Oracle just laid off up to 30,000 people. By email. At 6am. System access revoked before most finished their morning coffee. This from a company that posted a 95% jump in net income last quarter. Oracle isn’t in distress. It made a massive debt-funded bet on AI infrastructure $300B OpenAI deal, $50B in capex, $124B in total debt and is now converting its workforce into cash flow to service that debt. The workers who built Oracle’s products for decades were eliminated so the balance sheet could absorb a bet made by people who get paid regardless of how it turns out. The numbers across the economy are staggering: → 1.17M U.S. job cuts in 2025 — highest since COVID → 55,000 AI-attributed layoffs — 12x from two years earlier → Entry-level tech postings: down 67% → 13% employment drop for 22–25 year-olds in AI-exposed roles since 2022 → U.S. government: 300K+ jobs. Amazon: 30K. Oracle: 30K. Intel: 25K. Microsoft: 15K. Block: 40% of its workforce. The companies blaming AI are the same ones that hired 300 people they didn’t need during zero interest rates. AI is just the excuse. The real story: “We had no idea what half these people did, and now we have a socially acceptable reason to admit it.” Marc Andreessen said it plainly this week: AI is the “silver bullet excuse.” The real drivers are zero-rate COVID hiring binges and massive overstaffing. Every large company is overstaffed by 25-75%. And as he put it: “AI literally until December was not good enough to do any of the jobs they’re cutting. It just can’t have been AI.” Oxford Economics and Yale’s Budget Lab confirmed it , firms aren’t replacing workers with AI at scale. They’re using AI attributions because it gets a stock bump. Jensen Huang went further at GTC 2026. When asked why companies keep cutting while claiming AI boosts productivity, his answer was devastating: “Because you’re out of imagination. For companies with imagination, you will do more with more.” The CEO whose chips power every AI system on the planet said layoffs are a leadership failure. Not a technology constraint. His vision: “Every carpenter could now be an architect.” That’s what AI makes possible. Not fewer workers , more capable ones. So why aren’t we retraining? Why aren’t we reskilling? Why aren’t we building new business models that redeploy people instead of discarding them? Because it’s easier to cut than to lead. Meanwhile, 55% of employers who laid off workers for AI already regret it. Klarna replaced 700 people with AI, quality collapsed, and they rehired humans. We’re destroying the pipeline 89% of 2026 graduates now fear AI will eliminate entry-level roles. CS enrollment is projected to fall 20%. The junior analyst not hired today is the future CFO who cannot be promoted tomorrow. Before you send the next layoff email, answer one question: Did you spend even 10% of what those severance packages cost on retraining? Build AI skills. Build distribution. Learn to lead.
"The workers who built Oracle's products for decades" are now going to come together with local AI models, synergized edge inference, claws and MCP and vibe code alternatives to Oracle.
Oracle has a long reputation for their sales tactics. Even before cloud, they twisted elbows.Not a fan of that company for so many reasons.Just from what I have read in the news at various times: ORCL wants to stay commutative to tech giants in AI space, while banks flipped ORCL the bird with funding. So they are firing to raise money for AI.30k employees at an average of 200k cost to Oracle is about 6 billion, plus they be showing profits and pulling in investors ?I think they will be always chasing big tech and buying good innovations from smaller companies.
The principal-agent problem here is textbook. Management overhired during ZIRP because the cost of capital was effectively zero and headcount growth signaled momentum to the board. Now AI gives them cover to correct that misallocation without admitting the original error. The real economic signal isn't that AI replaces workers -- it's that companies are using AI narratives to restructure balance sheets and re-price labor markets downward. The 67% drop in entry-level postings is the scariest data point because it suggests firms are pulling up the ladder entirely rather than repricing specific roles.
Mark Minevich so is it surprising that employees are resisting AI integration and for whom fear is driving actions? These self-inflicted negative consequences won't appear on anyone's reporting dashboard, but they are real. The role of people at work is being reshaped. It does not need to be done with cruelty, and given the large documented gaps between employees' desire to be retrained for this new era vs how employers are closing the gap there are better ways to pursue this transformation—with humanity.
I love that one "The CEO whose chips power every AI system on the planet said layoffs are a leadership failure. Not a technology constraint. His vision: “Every carpenter could now be an architect.”
Mark Minevich, if AI was not good enough to do these jobs until December (Andreessen's own timeline), then every company that attributed earlier layoffs to AI lied to its shareholders. That is a disclosure story more than a labor story. The (next) problem: these same companies **cannot verify that the AI systems they claim justify the cuts actually perform the functions they describe.** They fired the humans and replaced them with systems no one in the C-suite can audit. The junior analyst line is the sharpest point in your post. We are cannibalizing the pipeline and calling it efficiency.
Oracle can't pay back the debt with private credit funds I guess - the funds are taking a hit already and limited withdrawals
They would kill us if they thought they could get away with it.
You graph is wrong. In the last 12 month there have been ~ 45k for Oracle, first batch was last year, Aug/Sept, that was ~ 15k, I know cause I was one of them.