OpenAI LP: Transition to Capped-Profit Structure

OpenAI has established OpenAI LP, a hybrid "capped-profit" company to accelerate the development of safe artificial general intelligence (AGI). This transition allows the organization to raise significant investment capital and attract talent with equity while ensuring that any returns beyond a specific cap are owned by the original OpenAI Nonprofit entity.

Capped-Profit Corporate Structure

OpenAI LP is designed as a hybrid of a for-profit and nonprofit entity to balance the need for massive capital investment with a mission-driven mandate. The organization determined that the most dramatic AI systems require immense computational power and algorithmic innovation, necessitating billions of dollars in investment for cloud compute, AI supercomputers, and talent retention.

Key structural features include:

  • Capped Returns: Investors and employees receive a capped return on their investment if the mission is successful. For the first round of investors, this cap is set at 100x their investment, with the expectation that future rounds will have lower multiples.
  • Nonprofit Control: OpenAI LP is controlled by the OpenAI Nonprofit board. The primary fiduciary obligation of OpenAI LP is to advance the goals of the OpenAI Charter.
  • Excess Value Distribution: Any monetary or value returns generated beyond the established caps are owned by the original OpenAI Nonprofit entity.
  • Governance Safeguards: Only a minority of board members can hold financial stakes in the partnership. Decisions where the interests of limited partners conflict with the mission—such as payout decisions—can only be voted on by board members without financial stakes.

Operational Focus and Capabilities

OpenAI LP employs approximately 100 people focused on three core areas of development:

  1. Capabilities: Advancing the functional abilities of AI systems.
  2. Safety: Ensuring AI systems are aligned with human values.
  3. Policy: Establishing appropriate governance for these systems.

While OpenAI Nonprofit continues to run educational programs (such as Scholars and Fellows) and host policy initiatives, OpenAI LP handles the technical development roadmap, building upon previous breakthroughs in reinforcement learning, robotics, and language.

AGI Safety and Risk Mitigation

OpenAI identifies significant risks associated with AGI, including machines pursuing misspecified goals, malicious human subversion of systems, and economic instability. To mitigate these risks, the organization is committed to the following:

  • Prioritizing Safety over Competition: As stated in the OpenAI Charter, the organization is willing to merge with a value-aligned organization to avoid a competitive race that would compromise safety priorities.
  • Mission Primacy: All investors and employees sign agreements stating that the obligation to the Charter takes precedence over their financial stakes, even if it means the loss of some or all of their financial investment.

Leadership and Governance

As of March 11, 2019, the board of the OpenAI Nonprofit consists of Greg Brockman (Chairman & CTO), Ilya Sutskever (Chief Scientist), and Sam Altman (CEO), alongside non-employees Adam D’Angelo, Holden Karnofsky, Reid Hoffman, Shivon Zilis, and Tasha McCauley. Elon Musk left the board in February 2018 and is not formally involved with OpenAI LP.

Sources