Every age has its own picture of intelligence. AI is changing that picture again. So what does intelligence look like next?
An Old Question
Artificial intelligence has forced an old question back onto the table.
What do we actually mean by intelligence?
We use the word easily enough. We call people intelligent because they remember a great deal, solve difficult problems, speak well, understand numbers or succeed in demanding professions. Now we also use the word for machines that can write, analyse, calculate, create images, build software and answer questions.
But the moment we try to say exactly what intelligence is, the word becomes less cooperative.
To understand this a little better, let us take the example of two very different people.
The first is a senior financial executive sitting in a modern office. He may have studied at an elite university, understand global markets, manage billions of dollars and make decisions that affect companies, governments and millions of people.
The second is a member of the Hadza people living in a remote camp near Lake Eyasi in northern Tanzania. Some Hadza communities still live largely through hunting and gathering. They know how to read the land, find water, track animals, gather wild plants and move through an environment in which a mistake can have immediate consequences.
Who is more intelligent?
Most of us in the modern world would probably choose the executive. Our schools, institutions and measures of success all point us in that direction.
But now let us move the executive out of his environment.
Take away the office, the computer, the legal system, the financial markets, the employees and the institutions through which his decisions become powerful. Then place him in the bush near Lake Eyasi, where a group needs to find food, water and shelter before nightfall.
His résumé remains impressive. The land remains unmoved.
The most important person may now be the one who can see a faint track in the dirt, know which plants are safe, sense a change in the weather and lead everyone back to camp.
The executive has not suddenly become unintelligent. The Hadza community member has not suddenly gained an ability that was not there before.
What changed was the world around them.
And when the world changed, a different kind of intelligence came forward.
This is why AI has made the question so urgent. It is not simply adding a new machine to the old world. It is changing the environment in which human intelligence has to operate. If a new environment brings a different kind of intelligence forward, what kind is ours bringing forward now?
To see where intelligence may be going, it helps to look at where it has already been.
The Storyteller, the Scholar and the Specialist
We often speak as though intelligence were one fixed thing and humanity simply became better at finding and measuring it.
History tells a different story.
Imagine a society before writing. Its history, laws, family lines, religious teachings and practical knowledge all have to live somewhere. That somewhere is the human mind.
In such a world, the person with an extraordinary memory is not performing a party trick. They are carrying part of the civilisation inside them. If they fail to pass that knowledge on, the community does not lose a file. It loses part of itself.
Let us call this person the Storyteller.
The name may sound modest, but the role is not. The Storyteller is not merely providing entertainment around a fire. They are the living memory of the group. Their stories carry knowledge about people, place, danger, duty, identity and survival.
In an oral world, this is one of the highest forms of intelligence because the world needs someone who can keep knowledge alive.
Then writing arrives.
For the first time, knowledge can leave the human head and sit still. It can be marked onto a surface, stored, returned to and passed to someone who never met the person who first recorded it.
Memory does not become useless. But it no longer has to carry civilisation by itself.
And so another figure moves towards the centre: the Scholar.
The Scholar’s great strength is not simply holding knowledge. It is being able to read what has been stored, compare it, question it, interpret it and build upon it.
The world changed, and our picture of the intelligent person changed with it.
Then the world became more complicated.
Cities grew. Trade stretched across greater distances. Governments became larger. Law, medicine, administration, science and machines developed into worlds of their own. There was now too much for one person to know, so knowledge was divided into smaller and smaller parts.
The Specialist moved forward.
The doctor knew the body. The lawyer knew the law. The engineer knew the machine. The accountant knew the numbers. Each could go deeply into one field, master its language and solve problems that others could not.
For a very long time, the Specialist has been our main image of intelligence. We built schools to produce specialists, professions to organise them and institutions to reward them.
But dividing knowledge did not remove the need for a whole. It made the whole harder to see.
Once specialists became more prominent, so did the people who could organise their work. The senior manager, the director and eventually the chief executive became important because someone still had to connect the different domains, hold the collective picture and decide where the whole thing was going.
Intelligence was now working on two layers. One did the work within the parts. The other organised those parts into a whole.
We valued both.
And this was not a break in the story. It was the next step in it. The Storyteller carried knowledge, the Scholar worked with stored knowledge, the Specialist mastered a part of knowledge, and the senior manager pulled the parts together. Each figure became prominent because the world created a need for that kind of intelligence.
The shape changed because the environment changed.
And the environment never stands still.
Intelligence Does Not Stand Still
This movement begins to look less like a fixed definition and more like evolution.
A species changes over time within an environment. Traits that help it survive come forward. But the species is never really finished. It is simply the form we happen to see at one point in a much longer movement.
Trying to define intelligence once and for all may be like taking a photograph of a species halfway through its evolution and saying, “There. That is what it is.”
We can describe the form in front of us. We can ask what conditions produced it. We can trace where it came from and watch the pressures now acting upon it.
But we cannot honestly say that its present form is its final form.
The Storyteller, Scholar, Specialist and executive are not different species of human being. They are different human abilities brought forward by different worlds. The same person may carry all of them. What changes is which ability the environment needs most, trains most and rewards most.
That gives us a more useful way to think about intelligence. It is not only what sits inside a person. It is the meeting point between a human ability and a world that gives that ability something important to do.
That changing relationship—between people, tools, knowledge and the world around them—is what we mean by an ecology of intelligence.
And now AI is changing that ecology. It is making some abilities easier to access, weakening old shortages and creating new ones.
So the question is not simply, “Is AI intelligent?”
The deeper question is:
What does intelligence become in a world with AI?
The coder gives us a quick glimpse of the answer.
When the Coder Stops Looking Like the Answer
The coder became one of the defining specialists of the information age.
Computers were extraordinarily powerful, but they required us to speak to them in a language they could understand. Someone had to turn human intention into syntax, and the people who could do that became rare, difficult to replace and extremely valuable.
For a while, many of us treated the coder as perhaps the clearest example of modern intelligence. Looking around, it was not hard to see why. The software age produced enough millionaires and billionaires to make the case look fairly convincing.
Then AI learned to code.
And the old question returned: if the machine can now do the work that looked most intelligent, what is intelligence?
The answer is not that coding has vanished. Coding has moved.
The syntax still exists, but natural language can increasingly produce it. We are still telling the machine what we want. We are simply doing more of that in our language instead of first translating ourselves into its.
The strongest coder therefore moves upward—from producing every line towards defining the problem, designing the system, deciding how the parts should connect, testing what comes back and directing several coding agents towards one result.
The work has evolved from writing more of the syntax to conducting more of the whole.
Which brings us back to intelligence.
We may not be able to trap it inside one final definition. But across all these changes, we can see one thing clearly: intelligence tends to operate in two modes.
The Worker and the Conductor
There is the mode in which we perform the task. We cut the vegetables, write the paragraph, fill in the plan, produce the code or carry out the calculation.
Let us call this the worker mode.
Then there is the mode in which we hold the whole. We decide what meal we are making, what the book is trying to say, where the plan is meant to take us, what the software needs to become or which mathematical relationship needs to be expressed in the first place.
Let us call this the conductor mode.
These are not two permanent classes of people. Both modes can exist inside the same person, and most worthwhile work requires us to move between them.
When we cook a meal, the conductor chooses what we are making, which ingredients belong together and when each part needs to be ready. The worker chops, mixes, heats and serves.
When we write a book, the conductor holds the argument, the structure and the experience we want the reader to have. The worker researches, drafts, edits and wrestles with a sentence that sometimes comes out beautifully and sometimes, after forty minutes, comes out worse.
When we build a plan, the conductor decides where we are going and how the pieces connect. The worker fills in the dates, numbers, tasks and details.
Even mathematics moves between the two. One kind of intelligence sees the pattern and forms the equation. Another works through the equation, performs the calculation and checks whether the result holds.
Coding follows the same pattern. The worker produces and tests the components. The conductor understands the problem, holds the architecture and brings those components into a working system.
We need both modes.
The conductor without the worker may have a beautiful idea that never becomes real. The worker without the conductor may perform every task perfectly and still build the wrong thing.
And AI did not create this distinction. We have been describing it throughout the essay.
The Specialist worked within the part. The executive was supposed to hold the whole. The coder produced the syntax; the software architect held the system. One layer carried out the work. Another organised it.
AI simply makes the distinction much easier to see because it can increasingly supply parts of the worker mode. It can research, draft, calculate, compare, organise, code, design and revise. It does not do these things perfectly, and it still needs knowledge, checking and responsibility around it. But it does enough to change where the shortage lies.
If more of the worker can be supplied, human value begins to move towards the person who can decide what should be done, see how the parts fit, judge what comes back and guide the work towards something that matters.
It begins to move towards the conductor.
But that does not mean every person with a senior title was conducting in the first place.
When AI Moves Up the Building
At first, the impact of AI seemed to fall mainly on workers.
Could it draft the report? Produce the image? Analyse the figures? Write the code?
Then the impact moved towards specialists because AI could reach across domains, explain professional language and perform parts of work that had once required scarce training.
But the movement does not stop there. It eventually reaches the executive floor.
Remember what surrounded the senior executive: researchers, analysts, advisers, assistants, accountants, lawyers, writers, designers, engineers and operational staff. Their work flowed upward until the person at the centre appeared to know everything.
Sometimes that appearance was deserved. A genuine leader could see what others could not yet see, recognise the real problem, bring the right people together, choose between competing paths and take responsibility for the whole. They were genuinely conducting the intelligence around them.
But sometimes the machinery was doing more of the work than the person.
A staff member found the problem. An analyst developed the answer. A junior employee shaped the idea. It travelled upward because someone higher in the organisation had the authority, budget and team required to carry it into the world. By the time the idea reached the public, it had somehow been born in the corner office.
Power has a strange way of improving authorship.
Until now, it was often difficult to tell the real conductor from the person who simply stood where all the work arrived. Both had the title. Both had the office. Both appeared with the finished result.
The machinery hid the distinction.
AI begins to expose it because versions of that machinery can now be placed around far more people. Research, analysis, writing, design, coding and planning can increasingly be called upon without first possessing an executive budget or a large organisation.
The worker has not disappeared.
In a sense, the worker has been cloned.
And once the worker is cloned, everyone is affected.
The worker is tested because the task can be supplied. The specialist is tested because access to the domain is widening. The executive is tested because a title and a hidden army can no longer prove an ability to conduct.
Genuine conductors will use AI to extend their reach. Those who depended mainly on access, machinery and position will find that those advantages no longer hide them as well as they once did.
This is why AI may be more than a great disruption.
It may also be a great equaliser.
The Great Equaliser
Equalisation does not mean equality.
A talented young person in a poor neighbourhood or a small town in sub-Saharan Africa does not suddenly gain the capital, education, networks and security of a corporate executive. The old gaps do not vanish.
But one important gap can shrink.
With a phone, an internet connection and access to AI, that person can begin to reach forms of research, writing, analysis, design and software development that once required money, institutional access or a team of trained people.
The distance between the person with the idea and the person with the machinery becomes smaller.
Not zero.
Smaller.
For most of history, intelligence could remain hidden because the environment gave it no instrument. A person might see something clearly but lack the education, language, money, team or social position needed to express it in a form the world could recognise.
We then made the mistake of assuming that the person with the visible result possessed all the intelligence, while the person who could not produce the result possessed none.
AI disturbs that arrangement.
The orchestra that stood behind the powerful is beginning to appear in front of the individual. And once the instruments spread, we should expect intelligence to emerge from places the old system overlooked.
That is the evolutionary pattern again. Change the environment, and abilities that were always present finally receive a field in which they can act.
The next important company, idea, theory, book or invention may come from someone who would never have entered the old institution. Not because intelligence suddenly appeared in that person, but because the machinery needed to express it finally reached them.
AI does not flatten human difference. It reveals more of it.
It gives more people the chance to show what they can see, what they can build and whether they can conduct.
At the same time, it turns the light back on those who always had the orchestra and asks:
What did you actually bring?
The Conductor Sees the Music Before It Exists
Giving everyone an orchestra does not make everyone a conductor.
It may simply create more noise.
We can already see this. AI can produce an enormous amount of polished writing, imagery, video, software and analysis. Much of it looks impressive for a moment, until it doesn’t. Behind the polish we see the hollow idea.
So in a sense AI has not removed bad ideas. It has given them excellent grammar and very attractive lighting.
So if production becomes easier, production alone cannot be the new measure of intelligence.
The value moves towards vision: the ability to see something that does not exist yet. It moves towards understanding: knowing what the whole is trying to become and what each part must contribute. And it moves towards anticipation and judgment: sensing how one decision will affect everything that follows, hearing when the work is out of tune and knowing when the plan must change.
This is why the conductor’s real instrument is not simply the baton.
The baton is the visible part. It can be handed to almost anyone. A title can do much the same thing.
The deeper instruments are vision and hearing.
The conductor must see the symphony before it exists. They must understand the melody, the movement and the relationship between the parts. Then they must listen as the work becomes real and notice where it is failing to become what they saw.
Vision without listening becomes fantasy.
Listening without vision becomes correction without direction.
The conductor also needs knowledge. They do not have to be the best player of every instrument, but they must understand enough to know what good sounds like. Otherwise polished nonsense can pass for excellent work—and AI can produce polished nonsense at remarkable speed.
This is also why conducting is not the same as prompting.
A prompt is an instruction. Conducting is the larger intelligence that decides why the instruction is being given, what role it plays in the whole, whether the answer is any good and what should happen next.
Most importantly, the conductor must connect the whole thing to reality—to a genuine need, problem, desire, market or human concern. Otherwise the orchestra may perform beautifully for an empty room.
AI can help us explore possibilities, test ideas and carry out the work. But it cannot remove the need for us to decide what matters, what should exist and what value we are trying to create.
That is intelligence operating at the level of the whole.
What Intelligence Means in the Age of AI
We began with two people in two very different worlds because intelligence always needs an environment in which it can act.
History shows us the same thing at a larger scale. When memory was scarce, the Storyteller came forward. When knowledge could be stored, the Scholar rose. When the world became too complex for one mind, the Specialist took the stage. When the parts needed organising, the senior manager rose above them. When machines required syntax, the coder became the star.
Each age rewarded the intelligence it needed most.
Now AI is changing the need again.
Memory, calculation, research, drafting, coding and specialist knowledge are becoming easier to call upon. They have not become worthless. They are simply no longer scarce in quite the same way.
And when scarcity moves, value moves with it.
Coding moves from syntax towards design. Work moves from performing every part towards directing more of the whole. Leadership moves from holding the title towards proving the ability.
Human value begins to shift from merely possessing intelligence to organising intelligence.
We will still need the worker. We will still need to learn, practise, calculate, write, code, build and sometimes do the difficult work ourselves. But AI changes the balance. It gives more of us access to workers and asks more of us to become conductors.
That may be the direction intelligence is taking now: towards those who can see what does not yet exist, understand how the parts fit, call upon the right forms of intelligence, listen to what comes back and guide the work towards something that matters.
AI may be remembered as one of the greatest disruptions in history. But disruption describes what the old world feels when its advantages begin to break apart.
Equalisation describes what becomes possible when organised intelligence begins to spread.
The worker has been cloned.
The machinery is no longer hidden.
The orchestra is within reach.
Intelligence is changing again.
The question is:
Do we know how to conduct?
