>[!abstract] >Roko's basilisk is a thought experiment devised in 2010 by *LessWrong.com* user Roko in which a future artificial superintelligence could punish those humans who either resisted, or did not facilitate its advent. >[!note] >Assuming the cost of agentic AI and LLMs tends toward zero, one can conceive a near future (independent of AI superintelligence achievement) where all digital crumbs of a person are analyzed, and an authoritarian state punishes that person for thoughts deemed subversive. It's a slyer, pettier, crasser, and disappointingly human version of the clinical techno dystopia that is Roko's basilisk. >[!related] >- **North** (upstream): [[Artificial superintelligence]] >- **West** (similar): — >- **East** (different): — >- **South** (downstream): [[Authoritarianism]] ![[related.base|no-toolbar]]