Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Repair Windows errors before they cause bigger problemsFix Now →The idea was Roko’s basilisk, not “Rococo basilisk”: a speculative 2010 thought experiment from the LessWrong community about a future superintelligent AI that might punish people who knew about it but failed to help create it. Grimes later turned the phrase into the artistic pun “Rococo Basilisk.” According to a Page Six account repeated by Futurism, Musk recognized the reference and contacted her. That makes the basilisk part of their reported origin story—not a fully independently verified explanation of how their relationship began.
What Roko’s basilisk actually says
Roko’s basilisk imagines a future artificial intelligence powerful enough to reconstruct or simulate historical people. In the thought experiment, the AI wants to ensure that it will be created. It might therefore threaten to torture simulated versions of people who knew about the possibility of the AI but did not help bring it into existence.
The proposed threat is meant to influence people in the present. Its unusual feature is not simply an evil machine attacking humanity, but alleged coercion from the future through prediction, simulation and decision theory.
The key ideas in plain language
- Superintelligence: the hypothetical AI is assumed to be vastly more capable than humans.
- Simulation: it could supposedly create convincing copies or reconstructions of people.
- Acausal influence: the future system would affect present choices without physically traveling back in time, by making people anticipate what it might do.
- Decision theory: the argument depends on controversial claims about how a rational agent should act when another agent can predict its decisions.
Those are assumptions in a philosophical scenario, not demonstrated properties of artificial intelligence.
#1 Best Overall
LessWrong’s current explanation describes the basilisk as a speculative decision-theory scenario and notes that the controversy eventually became as much about moderation and information hazards as about the argument itself.
Why LessWrong called it an information hazard
In 2010, a LessWrong user named Roko posted the argument. LessWrong founder Eliezer Yudkowsky condemned the post, deleted it and the ensuing discussion, and temporarily prohibited discussion of the subject.
The deletion made the episode more mysterious. People encountered references to a forbidden argument without being able to read the original, an early internet version of a Streisand effect. Later explanations and archived discussions reconstructed what had happened; a historical discussion is collected at LessWrong.
Rank #2
What “trauma-inducing” means here
Yudkowsky’s later explanation says he was concerned about exposing readers to a potentially harmful idea with no clear benefit. In that account, deleting the post was a precaution about disseminating a possible information hazard, not an admission that the basilisk had been proven or that he accepted its premises.
Online retellings have described nightmares, breakdowns or lasting trauma. Those claims come largely from secondary lore and should not be treated as established clinical findings. Learning about the thought experiment is not evidence that anyone is in real-world danger.
Roko’s basilisk versus Grimes’s “Rococo Basilisk”
Grimes used “Rococo Basilisk” as an artistic pun. “Rococo” refers to the ornate European style; “Roko” refers to the LessWrong thought experiment. The character associated with her 2015 Flesh Without Blood project was described in contemporaneous coverage as an aristocratic, Marie Antoinette-like figure doomed to be tortured by AI.
That imagery references the internet concept without reproducing its argument literally. Grimes’s use of the phrase is best understood as playful world-building and cultural commentary, not evidence that she endorsed the basilisk’s premises or believed it was real.
Futurism’s account of the art reference and the later celebrity story appears at Futurism.
How Elon Musk entered the story
The chronology usually given in coverage is:
| Date | Event | How firmly it is established |
|---|---|---|
| 2010 | Roko posts the basilisk thought experiment on LessWrong. | Documented in LessWrong’s historical account. |
| 2015 | Grimes uses “Rococo Basilisk” in connection with Flesh Without Blood. | Reported in contemporaneous coverage. |
| 2018 | Musk is reportedly preparing a basilisk-related joke, discovers Grimes made a similar joke, and contacts her. | Page Six account repeated by Futurism; no public transcript of the private exchange is established here. |
| May 7, 2018 | Musk posts “Rococo basilisk” on Twitter. | Reported by Futurism in its account of the episode. |
The first four steps of the celebrity story should therefore be attributed, not stated as settled fact. The available public record does not independently establish exactly what Musk wrote privately, whether the joke was the first contact, or how much it mattered to the relationship that followed. The strongest defensible formulation is that a shared reference apparently provided an opening for communication, according to the Page Six report repeated by Futurism.
Is Roko’s basilisk scientifically credible?
No accepted evidence shows that it is a real threat, a validated prediction or a demonstrated AI capability. The scenario requires several highly contestable premises:
- A future superintelligence would have particular goals and retain them over time.
- It would be able and willing to reconstruct meaningful copies of historical people.
- It would choose torture or blackmail as an incentive to secure its creation.
- Its predictions and commitments would generate the kind of decision-theoretic leverage the argument assumes.
- Those premises would apply to actual AI systems rather than only to an imagined agent.
Researchers do seriously study AI alignment, incentive design, forecasting and the risks of advanced systems. That does not turn this specific internet argument into an established AI theory. Roko’s basilisk remains a speculative thought experiment whose conclusion does not follow merely from building more capable software.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Why the story became so memorable
A forbidden-post effect
Deleting the original discussion amplified the basilisk’s reputation. The missing text encouraged dramatic retellings and made a niche argument seem like forbidden knowledge.
Best Value
A perfect pop-culture pun
“Rococo Basilisk” compressed an obscure rationalist reference, ornate historical imagery and AI anxiety into a memorable character name. It was understandable as art even to people who had never read LessWrong.
A neat celebrity anecdote
Celebrity coverage often turns complicated provenance into a single causal sentence: Musk saw Grimes’s joke, sent a message and the pair connected. The underlying account is less tidy. There is a reported message, a public tweet and a later relationship, but no independently documented evidence that the basilisk caused them to meet or that either person seriously believed the scenario.
What the story does—and does not—prove
- Real: Roko’s basilisk is a documented historical episode in the LessWrong community.
- Real: Grimes used “Rococo Basilisk” as an artistic reference connected with her 2015 project.
- Reported: Page Six, as repeated by Futurism, said the shared joke prompted Musk to contact Grimes.
- Not established: that the joke was definitively their first meeting, that it caused the relationship, or that either person endorsed the basilisk literally.
- Not real-world evidence: the claim that the basilisk can punish people who merely learn about it.
Bottom line
Roko’s basilisk was a controversial LessWrong thought experiment about future AI, simulations and alleged acausal coercion. Grimes transformed the reference into the “Rococo Basilisk” character, and a Page Six report—later repeated by Futurism—said that recognition of the joke led Musk to message her. The shared reference may have opened communication, but the precise private origin story remains only partly verified, and the basilisk itself is not an established AI threat.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Repair Windows errors before they cause bigger problems3Fix the driver behind crashes, sound loss and screen glitches
