Purpose and judgment
People define why the problem matters, what outcome is acceptable, what trade-offs are legitimate and who remains accountable.
AI → SI transition · AFA framework · September 2026
Super Intelligence (SI)
AFA is transitioning from AI-first language toward Super Intelligence (SI): a human-centred way to understand what happens when intelligence gains computational reach.
The intelligence is not artificial.
The amplification is computational.
The capability becomes super.
And the responsibility remains human.
The amplification is computational.
The capability becomes super.
And the responsibility remains human.
Definition
What AFA means by Super Intelligence.
Super Intelligence · two words
Human capability extended.
Super Intelligence is the extension of human intellectual capability beyond the practical limits of unaided cognition through computation, accumulated knowledge, models, machines and other instruments of intelligence.
The centre of the definition is not a machine replacing the person. It is the practical reach of intelligence becoming larger.
Artificial intelligence / AI
The technical substrate remains real.
AFA continues to use artificial intelligence for models, machine learning, APIs, agents, standards, regulations, procurement categories and established vendor terminology.
AI does not vanish from the site. It becomes more clearly situated inside a larger human-centred frame.
A necessary distinction
Super Intelligence is not the same as “superintelligence.”
The established one-word term superintelligence is commonly used for an entity whose intelligence exceeds human intelligence. AFA is making a different proposition with Super Intelligence, two words: intelligence extended through instruments while human purpose and responsibility remain central.
Reach and leverage
Models, agents, software, data and compute expand search, comparison, synthesis, simulation, generation and operational scale.
Action through machines
Robotics, sensors and embodied systems allow computational intelligence to enter the physical world. Ideas acquire hands.
What this means for AFA
The purpose stays the same. The architecture becomes clearer.
AFA still exists to help people and institutions use advanced technology responsibly and creatively. The new frame connects the six parts of the company around one question: what can people accomplish when intelligence is extended well?
Atkinson Academy
Develop judgment, literacy and the ability to direct intelligent systems well.
Intelligence Solutions
Turn AI models, agents and workflows into useful institutional capability.
Infrastructure & Discovery Labs
Give intelligence secure, appropriate computational foundations.
Marbles Cognitive OS
Connect knowledge, context, agents, software and eventually machines.
Physical Intelligence & Robotics
Extend intelligence into perception, movement and physical work.
Atkinson Studio
Use new instruments of intelligence to extend storytelling, simulation and imagination.
“The machine does not have to diminish the person. It can enlarge what a person can do.”
Common questions
Understanding the language change.
What does AFA mean by Super Intelligence?
It means human intellectual capability extended beyond the practical limits of unaided cognition through computation, accumulated knowledge, models, machines and other instruments of intelligence.
Is this the same as “superintelligence”?
No. AFA uses Super Intelligence as two words for a human-centred framework. The established one-word term “superintelligence” usually refers to an intelligence that exceeds human intelligence.
Is AFA abandoning the term AI?
No. AI remains the appropriate technical term for models, machine learning, agents, APIs, standards, regulations, vendor products and much of current search and procurement language.
How does robotics fit the framework?
Robotics gives computational intelligence the ability to act in the physical world. For AFA, that makes boundaries, safety, governance and human accountability even more important.
The transition
Why both terms will appear for some time.
The vocabulary used by governments, researchers, universities, standards bodies, vendors, search engines and procurement teams will not change at the same speed. AFA therefore uses a deliberate bridge rather than a global find-and-replace.
Super Intelligence first
Use the new frame when talking about human capability, education, institutional transformation, Marbles, robotics and the larger purpose of the technology.
AI remains precise
Keep AI where it identifies a field, model category, vendor product, regulation, standard, API, procurement term or established technical practice.
Bridge rather than erase
Maintain established AI terminology in metadata and high-intent service language while building clear semantic relationships to Super Intelligence.