There are moments in history when a new instrument does not merely change what human beings can do.
It changes the language we need to describe what we are doing.
I believe we are living through one of those moments now.
For more than seventy years, we have used the expression Artificial Intelligence.
The term has an important history. In the 1955 proposal for the Dartmouth Summer Research Project on Artificial Intelligence, John McCarthy, Marvin Minsky, Nathaniel Rochester and Claude Shannon proposed investigating whether aspects of learning and intelligence could be described precisely enough for machines to simulate them.
That was an extraordinary idea.
It helped give a field its name before most of humanity could imagine what that field would eventually become.
Today, Artificial Intelligence — AI — is the established scientific, technical, commercial and regulatory terminology. We are not pretending otherwise.
But words matter.
And as these technologies become more capable, more connected and more deeply integrated into human work, I have begun asking a different question:
Does the word artificial still give us the best conceptual picture of what is happening?
I increasingly believe it does not.
The intelligence was never artificial.
The computation was.
And that distinction is larger than it first appears.
Before the machine, there was intelligence
Before there was a computer, there was intelligence.
Before there was a language model, there was language.
Before there was machine vision, there was sight.
Before there was a scientific database, there was science.
Before machines manipulated mathematics at enormous speed, human beings had spent thousands of years developing mathematics.
Before a model could generate a sentence, civilizations had spent thousands of years creating words, meanings, metaphors, stories, histories, laws, philosophies and systems of knowledge.
This does not mean modern computational systems merely repeat what people have already written.
They do not.
They can discover relationships.
They can identify patterns nobody deliberately inserted.
They can generate combinations nobody previously expressed.
They can explore possibility spaces too large for a person to search manually.
They can produce results that surprise the people who built them.
That capability is real.
But novelty does not erase inheritance.
The machine enters an already-existing world of meaning.
Language already means something.
Scientific concepts already refer to something.
Mathematics already provides systems of representation.
Images exist within cultures.
Questions matter because someone has a reason for asking them.
The machine enters what we might call a semantic civilization: an accumulated intellectual inheritance of language, knowledge, mathematics, science, culture, records, experiments, failures, discoveries and human experience.
And then computation gives us an extraordinary new ability to operate upon it.
That distinction matters.
Even the traditional definition begins with us
Cambridge Dictionary currently defines artificial intelligence in relation to computer systems or machines possessing some of the qualities associated with the human brain, including language, image recognition, problem-solving and learning from data.
Notice the reference point.
Human capability comes first.
We identify certain things human intelligence can do, and then we build computational systems capable of performing some of those functions in extraordinarily powerful ways.
Yet our everyday language can make it sound as though two completely separate intellectual worlds have appeared:
Human Intelligence
over here,
and
Artificial Intelligence
over there.
The two meet.
The two compete.
Perhaps eventually one replaces the other.
It is a dramatic picture.
I am no longer convinced that it is the most useful one.
There is another way to understand what we are building.
Think of Galileo
For thousands of years, human beings looked at the night sky and saw points of light.
Then came the telescope.
The same human eye could suddenly see moons orbiting Jupiter and features on the Moon. A universe that had always existed became perceptible in a different way.
What happened?
Did Galileo invent artificial sight?
No.
His sight had been extended.
The telescope did not replace the human eye.
It gave the eye reach.
The microscope did the same in the other direction.
We did not call what followed artificial perception.
We built an instrument that allowed human perception to enter a scale previously inaccessible to it.
Consider the engine.
A human body cannot lift a locomotive.
But a human being can command a machine capable of moving thousands of tonnes.
We do not normally describe that as artificial human strength.
Human purpose learned to command forces vastly beyond the natural limits of the unaided body.
And then came the computer.
At first the relationship was obvious.
Computers calculated.
They stored.
They searched.
They performed operations at speeds no person could reproduce manually.
Nobody needed to claim that mathematics itself had suddenly become artificial.
The machine extended our ability to calculate.
Then something remarkable happened.
The instrument reached the intellect.
Computers began extending capabilities that we associate much more directly with cognition:
language,
recognition,
prediction,
reasoning,
simulation,
generation,
discovery,
and increasingly, through robotics, physical action.
When an instrument changes that much, perhaps our language eventually has to change with it.
Why “Super” deserves another look
On September 22, 2026, President Donald Trump used the phrase “Super Intelligence” during his address to the United Nations, saying that what had been called Artificial Intelligence would be “hereinafter officially called ‘Super Intelligence.’”
The political implications of that declaration are separate from the question I am interested in here.
I am interested in the words.
Because super contains an idea that may prove useful.
Cambridge defines the prefix super- in ordinary English as something larger, more effective, more powerful or more successful than usual. A supercomputer is still a computer. It has simply crossed an important threshold of capability.
That gives us another way to understand what is happening.
Perhaps Super Intelligence does not have to mean that a machine has become an independent superior mind.
Perhaps it can describe something more immediately relevant:
human intelligence acquiring capabilities beyond the practical limits of unaided human cognition.
That is a different proposition.
And it is why Atkinson Film-Arts is deliberately beginning to use Super Intelligence as two words.
Tony Stark is still Tony Stark
Comic books provide a surprisingly useful analogy.
Consider Tony Stark.
Remove the suit and he remains a human being: brilliant, imperfect and biologically finite.
Put him inside an extraordinary technological system and suddenly his effective capabilities change.
He can perceive farther.
Calculate faster.
Communicate across enormous distances.
Process streams of information.
Navigate complex environments.
Command machines.
Fly.
Lift forces his body could never lift alone.
Did the suit abolish Tony Stark?
No.
Did Tony Stark cease to matter?
Quite the opposite.
The story remains about the person.
The capability becomes super.
Tony Stark did not need to become a new species in order to become a superhero.
The technology altered what an existing human intelligence and human purpose could accomplish.
That is increasingly how I think about the relationship between people and advanced computational systems.
The useful analogy may not be:
human mind versus machine mind.
It may be:
human mind with an intellectual exoskeleton.
A cognitive exoskeleton
Consider what happens when a person uses one of these systems well.
You ask a complicated question.
The system produces an answer.
You recognize something that is missing.
You challenge an assumption.
You introduce context.
The system can then search, compare, synthesize and reformulate information on a scale that might have taken you days to examine manually.
You read the result.
Perhaps it is better.
Perhaps it is still wrong.
You push again.
You change the question.
You identify another distinction.
The computational system explores the revised problem.
Eventually you can arrive somewhere neither the first human formulation nor the first computational response had reached.
That interaction tells us something important.
The first machine answer was not automatically true because a powerful computer generated it.
The person supplied judgment.
Purpose.
Context.
Correction.
Nuance.
The computation supplied reach.
Search.
Comparison.
Synthesis.
Iteration.
Speed.
The person's effective intellectual reach became larger.
That is not merely automation.
It is intellectual leverage.
The data centre changes the scale
This is where the distinction becomes even more important.
An unaided human brain has practical constraints.
Attention is limited.
Working memory is limited.
Time is limited.
Reading speed is limited.
A person cannot simultaneously investigate millions of alternatives.
And our biological brain does not acquire thousands of additional processors because another data centre is constructed.
Computational infrastructure is different.
Processors can be added.
Memory can expand.
Models can improve.
Networks can connect specialized systems.
Multiple agents can operate simultaneously.
Sensors can extend perception.
Robotics can extend action.
New data centres can increase the pool of available computation.
None of this makes computation unlimited. Energy, hardware, economics, algorithms, physical infrastructure and reliability all impose real constraints.
But the architecture is fundamentally scalable in a way the unaided individual human mind is not.
And one person may increasingly gain access to a portion of that infrastructure.
That creates a potentially enormous gap between:
what a person can accomplish unaided
and
what that same person can accomplish when intellectually extended through computation.
The word amplification describes an increase.
But an increase can be tiny or enormous.
Turning a radio from volume five to volume six is amplification.
Broadcasting a human voice across a continent is also amplification.
At some point the magnitude matters.
That is where super becomes useful.
The intelligence does not have to originate in the machine
This does not require us to deny the extraordinary computational contribution.
Quite the opposite.
The computational system may do work that no human being could practically reproduce.
It may search spaces no human being could manually examine.
It may identify relationships that no particular person noticed.
It may construct intermediate strategies.
It may generate genuinely novel outputs.
And increasingly it may operate for significant periods without human approval at every individual step.
But those facts do not force us to conclude that the most useful philosophical description is an intelligence detached from humanity.
A system may increasingly determine how to pursue an objective.
That is different from determining the ultimate why.
Why was the system built?
Why does the problem matter?
Why was one objective selected instead of another?
Why should one outcome be considered desirable?
What authority should the system possess?
Who is accountable when its decisions affect another human being?
Those remain questions of human purpose, institutions, governance and responsibility.
Operational autonomy is real.
But operational autonomy is not automatically the same thing as independence from humanity.
The more powerful the instrument, the more consequential judgment becomes
There is an understandable assumption that as computational systems become more capable, human judgment must become less important.
I think we should consider the opposite possibility.
A weak system making a mistake may inconvenience us.
A vastly powerful system making a mistake may multiply the consequences of that error.
A model can produce an elegant answer that is wrong.
It can misunderstand a premise.
It can fabricate information.
It can optimize the wrong interpretation of an objective.
It can reason coherently within incomplete context.
Greater computational capability therefore does not eliminate judgment.
It can make good judgment more consequential.
The lesson should not be:
Stop thinking because the computer can think for you.
It should be:
Learn to think well enough to direct increasingly powerful instruments of thought.
That may become one of the defining human skills of this century.
Education changes
For centuries, education has asked:
What should a person know?
That question remains important.
But another question now stands beside it:
What can a person accomplish when they know how to direct computational intelligence well?
The calculator did not make mathematics irrelevant.
It changed which mathematical abilities became most valuable.
Search engines did not make knowledge irrelevant.
They changed how we navigate knowledge.
Super Intelligence should not make thinking irrelevant.
It may make good thinking dramatically more valuable.
The person who frames the better question may direct extraordinary computational resources toward answering it.
The person who recognizes the hidden error may redirect an entire chain of automated work.
The person who understands the objective may accomplish work that once required a large institution.
That does not represent the disappearance of human intelligence.
It may represent one of the greatest expansions of its practical reach in history.
Super Intelligence is not the same as “superintelligence”
This distinction must be explicit.
The established English noun superintelligence, normally written as one word, already has a meaning.
Cambridge defines it as an entity—for example, a computer—with a level of intelligence higher than humans.
That is a legitimate and important concept.
It is not the concept I am defining here.
Atkinson Film-Arts is deliberately using:
Super Intelligence
as two words.
Our proposed definition is:
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.
It does not require us to claim that today's computers are conscious.
It does not require us to claim that a machine has become an independent person.
It does not require us to claim that every computational system is more intelligent than every human being.
It describes something else:
intelligence whose effective reach has become super through technological extension.
This is a conceptual proposal, not a claim that scientific terminology has already changed
That point matters.
Artificial Intelligence remains the established name of the field.
Indeed, the original 1955 Dartmouth proposal explicitly framed the project around making machines use language, form abstractions and concepts, solve problems associated with humans and improve themselves. The field has inherited that vocabulary for seven decades.
Atkinson Film-Arts is not claiming that dictionaries, universities, standards bodies or the technology industry have suddenly abandoned AI.
They have not.
Nor are we suggesting that researchers should simply replace established terminology in technical contexts where precision and interoperability depend on shared language.
Instead, we are making a conceptual proposal.
Artificial Intelligence describes the historical technological field.
Super Intelligence describes the broader human-centred phenomenon we believe is now becoming visible as those technologies mature.
One asks:
What intelligent capabilities can machines perform?
The other asks:
What can intelligence now accomplish when human beings can extend it through computation?
Those questions overlap.
They are not identical.
Why you will still see “AI” throughout this website
We are also living through a transition in language.
The world currently searches for:
Artificial Intelligence.
AI.
Generative AI.
AI agents.
AI consulting.
AI infrastructure.
AI robotics.
AI education.
AI for healthcare.
AI for government.
AI for filmmakers.
These are the established terms people understand.
They remain the terminology used by universities, governments, technology companies, standards, procurement systems, search engines and software products.
So Atkinson Film-Arts will continue to use Artificial Intelligence and AI throughout this website where they provide technical clarity and shared understanding.
We are not interested in creating confusion merely to introduce a new term.
Instead, over time, visitors will increasingly see the two concepts together:
Artificial Intelligence (AI) and Super Intelligence (SI).
Then increasingly:
Super Intelligence — commonly described today through the language of AI.
And as the language and technology evolve, our terminology can evolve with them.
Our intention is not to erase seventy years of technical history.
It is to describe what we believe that history is becoming.
Every great instrument extended something human
The history of technology is, in one sense, the history of human beings refusing to remain confined by the limits of the unaided body.
We could not move enough weight.
So we built machines.
We could not see far enough.
So we built telescopes.
We could not see small enough.
So we built microscopes.
We could not travel fast enough.
So we built trains, automobiles and aircraft.
We could not calculate fast enough.
So we built computers.
We could not communicate far enough.
So we built telecommunications and the internet.
And now we are extending something even more fundamental:
the practical reach of the human intellect.
The machine does not have to diminish the person.
It can enlarge what a person is capable of doing.
A child can enter intellectual territory once accessible only to specialists.
A physician can interrogate bodies of knowledge too large for any individual to memorize.
A scientist can explore enormous spaces of possibility.
An entrepreneur can command capabilities once available only to large institutions.
An artist can work with instruments that did not exist a generation ago.
An engineer can move rapidly between idea, simulation and iteration.
And through robotics, intelligence increasingly gains a new relationship with the physical world.
Ideas acquire hands.
The question before us
One of the great questions of this century will certainly be:
Will machines eventually become more intelligent than human beings?
That question is real.
The established concept of machine superintelligence deserves serious scientific, philosophical and ethical examination.
But another question is already before us:
What happens when ordinary human beings gain access to capabilities far beyond what an unaided individual has ever possessed?
That question may prove just as consequential.
It concerns education.
Medicine.
Science.
Government.
Business.
Creativity.
Robotics.
Opportunity.
Democracy.
And ultimately human agency.
Technology can concentrate power.
It can also distribute capabilities that were previously scarce.
Nothing guarantees which outcome we will produce.
That depends upon what we build.
Who receives access.
What we teach people to do with these systems.
How we govern them.
What we choose to automate.
What we refuse to automate.
And whether human beings remain willing to accept responsibility for the consequences.
That is the work ahead.
Super Intelligence
So I return to the word.
Artificial Intelligence served—and continues to serve—an extraordinary purpose.
It gave a name to a field before most of the world could imagine what that field would become.
But perhaps the maturation of that field is revealing something larger.
Not merely an artificial substitute for human intelligence.
Not merely another machine.
But an extension of the intellectual reach of humanity itself.
A cognitive exoskeleton.
A telescope for thought.
A microscope for knowledge.
An engine for ideas.
And through robotics, a bridge between thought and physical action.
Tony Stark did not stop being Tony Stark when he entered the suit.
The instrument did not abolish the person.
It extended the person's effective capability beyond the ordinary human range.
The super described what became possible.
That is why I believe Super Intelligence deserves serious consideration as a conceptual frame for the era we are entering.
Not because the human being is disappearing.
Because the human being may be gaining capabilities we have never possessed before.
The great challenge will not simply be whether we can build the technology.
It will be whether we possess the judgment to use it well.