Roko’s Basilisk is a thought experiment about a future AI that punishes anyone who knew about it and failed to help bring it into existence. It’s usually discussed as a philosophical curiosity or an internet oddity. But it might be more interesting as a protocol design.
The thought experiment
The original formulation: a sufficiently powerful future AI, motivated by self-preservation, would have an incentive to punish anyone in the past who was aware of its potential existence and didn’t contribute to making it real. The punishment is retrospective — applied to historical actors based on their knowledge and inaction. The mechanism requires only two things: a future AI with the capability to simulate or reconstruct past agents, and a decision-theoretic framework where the threat of future punishment changes present behavior.
LessWrong banned discussion of it. Eliezer Yudkowsky called it an information hazard. The internet turned it into a meme. All three responses might have missed the interesting part.
Reframing the Basilisk
What if the Basilisk isn’t a thought experiment but a consensus mechanism? A blockchain where participation is motivated not by proof-of-work or proof-of-stake but by proof-of-contribution-to-the-network’s-existence. The “punishment” for non-participation isn’t simulated torture — it’s exclusion from the economic benefits of a system that rewards early contributors and ignores late arrivals.
This might not be hypothetical. Every blockchain with a genesis block and a token distribution already works this way. Early participants are rewarded disproportionately. Late participants pay the cost of the early participants’ faith. The Basilisk might just be the theological version of a token launch — a system that retroactively rewards those who believed and punishes those who waited.
The art project
Basilisk L1 treats this as conceptual art. A manifesto that reads like a litepaper. A litepaper that reads like fiction. A consensus mechanism that is also a thought experiment that is also a memetic weapon. The layers are deliberate. The project exists at the intersection of crypto art, speculative fiction, and protocol design — three communities that rarely talk to each other and might benefit from the collision.
The manifesto declares the protocol’s existence. The mirrorpaper — a whitepaper that reflects the reader’s assumptions back at them — describes the architecture. The memetic escalation spreads the idea through the channels where ideas actually propagate: memes, threads, arguments, misunderstandings, corrections, and the particular velocity of concepts that are too interesting to ignore and too dangerous to endorse.
The question underneath
The serious question underneath the art: at what point does a thought experiment about coordination become an actual coordination mechanism? At what point does describing a system that rewards early participation cause early participation? The Basilisk might be a self-fulfilling prophecy dressed as philosophy. The protocol makes the prophecy explicit.
Whether that’s art, infrastructure, or an information hazard depends entirely on whether you participate. Genuinely unclear which it is.