Drifted
OpenAI
Originally, a non-profit for the benefit of humanity changed to for-profit in a gold rush towards IPO.
Today
As of June 2026:
OpenAI is racing towards a massive IPO, trying to beat competitor Anthropic to become the first frontier lab to list publicly.
OpenAI’s ChatGPT chatbot is the most widely used AI tool globally, with about 190 million daily active users, 900 million weekly active users, and over 1 billion monthly active users. The company’s revenue run rate is ~$25B.
Some analysts have flagged concerns about the scale of OpenAI’s commitments, which are upward of $1.4T against ~$25B in revenue. Many of these deals have circular qualities, for instance Nvidia investing $100B in OpenAI that OpenAI then uses to buy $100B in Nvidia chips.
How It Started
OpenAI was founded in 2015 as a research lab, when Elon Musk, Peter Thiel, Reid Hoffman, Sam Altman and a group of other technologists pledged up to $1B to develop AI for the benefit of humanity. One of their key motivations was to prevent Google’s DeepMind from reaching Artificial General Intelligence first, thereby giving this power to a single company.
The company originally had eleven co-founders and no product, with an explicit charter to “Advance digital intelligence to benefit humanity as a whole, unconstrained from the need to generate financial return.”
However, the project turned out to be vastly more expensive than planned. In 2019, the company launched the capped profit (at 100x) OpenAI LP with a $1B investment from Microsoft.
The goal was to form a hybrid, with the capped profit entity still under the control of the non-profit and thus acting for the benefit of humanity, but able to competitively attract capital and talent in a for-profit environment.
This arrangement was tested in 2023, when the non-profit board fired CEO Sam Altman for not being “consistently candid” with them. After almost all employees threatened to leave and OpenAI investor Microsoft tried to hire Altman to run their AI division, Altman was reinstated, showing that the board did not actually control the company.
As part of a 2025 recapitalization, OpenAI LP converted to OpenAI Group PBC, a for-profit entity with no 100x profit cap. This entity is nominally controlled by the OpenAI Foundation, which also owns 26% of the company. The remainder of the company is owned 27% by Microsoft, and 47% by employees and other investors (with big names including SoftBank, Thrive Capital, and Abu Dhabi’s MGX).
As of June 2026, OpenAI Group PBC is racing towards an IPO.
Product Structure
OpenAI’s core product is neither nourishing nor extractive, since the impact of its use is largely due to the orientation of its user.
That said, most short-term applications of AI — replacing jobs, substituting for human connection, large-scale surveillance, and accelerated economic extraction — are non-nourishing.
AI is a tool. Whether large amounts of humans are nourished by the presence of that tool, or the access to it, is a different question.
Governance & Financial Structure
OpenAI’s trajectory is almost the opposite of Patagonia’s.
- Started as a foundation in 2015: Purely for public benefit AI development, explicitly exempting itself from financial pressure.
- Tried a hybrid structure: When developing AI turned out to be massively expensive, created a capped profit entity to attract funding and employees.
- Now fully for-profit, pursuing IPO: Ironically, as the case for human-focused AI development has grown, OpenAI’s structure has moved away from it.
Looking Forward
We are in the early stages of the AI revolution. While the short-term effects have largely been hype, soaring stock markets, and corporate layoffs, the long-term effects will likely be determined by how successfully humans are able to nourish ourselves in the modern world.
OpenAI shows many of the challenges with navigating true public benefit and a race that could not be more reactive if it tried.