well as third-party ones like Claude and Cursor. That didn’t protect him. If anything, he wonders whether it contributed to his job loss. “Over the last year that we were strongly encouraged to use all these AI tools, we were laying the foundations for our own replacement,” Kenji tells me. “If you show the tool how to do a task once or twice, it can kind of take it from there.”
The question now is whether there will be a lot more Kenjis in the months ahead. Dorsey and plenty of industry executives clearly think so. Since last week’s announcement, though, critics have argued that Block may have simply overhired during the pandemic. Maybe this wasn’t AI making it possible to halve the workforce so much as a company correcting a bloated headcount. AI, in that telling, was merely the excuse, not the underlying cause of such sweeping cuts.
My sense is that more companies will follow, as Dorsey predicted — just not as quickly as he suggests. As I wrote last year, startups building from scratch in today’s AI age are already operating with dramatically leaner teams. It’s not hard to imagine large companies eventually getting there, too.
The Googles and the Amazons of the world have entrenched workflows, team structures, and job descriptions they’ll need to redesign around AI, and that overhaul will take time. The more likely scenario is that layoffs come in waves over several years, alongside persistently low hiring, steadily shrinking companies’ headcounts.
For now, Kenji plans to take a break, thanks to the sizable severance Block offered him (the company offered laid-off employees at least 20 weeks of pay, with an additional week of pay for every year of tenure). Then, he’ll start looking for a new job. “I have enough knowledge, enough context, enough connections that I’m quite confident I could land another job in the near term,” he says. Machine learning might not be the hot new thing it was a few years ago, but it’s still in demand.
The longer term is a different story. “If I land a job tomorrow,” he tells me, “I have zero confidence that it, too, couldn’t be automated away in a couple years.”
Aki Ito is a chief correspondent at Business Insider.
Business Insider’s Discourse stories provide perspectives on the day’s most pressing issues, informed by analysis, reporting, and expertise.
