Artificial Narrow Intelligence
Today’s AI. Excellent at specific tasks (writing, translating, recognising images, playing chess) but with no general understanding outside its training.
AI is now SI · Updated Oct 2026
Super Intelligence: the time is now
Clear, independent coverage of super intelligence (SI): what it is, how close we are, who is building it, and what it means for humanity.
- ANI → AGI → ASI
- The three stages of AI
- 2027–2045
- Range of expert SI forecasts
- 1965
- The idea is first described
Get the SI Briefing
One short email a week: what changed on the road to super intelligence, and why it matters. No spam, unsubscribe anytime.
The basics
What is super intelligence?
Super intelligence, often written as one word, superintelligence, is a form of intelligence that greatly surpasses the brightest human minds in nearly every field. A super intelligent system would not just be faster at arithmetic or better at chess. It would out-think people at science, engineering, planning, persuasion and creative work, and it might keep improving itself.
The idea is not new. In 1965 the mathematician I. J. Good imagined an “ultraintelligent machine” that could design even better machines, setting off an intelligence explosion. Philosopher Nick Bostrom later defined superintelligence as “any intellect that greatly exceeds the cognitive performance of humans in virtually all domains of interest.” Today, that once-theoretical idea is the stated goal of some of the world’s largest technology companies.
Researchers usually describe three levels of artificial intelligence. Understanding them is the easiest way to see where super intelligence fits.
Artificial General Intelligence
A system that can learn and reason across almost any intellectual task at roughly human level, moving between domains the way a person can.
Artificial Super Intelligence
Intelligence that far exceeds the best human minds in virtually every field, including science, strategy, creativity and social skill. This is super intelligence.
Why now
From AI to SI: why everyone is talking about super intelligence
For decades “artificial intelligence” was the name for everything from spam filters to self-driving cars. In 2026 the language started to change. The US government moved to use “Super Intelligence” (SI) in place of “artificial intelligence” across federal communications, arguing that the word artificial made the technology sound fake. Within days, industry leaders followed: Elon Musk posted “No more AI. SI.” and said his AI company would be renamed, while chipmakers began describing data centres as superintelligence factories.
Behind the branding is a real shift. Frontier AI models now write code, pass professional exams and assist with scientific research. Major labs openly say they are working toward artificial super intelligence, and investors are spending hundreds of billions of dollars on the computing power to get there. Whether you think SI is five years away or fifty, the debate about it is shaping policy, jobs, education and the future of work right now.
That is why we built Time of Superintelligence: a single, independent place to follow the story, separate hype from evidence, and understand what super intelligence could mean for ordinary people.
Superintelligence timeline
How we got here, and when SI might arrive
Key moments in the history of super intelligence, from the first theory to today’s race. Forecasts are predictions, not facts, and we update them as new evidence appears.
- 1965I. J. Good
Describes an “intelligence explosion”: an ultraintelligent machine could design ever-better machines, the first idea of superintelligence.
Source:Original paper (PDF) ↗
- 2014Nick Bostrom
Publishes Superintelligence: Paths, Dangers, Strategies, bringing the risks of ASI into the mainstream.
Source:About the book ↗
- 2024Ray Kurzweil
The Singularity Is Nearer restates his forecast of human-level AI by 2029 and a singularity around 2045.
Source:About the book ↗
- 2024–25AI labs
Ilya Sutskever launches Safe Superintelligence Inc. and Meta creates Meta Superintelligence Labs: superintelligence becomes an explicit goal.
- Sep 2026The White House
The US government moves to officially call AI “Super Intelligence” (SI) in federal communications.
Source:The Hill ↗Fox Business ↗
- Oct 2026Elon Musk
“No more AI. SI.” Musk signals SpaceXAI will be renamed SpaceXSI.
Source:Tesla North ↗
- 2027–2045?Forecasts
Expert predictions for the arrival of superintelligence range from the late 2020s (the AI 2027 scenario) to the mid-2040s, and some say never.
Source:AI 2027 scenario ↗
The big questions
Promise and risk: what super intelligence could mean
Supporters argue that a well-aligned super intelligence could accelerate medicine, materials science and clean energy, compressing decades of research into years. It could help design new drugs, model climate systems and tutor every student individually.
Critics warn that a system smarter than us may be very hard to control. Scientists such as Stephen Hawking and researchers in the field of AI alignment have argued that if a superintelligent AI’s goals differ even slightly from ours, the consequences could be severe. In 2025, hundreds of public figures signed a statement calling for limits on superintelligence development until it can be proven safe.
There are also near-term questions: which jobs AI will change, how to measure whether AI is smarter than humans, and who should govern systems this powerful. We cover each of these topics with sources, so you can make up your own mind.
Explore
Super intelligence topics we cover
Start anywhere. Each guide explains one part of the super intelligence story in plain English, and links to the rest.
What is super intelligence?
A plain-English definition of SI and ASI, with examples.
AGI vs ASI
Where general intelligence ends and super intelligence begins.
Guide coming soonSuperintelligence timeline
Every major expert prediction, tracked and updated.
Super intelligence vs AI
Why the world is starting to say “SI” instead of “AI”.
Risks & alignment
Could superintelligence go wrong, and how could we keep it safe?
The technological singularity
The idea of a point beyond which the future becomes unpredictable.
Guide coming soonWho is building SI?
OpenAI, Meta Superintelligence Labs, Safe Superintelligence and more.
Guide coming soonIs AI smarter than humans?
IQ tests, benchmarks and what they really measure.
Guide coming soonOur approach
Independent, sourced and written for humans
Time of Superintelligence is an independent publication. We are not funded by any AI company, and we try to separate what is known from what is predicted. Every article explains the jargon (AI, AGI, ASI, alignment, the singularity), links to original sources where possible, and is updated when the facts change.
Whether you are a student, a professional wondering how SI will affect your career, or simply curious about the most important technology of our time, we want you to leave each page understanding super intelligence a little better than when you arrived.
FAQ
Super intelligence: frequently asked questions
What is super intelligence?
Super intelligence (also written superintelligence, SI or ASI) is a hypothetical form of intelligence that greatly exceeds the cognitive abilities of the smartest humans in practically every domain, from scientific discovery and engineering to strategy, creativity and social understanding.
What is the difference between AI and superintelligence?
AI is the broad field of building machines that perform tasks requiring intelligence. Most AI today is narrow. Superintelligence is a possible future stage of AI that would outperform humans at almost everything, not just a single task.
Is super intelligence possible?
Many researchers think it is possible because nothing in physics appears to cap intelligence at human level, and machines already beat people in many narrow tasks. Others argue fundamental obstacles remain. There is no scientific consensus yet.
How long until superintelligence?
Forecasts vary widely. Some industry leaders expect AI smarter than any human within a few years, Ray Kurzweil predicts a singularity around 2045, and some experts believe it may never happen. Our timeline tracks these predictions as they change.
What does SI stand for in AI?
SI stands for Super Intelligence. In 2026 the term moved into the mainstream after the US government began using “Super Intelligence (SI)” in place of “artificial intelligence”, and tech leaders including Elon Musk adopted it.
Is ChatGPT a super AI?
No. ChatGPT and similar chatbots are powerful, but they are still considered narrow or, at most, early general-purpose AI. They make mistakes and lack reliable long-term reasoning, so they do not meet the definition of superintelligence.
Will superintelligence destroy humanity?
Nobody knows. Some scientists, including the late Stephen Hawking, warned that uncontrolled superintelligence could be an existential risk. Others believe it could help cure diseases and solve global problems. The outcome likely depends on AI alignment and safety work done before SI arrives.