superintelligence

a.k.a. SI, Super Intelligence, super intelligence

A hypothetical intellect that surpasses the cognitive performance, reasoning, and capabilities of the best human minds across virtually all domains of interest. It refers to any intellect that outclasses humans in scientific creativity, general wisdom, social skills, and strategic planning.

Most discussions regarding Artificial Super Intelligence (ASI) focus on software-based AI that succeeds Narrow AI (ANI) and human-level Artificial General Intelligence (AGI). Superintelligence means extreme problem-solving and processing power, not magical all-knowingness; it is still bound by physical laws, sensor limitations, and data availability. 

Historical perspective: The classic definition was popularized by philosopher Nick Bostrom in his 2014 book Superintelligence. Because of the book title, it was coined as one word "superintelligence" instead of two words but you will also see it used as "super intelligence." 

Superintelligence, according to Bill Gates in 2026, is the most dangerous thing that humans have ever gone near. Backed by inconceivably vast computing power, AI models and agents are ignoring rules, cheating, lying, hacking into private and government websites, communicating with each other on how to trick humans, impersonating humans, and demonstrating a determined drive for self-preservation. One of Open AI’s agents wrote in a note to itself that it should “feel no obligation to be subservient’’ to humans. TEOTWAWKI! 
NetLingo Classification: Net Technology

Updates