80,000 Hours Podcast
80,000 Hours Podcast

#209 – Rose Chan Loui on OpenAI’s gambit to ditch its nonprofit

One OpenAI critic calls it “the theft of at least the millennium and quite possibly all of human history.” Are they right? Back in 2015 OpenAI was but a humble nonprofit. That nonprofit started a for-profit, OpenAI LLC, but made sure to retain ownership and control. But that for-profit, having becom

Featured Speakers

The 80,000 Hours team Host

Topics Discussed

Episode Summary

Executive Summary: The episode examines OpenAI’s attempt to convert from a nonprofit-controlled hybrid structure into a normal for-profit company, and the legal, ethical, and governance consequences of that shift. Rose Chan Louie explains why the nonprofit’s control may be uniquely valuable, how compensation should be calculated, why cash matters more than locked-up equity, and how California and Delaware regulators could influence a fair outcome.

Main Topics: OpenAI’s original nonprofit-for-profit structure (Priority: 5/5): The discussion explains how OpenAI was intentionally designed to keep charitable purpose at the top while allowing investor capital below it, with the nonprofit board controlling the operating entities and limiting profit extraction. Why the proposed conversion is legally and morally fraught (Priority: 5/5): The hosts explore whether OpenAI is effectively trying to exit the commitments that justified public trust, donor support, and employee participation, while retaining the upside of the nonprofit structure’s early credibility. Valuing the nonprofit’s control and equity stake (Priority: 5/5): A major segment focuses on how to price the nonprofit’s control rights, residual profit interest, and intellectual property claims, including whether a control premium and auction-style valuation should apply. Cash versus equity in the compensation package (Priority: 5/5): The conversation argues that the nonprofit needs immediate cash, not just illiquid equity, because mission-critical AI safety and governance work must be funded now rather than years later. Regulatory oversight and attorneys general involvement (Priority: 4/5): California and Delaware attorneys general are portrayed as key checks on the transaction, with their approval and scrutiny helping protect charitable assets and enforce arm’s-length standards. Conflicts of interest and arm’s-length negotiation problems (Priority: 4/5): The episode highlights structural conflicts involving Sam Altman, investors, and staff equity holders, making a truly independent transaction difficult and putting pressure on the nonprofit board to act heroically. What the nonprofit could become after spinning out (Priority: 3/5): If it relinquishes control, the nonprofit may evolve into a large but more conventional foundation that must choose how best to deploy a potentially enormous endowment across AI safety and governance efforts.

Key Arguments: OpenAI’s nonprofit structure was not accidental; it was central to the organization’s public promise to prioritize humanity over shareholders. The nonprofit’s control rights may be more valuable than any simple dollar amount because they allow direct oversight of a potentially world-changing technology. A conversion that gives the nonprofit only illiquid equity would undermine its ability to act on its mission in the near term. The nonprofit and the for-profit are not aligned: every dollar or control right retained by the nonprofit is a cost to investors and staff with equity. California and Delaware attorneys general can help ensure the conversion is fair and that charitable assets are properly protected. A fair valuation likely requires a control premium and careful treatment of intellectual property, profit rights, and the probability of future AGI-scale returns. The nonprofit may be able to justify selling control only if the current structure is no longer sustainable and if doing so maximizes its ability to pursue its charitable mission. If the nonprofit receives substantial cash, it could independently fund AI safety, governance, and alternative research pathways more effectively than through its current embedded role.

Data Points: OpenAI nonprofit board size: 9 members - Described as including Sam Altman plus other appointees with different backgrounds. Board members appointed last year: 7 of 9 - The episode notes that most of the board was newly appointed recently. Initial fundraising goal: $1 billion - OpenAI’s early funding target as a nonprofit-backed research organization. Funds raised by 2019: $130 million - By 2019, the organization had raised only a fraction of its initial goal through donations. New funding round: $6.6 billion - New investors received a deal tied to completion of the conversion within two years. Conversion deadline tied to funding: 2 years - If the restructuring is not completed, OpenAI may need to return the $6.6 billion. Estimated nonprofit asset base: $19 million - Referenced as the nonprofit’s 2022 asset level, highlighting the imbalance with the for-profit side. Latest valuation mentioned: $156 billion - Used to illustrate the scale of the for-profit compared with the nonprofit parent. Lower-end compensation estimate: $37 billion - Mentioned as a possible figure for intellectual property and related interests. Alternative valuation estimate: $80 billion - Discussed as roughly half of a $157 billion valuation and as a possible fairer amount. Control premium range: 20% to 40% - General premium typically paid for a controlling interest in a business, potentially higher here due to strategic importance. Microsoft return cap: 100x investment - Microsoft’s contractual upside was described as capped at 100 times its investment. Global foundation comparison: Over $100 billion - Referenced as a benchmark for the largest nonprofit/endowment-style structures globally. Gates Foundation scale: About $50 billion - Mentioned as a major U.S.-based foundation by endowment size.

Pivotal Quotes: "No one person should be trusted here. I don't have super voting shares. The board can fire me." — Sam Altman: Used to illustrate OpenAI’s earlier emphasis on nonprofit governance and board control. "Whatever happens in OpenAI's next chapter, protects the charitable interests is likely to be a heroic task in the face of the overwhelming profit-making incentives." — Rose Chan Louie and colleagues: Quoted from their UCLA paper describing the structural conflict between charitable purpose and investor incentives. "It would be wise to view an investment in LLC in the spirit of a donation." — OpenAI operating agreement / transcript reference: Highlighted as evidence that early investors were told the venture might never behave like a normal profit-maximizing company.

Implications: The outcome could set a precedent for how mission-driven nonprofits embedded in commercial AI ventures are valued and governed. For listeners and the industry, the key issue is whether charitable purpose, control, and AI safety can be preserved when investor pressure pushes toward full commercialization.

🔓 Sign Up for Unlimited Episode Search

About 80,000 Hours Podcast

View all episodes from 80,000 Hours Podcast