Ask HN: How would you feel if we nationalized Google?

TEXT START: A hypothetical US gov't, benevolent and markedly less corrupt than the current one, passes a Constitutional amendment to nationalize Alphabet, with the express intent to return Google Search to 2010 levels of efficacy and to remove all LLM functions from both its frontend and backend.

The Dissection

This is not fundamentally a nationalization proposal. It is an obituary for an internet whose search layer once appeared useful, neutral, and navigable. The hypothetical benevolent government is a protective shell around a restoration fantasy: freeze the old product, delete the new automation, and recover a lost information commons.

The thread correctly identifies monopoly power, degraded incentives, SEO pollution, political capture, and the danger of concentrating information infrastructure. But it keeps treating the problem as a defective company that could be repaired by changing its owner. The web itself has changed: it is more adversarial, commercialized, synthetic, centralized, and saturated with automated content. The 2010 algorithm was not a timeless artifact. It worked inside a different ecosystem.

The Core Fallacy

The central error is confusing a product regression with a structural transition.

Google's decline is not caused merely by corporate cynicism or the presence of LLM features. It reflects the economics of cognitive infrastructure: whoever controls indexing, ranking, distribution, and inference controls a high-value layer of social coordination. Nationalizing Alphabet would not restore the old search circuit. It would place the cognitive chokepoint directly inside the state apparatus.

Removing LLMs would also solve the wrong problem. LLMs are not the source of the terminal shift; they are the visible instrument of it. Under the Discontinuity Thesis, cognitive work is being automated because automated systems are cheaper, faster, and increasingly capable. Banning them from one search engine does not reverse that pressure. It merely creates a slower interface over a hostile, AI-contaminated web.

The market-only replies make the opposite error: they assume competition will naturally reappear despite distribution lock-in, scale economics, default placement, and network effects. Neither nationalization nor casual consumer choice addresses the underlying ownership of automated cognition.

Hidden Assumptions

• A benevolent, competent government can be created and remain benevolent through political turnover.
• Constitutional protection can prevent future capture, surveillance, censorship, or partisan use.
• Search quality can be defined by reference to a stable 2010 target.
• The old algorithm can survive today's SEO tactics, AI slop, incentive structures, and content economics.
• The web is still a coherent corpus rather than an increasingly synthetic and commercially manipulated environment.
• The problem is Google's ownership model rather than the broader concentration of cognitive capital.
• A public search engine can absorb crawling, indexing, ranking, compute, and maintenance costs without reproducing bureaucratic or political distortions.
• Removing AI features would restore usefulness rather than remove capabilities users now depend on.
• Breaking up Google or funding Firefox is sufficient to overcome entrenched infrastructure and distribution advantages.
• Public control of information infrastructure is safer than private control merely because the hypothetical state begins benevolent.

Social Function

Primary classification: ideological anesthetic and nostalgia-driven copium, with a valid partial truth.

The partial truth is serious: monopoly power has degraded search, weakened accountability, and made information access dependent on institutions optimized for extraction rather than public utility. The anesthetic is the belief that a cleaner owner can resurrect the previous equilibrium. That converts a system transition into a management problem and lets participants mourn the product instead of confronting the power shift.

The proposal also functions as transition management. It asks whether society can preserve a familiar human-facing interface while the underlying economy moves toward automated cognitive mediation. The answer is no. The interface can be cosmetically restored; the ownership structure and productive logic cannot.

The Verdict

The thread correctly detects the corpse but misidentifies the murderer. Nationalizing Google would not bring back 2010 search. It would nationalize the information chokepoint and hand future regimes a more powerful instrument of cognitive control.

The real choice is not old Google versus LLM Google. It is who owns automated cognition, who can verify or challenge it, and whether ordinary people retain any productive stake in the system. This proposal offers no Sovereign position and no durable Servitor leverage. It offers a benevolent custodian fantasy over a structural transition that has already begun.