Amazon, Google See High Turnover, Not Mass Layoffs, Blind Data Shows

TEXT START: As artificial intelligence advances rapidly, U.S. big tech companies are diverging sharply in whether their headcounts are growing or shrinking.

The Dissection

The text converts selective hiring churn into evidence against mass AI displacement. It documents labor being pulled toward chipmakers and frontier-model firms while weaker companies bleed staff. That is not proof that employment survives. It is proof that AI concentrates productive power and reallocates a shrinking share of valuable labor toward the firms controlling the new machinery.

The Blind sample tracks users who changed employers, not total employment, hours, wages, involuntary exits, eliminated roles, or workers who vanished from the platform because they no longer found a comparable job. A revolving door among elite technology workers is not the labor market. It is the visible movement of survivors inside the narrowing core.

The Core Fallacy

The central error is confusing workforce composition with productive participation.

Under the Discontinuity Thesis, AI does not need to produce universal layoffs immediately. It only needs to make cognitive labor cheaper and strategically subordinate to AI capital. Companies can simultaneously hire scarce AI specialists, remove ordinary knowledge workers, automate functions, and preserve or even expand aggregate headcount temporarily through transition hiring.

The article treats incoming workers as evidence that AI creates jobs. More accurately, it shows that AI creates demand for owners, builders, and controllers of AI systems while eroding demand for the labor those systems replace. This is labor concentration, not system health.

Hidden Assumptions

• That employee inflows and outflows approximate total job creation and destruction.
• That workers leaving one firm can reliably find equivalent work elsewhere.
• That hiring AI specialists offsets the displacement of the much larger population whose tasks are being automated.
• That company-level headcount stability implies preservation of the wage-to-consumption circuit.
• That voluntary job changes are separable from structural pressure and declining prospects.
• That business strategy can preserve human labor demand indefinitely rather than merely determine which firms capture the transition rents.
• That “not mass layoffs” means “not mass obsolescence.” It does not.
• That the Blind population represents workers generally. It mainly represents a self-selected, credentialed technology stratum.

Social Function

Classification: partial truth, transition management, and ideological anesthetic.

The partial truth is real: AI leaders are hiring aggressively, and firms with weak AI positions are losing personnel. The anesthetic is the framing. By spotlighting movement among recognizable companies, the text reassures workers that the market is merely rearranging opportunities and instructs them to choose the right employer. It turns a structural ownership crisis into an individual placement problem.

Its practical message is useful only to workers near the remaining frontier: move toward AI capital, acquire scarce control or verification capabilities, and avoid roles exposed to automation. For everyone else, “look beyond the company name” is career-market advice issued from inside a burning building.

The Verdict

The article does not refute AI job-loss fears. It records the early labor-market signature of the Discontinuity Thesis: concentrated hiring at AI-capital firms, attrition at laggards, and rapid sorting of workers by proximity to the new productive core.

The system is not preserving mass employment. It is selecting a smaller number of AI owners, builders, and indispensable servitors while using turnover to conceal the broader contraction. The revolving door is not evidence that the machine still needs everyone. It is the sound of the machine deciding whom to keep.