In April 2026, OpenAI dropped a heavyweight document on the world. CEO Sam Altman, in an Axios interview, declared that superintelligence is arriving faster than most people expect — technology so powerful and disruptive that society will need a brand-new social contract to govern it. The accompanying official blueprint, Computing for the Age of Intelligence: With a Human-Centered Philosophy, runs 13 pages and for the first time systematically lays out six radical policy ideas for the AI era.
The Superintelligence Threshold
Altman revealed in the interview that OpenAI is transitioning toward superintelligence — AI systems that outperform the smartest humans even when those humans have AI assistance. Current models already help scientists make small discoveries, but the next generation (the rumored Spud) will leap qualitatively. Altman expects people to start saying: “this model helped me make the most important discovery of my decade — maybe of my career.” At the same time, the risks climb steeply. Altman warned that advanced cyberattacks and biological threats cannot be ignored, and that as AI advances, massive labor displacement becomes real: AI will surpass human labor.
The Six Radical Ideas
Six proposals from the official blueprint, aimed at keeping society functional through the superintelligence transition:
1. Public wealth funds. AI companies would fund national sovereign funds that distribute AI-created wealth directly to citizens — an explicit echo of the Alaska Permanent Fund’s oil-dividend model. The point: when AI generates enormous wealth, ensure ordinary people share the dividend rather than watching it concentrate in a handful of tech giants.
2. A robot tax. Tax automated labor and corporate AI gains to offset the coming employment shock, funding retraining and social programs for displaced workers. The blunt logic: when machines replace human work, who pays the people who lose income? The machines’ own wealth should.
3. Higher capital-gains taxes. Raise rates on the wealthy to finance the transition and redistribute AI-concentrated wealth — aimed squarely at the problem of AI-generated value pooling in a few companies (OpenAI, Google, Microsoft among them). The insight: the AI era may widen inequality dramatically, and taxation is the redistribution instrument.
4. The four-day work week. Incentivize four-day weeks without pay cuts, sharing AI’s productivity gains directly with workers — improving quality of life and mental health while distributing the automation dividend. The framing question: if AI makes productivity 10x, should humans work longer? No — work less, live more.
5. Universal AI access. Treat basic AI access — universal basic computing — as a fundamental right, on par with electricity or literacy. Everyone, especially underserved communities, should reach foundational AI models so no one is excluded from the AI-driven economy. In the intelligence age, those without AI access will be marginalized; access belongs in the infrastructure layer, like internet connectivity.
6. A rapid-response social safety net. Massively expand unemployment insurance, healthcare subsidies, and social security for workers permanently displaced by AI — with adaptive systems that scale up fast under economic shock and wind down once things stabilize. Traditional safety nets were built for gradual change; AI’s impact on employment may be sudden and severe, and the transition will not wait for legislative cycles.
Why This Matters
The blueprint’s clearest signal: AI is no longer the distant future; it is the happening present. The six ideas share one logical spine: admit reality — AI will displace enormous amounts of work and create enormous wealth; distribute fairly — via funds and taxes so the gains reach everyone; secure the floor — with a safety net so no one falls through; raise quality of life — shorter hours, more human freedom; and universalize the right — AI access as a basic entitlement. Altman calls for action now, so the world can safely harness this explosive new intelligence.
Reading the six ideas together reveals the blueprint’s real architecture. Ideas one through three are the money layer — a public wealth fund, a robot tax, and higher capital-gains rates form one coherent mechanism for routing AI-generated value toward the public. Ideas four through six are the human layer — the four-day week, universal AI access, and a rapid safety net address what people do with time, capability, and security once labor economics break. The two layers are deliberately paired: redistribution without access produces dependency, while access without redistribution produces a two-tier society of AI owners and everyone else.
Skeptics will note the messenger problem — the CEO of the company best positioned to capture AI’s gains proposing the tax regime that constrains it can be read as strategy as easily as sincerity. And the legal reality is unforgiving: sovereign wealth funds require legislation, robot taxes have failed in every legislature that considered them, and US federal preemption fights are already underway over state AI rules. But the blueprint’s function may be less legislative than normative — moving ideas like universal basic computing and automated-labor taxation from the fringe of the discourse into the center, the way universal service obligations migrated from utilities into telecoms. Policy windows open when the underlying facts move fast enough; Altman’s bet is that superintelligence moves them faster than any technology before it.
The Outlook
The document’s significance exceeds its policy specifics: it marks the moment the tech industry started seriously asking how society should organize when the technology genuinely changes everything. OpenAI’s answer: guide the transition through democratic process, giving people real power to shape the AI future they want. Whether these six ideas ever become law is almost beside the point — they supply a thinking framework for the question that matters: in the superintelligence era, what actually counts? References: OpenAI official — AI progress and recommendations; Sina Finance coverage of the Axios interview.
