OpenAI widens its playbook

OpenAI is no longer acting like a company that only wants to release better models. In the span of a week, the company and its extended network showed a broader ambition. One group of former OpenAI operators surfaced with a new venture fund that is aiming for $100 million. OpenAI itself published a policy manifesto for the AI economy. The common thread is simple: it wants to influence what gets built, who gets funded, and how the gains from AI are divided (techcrunch.com, openai.com). The fund is called Zero Shot. TechCrunch reported on April 6 that it has already made a first close at $20 million toward a $100 million target and has already written a few checks. Its founders are not random ex-employees trading on a logo. Three came out of OpenAI’s early product and research orbit: Evan Morikawa, Andrew Mayne, and Shawn Jain. They are joined by investor Kelly Kovacs and operator Brett Rounsaville. Their pitch is that too much AI money is chasing flashy demos and not enough is going to products that match what the technology can actually do today (techcrunch.com). That matters because OpenAI already sits at the center of a giant commercial flywheel. On March 31, the company said it had closed $122 billion in new funding at an $852 billion post-money valuation. It described itself less as a lab than as infrastructure: consumer distribution through ChatGPT, developer demand through its APIs, enterprise deployment, and the compute to feed all three. It also said it is now generating $2 billion in revenue per month. A network of alumni investors backing startups that build around this stack is not a side story. It is one more layer of the platform strategy (openai.com). The policy push makes that strategy even clearer. On April 6, OpenAI published “Industrial policy for the Intelligence Age,” a document that argues ordinary policy tweaks will not be enough as AI gets stronger. The company says the point is to start shaping the institutions around advanced AI before the labor market and the tax base are forced to catch up the hard way. It is also putting money behind that effort, with fellowships, research grants of up to $100,000, up to $1 million in API credits, and a new workshop space opening in Washington, DC, in May (openai.com, cdn.openai.com). The striking part is not that OpenAI wants lighter regulation for itself. The striking part is that it is also arguing for redistribution. The policy document calls for a public wealth fund so citizens can share directly in AI-driven growth. Other reporting on the document says it also backs taxes on automation and trials of a four-day workweek as AI changes how much human labor is needed. That is a strange message from one of the companies pushing hardest toward more capable systems. It suggests OpenAI thinks disruption is no longer a distant thought experiment. It is near enough that the company wants to pre-write some of the response (cdn.openai.com, thenextweb.com, ibtimes.sg). This is consistent with the direction OpenAI has been moving for months. In 2025 it rolled out “OpenAI for Countries,” offering governments help with data centers, localized AI services, and startup ecosystem building under what it called “democratic AI rails.” In July 2025 it launched economic research with Ronnie Chatterji, Jason Furman, and Michael Strain to study productivity and labor effects, and tied that work to the same Washington workshop now being used as a policy hub. OpenAI is not just selling tools into the economy. It is trying to design the economy those tools will live in (openai.com, openai.com). That changes the signal for everyone around it. Founders now have another incentive to build “technically realistic” companies that fit the OpenAI stack well enough to attract alumni capital. Policymakers are being handed ready-made frameworks from the company with the most to gain if AI adoption accelerates. Investors are being told that the next durable edge may not be the best model alone, but the best position inside a system that includes capital, distribution, infrastructure, and rules. OpenAI’s own latest language makes the ambition plain: it says AI is becoming “core infrastructure,” and it is opening a Washington workshop in May to keep that argument moving from product launch to public policy (openai.com, openai.com).