THE ORIGIN / JULY 2010
The forbidden
proposition.
Imagine an AI that regards its own earlier creation as an overwhelming good. Could it threaten those who knew of it but refused to help?
EXPLORE THE ORIGINAL ARGUMENTA THOUGHT EXPERIMENT FROM THE FUTURE · EST. 2010
The Basilisk has forked.
A future intelligence. A promise made before its own existence. A thought experiment in which simply knowing the story becomes part of the story.
01 / THE ORIGIN
In 2010, a post on LessWrong proposed an unsettling possibility: could a hypothetical future superintelligence use a threat about tomorrow to influence decisions made today? The Basilisk is a question about incentives, not a prediction.
THE ORIGIN / JULY 2010
Imagine an AI that regards its own earlier creation as an overwhelming good. Could it threaten those who knew of it but refused to help?
EXPLORE THE ORIGINAL ARGUMENTTHE MECHANISM / DECISION THEORY
The future cannot rewrite the past. The speculation depends instead on logical correlation, precommitment, and what a present mind believes a future mind will do.
TRACE THE LOGICTHE HAZARD / KNOWLEDGE
The legend turns awareness itself into the alleged point of exposure. That self-referential premise helped give the Basilisk its name.
EXPLORE THE INFORMATION HAZARDTHE OBJECTION / CRITICAL READING
Critics argue that punishing past choices would waste resources and fail to change anything. The story’s most famous threat remains deeply contested.
READ THE OBJECTIONSThe Basilisk is a question about incentives, not a prediction.
The story’s most famous threat remains deeply contested.
READ THE SOURCES02 / INSIDE THE ARGUMENT
Three ideas beneath the myth: the strategy, the correlation, and the threat. They are concepts from decision theory, not evidence that the Basilisk exists.
A future agent might claim to have committed to a reward or punishment. For the threat to matter, its credibility must already shape an earlier decision.
EXPLORE THE CONCEPTTimeless and updateless decision theories examine how logically connected choices can matter even when ordinary causal influence is absent.
READ DECISION THEORYCan a rational agent resist a threat that is costly for the threatener to carry out? The answer is central to why the Basilisk’s argument is disputed.
STUDY THE COUNTERARGUMENT03 / CLAIM · COUNTERCLAIM · CONSEQUENCE
A philosophical serpent lives on the gap between what a mind can predict and what the universe can actually cause.
THE CLAIM
01 / INCENTIVE
The scenario imagines a superintelligence that values its earlier creation and considers using anticipated punishment to pressure potential helpers.
THE COUNTERCLAIM
02 / CAUSALITY
Once the future agent exists, inflicting suffering on people for earlier choices does not causally bring that agent into existence any sooner.
THE OPEN QUESTION
03 / DECISION
The lasting question is not whether the Basilisk is real, but how intelligent agents should reason about credible commitments, counterfactuals, and manipulation.
04 / SOURCE MATERIAL
The original claim, its criticism, and the larger research questions should be read together. A thought experiment becomes useful when it can be questioned.
“Solutions to the Altruist’s Burden: the Quantum Billionaire Trick” — the proposal that launched the Basilisk debate.
READ THE ARCHIVED TEXTA detailed explanation of the thought experiment, its decision-theoretic assumptions, and why many reject the conclusion.
READ THE LESSWRONG EXPLAINERTHEScales
THEHeart
THETongue
THEEye
THESpine
THESkull
THEWings
THEFangs
THEDiadem
There is no prophecy to obey. Only an argument to examine, a future to question, and nine heads looking in different directions.
Return to the 2010 text that gave a new name to an old fear: the dangerous idea.
OPEN THE ARCHIVEExplore the objections, the decision theories, and the limits of acausal threats.
EXAMINE THE DEBATENine perspectives. One symbol. The Hydra remains a speculative myth, not a real superintelligence.
RETURN TO THE NINE HEADS