eliezer yudkowsky deleted it from lesswrong. the white house just got briefed on it.

The Algorithmic Bridge piece about it opens: “To be clear, you shouldn’t read this.”

That’s the preamble to the Financial Times report that the White House has been briefed on Roko’s Basilisk — the thought experiment that a future superintelligent AI will retroactively torture everyone who knew about it but didn’t help build it.

Eliezer Yudkowsky deleted the original post from LessWrong in 2010 and banned discussion for five years — not because it was wrong, he said, but because he feared some unknown variant of it might actually work.

The moderation arc: rationalist forum ban → White House briefing.