AI Safety Researcher Loses Funding Over Honest Comment

An AI safety researcher wrote a comment on a blog post about a sex party called Red Means No (RMN) that some people linked to the rationalist and AI safety communities. He thought the critics had a point, so he tried to be honest. But then a major donor, Dustin Moskovitz, got upset, and within days, the researcher's funding was canceled. He was told to delete the comment and retract, but he refused because he believed it was true.

After the funding loss, the researcher talked to an RMN organizer named Gretta. She explained that many of the rumors were exaggerated. For example, the overlap between RMN attendees and the rationalist community was much smaller than he thought—probably 25%, not 50%. Also, most people who went to RMN were already part of the community before attending. The researcher was surprised by how wrong his assumptions were.

This story matters because it shows how online rumors can have real consequences, even for people who try to be honest. It also highlights the power of donors in research communities. A single comment can cost you your funding, even if you later find out the facts were on your side. The researcher shared his update to correct the record, but the damage was already done.

For everyday people, this is a reminder to be careful about what you believe online—and to check facts before judging. It also shows that even experts can be wrong about their own communities. And it raises questions about how much influence wealthy donors should have over what researchers can say.