A mystery letter about Q* may have sparked OpenAI's chaos

Just when you thought the OpenAI boardroom drama had reached peak Silicon Valley absurdity, a new detail emerged about what actually lit the fuse.

Ahead of Sam Altman's chaotic four-day exile, several staff researchers sent the board of directors a warning letter about a powerful AI breakthrough that they believed could threaten humanity, according to two sources familiar with the matter reported by Reuters.

The mystery of Q*

Sources say long-time executive Mira Murati told employees that a letter detailing a project called Q* (pronounced Q-Star) precipitated the board's abrupt move to oust Altman.

Inside the company, some researchers viewed Q* as a major leap forward in OpenAI's pursuit of artificial general intelligence, which the company defines as AI systems smarter than humans.

Fueled by massive computing resources, the new model was capable of solving certain mathematical problems. While the feats were only at the level of grade-school math tests, researchers were extremely optimistic about its trajectory. Solving math is fundamentally different from generating text, because it requires genuine reasoning toward a single correct answer. Reuters noted it could not independently verify these claimed capabilities, and OpenAI declined to comment.

72 hours of chaos and leverage

While the letter was only one factor among a list of board grievances, it catalyzed a wild chain of events. President Greg Brockman quit in protest, and the board scrambled to name former Twitch boss Emmett Shear as interim CEO.

Shear later posted on X that he worked roughly 72 intense hours to stabilize the ship. But the board completely lost its leverage once Microsoft stepped in. Satya Nadella offered to hire Altman, Brockman, and anyone else who wanted to leave.

Armed with the promise of Microsoft's compute power, over 700 OpenAI employees signed a letter threatening a mass exodus unless the board resigned and reinstated Altman. By Tuesday night, just ahead of the Thanksgiving deadline, Altman reclaimed the CEO seat.

Why it matters

This revelation pulls back the curtain on the actual ideological civil war inside AI labs: raw velocity versus existential fear.

If Q* represents a genuine leap in machine reasoning, the implications for software development, scientific discovery, and automated problem-solving are enormous. Current large language models are essentially predictive text engines; an architecture that masters math points toward reasoning engines that can verify their own truth. But you will not be trying this anytime soon, as it remains strictly an internal research project under immense scrutiny.

In our view, the board's execution was catastrophic. Rather than establishing clear, transparent safety protocols, a few directors panicked over an internal research milestone and triggered an uncoordinated coup.

It looks like the coup backfired completely. By pushing the company to the brink of collapse, the board ended up transferring virtually all leverage to the commercialization wing backed by Microsoft. The lesson Silicon Valley took away is clear: once the commercial flywheel is spinning at this scale, safety concerns expressed through opaque boardrooms can be steamrolled over a single holiday weekend.