The AI Washing Scapegoat

A CEO announces he's cutting 20% of the workforce. In the memo, he explains that AI is now handling tasks that used to require entire teams. The press covers it. LinkedIn mourns. And somewhere in a boardroom, a CFO quietly notes that the company had been haemorrhaging cash on headcount it never needed.

This is the pattern repeating across tech right now. And most of the time, AI is just the cover story.

The numbers don't support the narrative

The Yale Budget Lab has been tracking AI's impact on the US labor market across multiple quarterly reports. Their finding, consistent through early 2026: no discernible disruption in aggregate employment attributable to AI. Unemployment sits at 4.3%. Software developer job postings are up 15–30% year-over-year — in the category most directly impacted by AI coding tools.

If AI were genuinely eliminating work at the scale being claimed, this data would look very different.

What it actually shows is that we're three years into the most significant technology wave since the internet, and the labor market remains fundamentally stable. The job apocalypse that was confidently predicted has not materialised.

What's actually happening

Chamath Palihapitiya put it bluntly on the All-In podcast: companies are using AI as a scapegoat to clean up what was, in reality, very poor management over the last five to ten years.

He's right. During 2020 to 2022, major tech companies hired aggressively into a zero-interest-rate environment that made capital feel infinite. Google, Meta, Amazon — they ran talent-hoarding strategies, deliberately taking people off the market to keep them away from competitors. This was an explicit strategy, not a conspiracy. It was openly discussed at the time.

Now the capital environment has changed. Public markets are punishing inefficiency. And these same companies need to get back to fighting weight. So they announce AI-driven restructurings, put "measurer" or "coordinator" labels on the roles being cut, and the media writes the AI-displacement story they're angling for.

Jack Dorsey cut half of Block's workforce and attributed it to AI. Financial analysts on X identified within 24 hours that Block had been running at significantly lower efficiency than its sector peers for years — and had needed this cut long before AI became a convenient framing. The Cloudflare memo did the same thing, describing "measurement roles" as if the company had pioneered a new category of human obsolescence, rather than trimming overhead it had accumulated through weak management. We've written about the market mechanics behind that Block cut in 7.5 X.

David Solomon, CEO of Goldman Sachs, wrote an op-ed in the New York Times in May 2026 making the same basic case: the AI job apocalypse is overblown. His argument: AI will automate roughly 25% of work hours over the next decade, but that's not the same as eliminating 25% of jobs. It means workers shift toward higher-complexity tasks. Goldman's own 200,000-employee AI deployment internally is a test case — not a headcount reduction, but a reorientation.

This matters coming from Solomon specifically. Goldman is not a company known for softening difficult truths for public relations purposes. When the CEO of one of the most hard-nosed institutions in finance says the doom narrative is wrong, it's worth taking seriously.

History doesn't support it either

Bill Gurley made the most clarifying point of recent months when he appeared on the same All-In episode. Pope Leo XIV framed his AI encyclical as a deliberate echo of Pope Leo XIII's 1891 encyclical warning that the Industrial Revolution would harm workers and concentrate wealth in dangerous ways.

Gurley laid out what actually happened between 1891 and today. The work week fell from over 60 hours to 34 hours globally. Real wages rose 8–10x adjusted for inflation. The median worker now earns more than a doctor did in 1891. Global GDP per capita went from $1,500 to over $20,000. Child labour in the US dropped from 18% to zero. Workplace deaths fell by 40x. Life expectancy rose 60%. Global poverty fell from 75% of humanity to under 10%.

The pope in 1891 got it precisely backwards. Every major concern he raised was resolved — not despite technological change, but because of it.

The same pattern has held across every major wave of automation since. Bank tellers increased after ATMs were introduced. Live entertainment became more popular after television. The number of software developers has risen through every iteration of tools meant to reduce the need for software developers.

Gurley's conclusion, and I think he's right: the threat isn't AI. It's refusing to use it.

So what actually threatens your job?

Not the technology. The people using it.

David Sacks made the observation that proficiency in AI tools is probably the single most marketable skill in the economy right now. He compared it to being the only person in the office who knows how to use a spreadsheet in 1985 — the advantage is asymmetric and real, and it compounds.

The data on software developers makes this concrete. Even as AI writes most routine code, job postings for developers are surging. Why? Because the amount of software being built has expanded dramatically. GitHub saw code commits go from 1 billion in a year to 1.1 billion in a single month. When you make something easier, more of it gets done. The developer's job changes — but it doesn't disappear.

The same logic applies across every professional category I work with: finance, consulting, healthcare administration, legal services. The work isn't going away. The distribution of who does what shifts. The people who adapt to that shift thrive. The people who refuse to adapt will find the landscape increasingly hostile — not because a machine replaced them, but because a person with a machine did. It's the same point we made in You Can Be Tom Cruise.

What this means in practice

The AI washing narrative serves no one well. It lets bad managers hide operational failures behind a technological story. It frightens workers who would be better served by learning how to use these tools. And it distracts from the real question every executive should be asking: where in my business does AI change what's possible, and am I building the capabilities to take advantage of it?

That question is harder than cutting headcount and blaming the algorithm. It requires actually understanding what the technology can and can't do. It requires workflow-level specificity, not generic deployment of tools that go unused after week two. I've written about why that generic approach fails in Plug-and-Play AI Is a Myth.

The companies I see getting real value from AI are not the ones making headlines for layoffs. They're the ones quietly rebuilding how work gets done — one specific process at a time.

That's the harder path. It's also the one that leads somewhere.