AI is Corporate Ozempic

Yesterday, Jack Dorsey cut nearly half of Block's workforce -- over 4,000 people -- and told shareholders exactly why: "A significantly smaller team, using the tools we're building, can do more and do it better." The stock surged 25%.

Then he said what most CEOs are thinking but won't admit: "I don't think we're early to this realization. I think most companies are late."

This is the moment. AI isn't just automating tasks anymore. It's restructuring the org chart.

Critics call it "corporate Ozempic" -- a crash diet dressed up as strategy. The implication is cynical: companies are starving themselves thin to juice quarterly earnings, and the weight will come right back.

But what if that framing misses the deeper mechanism?

Ozempic doesn't just suppress appetite -- it resets the body's metabolic set point, helping patients maintain a healthier weight long after treatment. It forces a reckoning with consumption habits that were never sustainable. It's painful. It works. And the weight stays off.

AI is doing the same thing to companies. This isn't a layoff cycle. It's a margin regime shift.

The Numbers Are Brutal -- And They Should Be

The data from 2025 is in: 244,000 tech jobs cut globally. Q4 saw the highest quarterly layoffs since 2008. And 2026 is accelerating -- Block alone just erased 4,000 positions in a single announcement.

The reflex is to call this a crisis. But zoom out and ask: was the pre-AI headcount sustainable?

Consider the hiring binge of 2020-2022. Tech companies added headcount at rates that would have embarrassed a government agency. Block went from under 4,000 employees in 2019 to over 10,000 before this week's cuts. Meta went from 48,000 employees in early 2020 to over 87,000 by late 2022 -- an 80% increase while revenue grew 40%.

The math never worked. Everyone knew it. Nobody wanted to be the one to say it.

AI said it.

The Meta Case Study

When Zuckerberg declared 2023 the "Year of Efficiency," critics called it damage control for metaverse losses. Then the results came in.

Meta cut 21,000 jobs. The stock surged 20% on the next earnings report. The company gained $197 billion in market value in a single day -- the largest one-day gain in stock market history. Revenue per employee jumped. Margins expanded. The company announced its first-ever dividend.

Two years later? Meta's market cap has climbed past $1.8 trillion. The efficiency stuck.

This wasn't financial engineering. This was a metabolic reset.

The bloat wasn't in bad people -- it was in processes that had calcified, layers that had accumulated, projects that existed because someone needed to manage them. AI didn't replace workers so much as it revealed how much work wasn't necessary.

Here's where the Ozempic metaphor gets uncomfortable. The drug doesn't just eliminate fat -- it reduces appetite for the consumption that created the problem. Patients report losing interest in alcohol, gambling, even shopping. The medication doesn't just treat symptoms; it changes behavior.

AI is doing the same thing to organizations. It's not just automating tasks -- it's eliminating the organizational demand for those tasks.

The middle management layer that existed to translate between executives and workers? AI can summarize and synthesize directly. The reporting structures that required human intermediaries? Dashboards and agents handle it. The coordination costs that justified large teams? Plummeting.

This isn't about replacing humans with robots. It's about recognizing that much of what companies did wasn't work -- it was organizational overhead masquerading as work.

The Cautionary Tale: Klarna

Not every company gets the dosage right. Klarna announced in early 2024 that its AI assistant was "doing the work of 700 people." The CEO paraded the stat everywhere, positioning Klarna as the template for AI-native efficiency.

Then customer satisfaction cratered. Complex issues went unresolved. The company scrambled to rehire human customer service workers and even forced engineers to work the phones as the AI failed.

The lesson isn't that AI efficiency is a mirage. The lesson is that you can't just excise muscle and call it fat. Klarna cut into customer relationships -- the one area where human judgment still compounds. They confused headcount reduction with value creation.

Meta didn't make that mistake. They cut coordinators, not creators. Processes, not products. Dorsey seems to understand the distinction too -- Block's cuts targeted roles that AI can genuinely replace, while the company continues hiring for AI development.

The Uncomfortable Prescription

The Ozempic discourse has a parallel that's instructive here. Critics argue the drug is a "shortcut" -- that patients should simply eat less and exercise more. But decades of data show that willpower rarely defeats metabolic dysfunction. The set point is real. Without intervention, bodies fight to return to their dysfunctional equilibrium.

Corporate organisms work the same way. Companies don't bloat because executives are stupid. They bloat because every hire creates a constituency, every process creates a defender, every team creates a Slack channel that will fight to justify its existence. Parkinson's Law is metabolism, not malice.

AI breaks the cycle because it doesn't care about constituencies. It optimizes without politics. It's the intervention that lets organizations escape their own institutional gravity.

What Comes After the Cut

Novo Nordisk -- the company that literally makes Ozempic -- just announced it's cutting 9,000 jobs as generic competition heats up. Even the pusher isn't immune to the prescription.

But here's what the doomsayers miss: companies that emerge from AI restructuring aren't just leaner. They're faster. More responsive. More capital-efficient. They can experiment more because each experiment requires fewer resources. They can pivot because there are fewer layers to convince.

Companies with AI-driven productivity gains are outperforming the S&P 500 by 29%, according to an analysis of 217 firms. The market isn't being fooled. It's pricing in durable efficiency.

The Long Game

Schumpeter called it "creative destruction" -- the gales that periodically sweep through capitalism, clearing out deadwood and making room for new growth. It's painful for the trees that fall. But the forest is healthier for it.

AI is the strongest gale in decades. And like Ozempic, the benefits aren't just the immediate weight loss. They're the metabolic reset that makes the new weight sustainable.

Dorsey put it bluntly: he'd rather act now "honestly and on our own terms" than be "forced into it reactively." That's the CEO equivalent of scheduling your own intervention before your body forces one on you.

The companies that survive won't just be smaller. They'll be structurally incapable of bloating the same way again. That's not a bug of AI adoption. That's the feature.