The tuxedo already had a life before Saint Laurent.

Its vocabulary was familiar: black wool, satin lapels, white shirt, bow tie. It carried restraint, ceremony, and a particular kind of male authority. So when Le Smoking appeared in 1966, Saint Laurent was not introducing an unknown object.

He was putting a known object somewhere it was not supposed to be.

On a woman.

In retrospect, that move looks almost obvious. This is one of the problems with successful cultural shifts: once they settle into everyday life, it becomes difficult to reconstruct why they once felt strange. We see a woman in a trouser suit and barely register the code. But in the 1960s the tuxedo still belonged to a very specific social and gender vocabulary.

Pierre Bergé later compressed this shift into a sentence:

I’m wary of grand statements like this. Fashion did not single-handedly give women power, and Saint Laurent did not cause the social transformations already happening around him. But the sentence points to something more precise that I do want to keep.

And even that description is slightly too simple, because he did not merely put a man’s suit on a female body. The proportions and construction changed. The recognizable elements remained, but they were translated around another silhouette. Enough of the original code survived for us to read “tuxedo”; enough changed for it to become something else.

This was the first moment in the material where I stopped thinking about “breaking rules” and started thinking about reassignment.

We already do this constantly, although we rarely describe it this way. A critique format that belongs to art school becomes a method for developing a business idea. A sprint created for software teams moves into strategic work. A diagnostic structure from medicine becomes a way to explore an organisation. Game mechanics migrate into learning. A laboratory protocol becomes a way of testing a product hypothesis.

The object, format or logic may already exist. The novelty comes from the new relationship around it.

And Le Smoking goes one step further, because clothes do not just cover bodies. They change what bodies can comfortably do.

A long evening dress and a trouser suit do not ask the same thing of the person wearing them. You walk differently. Sit differently. You can put your hands in your pockets. You occupy space differently. The change begins physically, but it does not remain physical for very long.

That made me think about permission.

Most of the language around products is built around capability. A course lets me learn something. A tool helps me perform a task. A service helps me solve a problem.

But knowing that something is possible and feeling that it is available to me are not quite the same thing.

I can know how. I can be capable. I can even have enough confidence. And still feel that I am somehow not supposed to do it.

This is especially visible with expertise. People often know far more than they allow themselves to turn into a method, a product, or a point of view. The missing piece is not necessarily another piece of information. Sometimes what changes everything is a structure that legitimizes the action: a framework that lets you make the decision, a format that makes experimentation acceptable, or a game that allows you to approach something that felt too serious or too risky head-on.

A product can make something usable not because it creates the capability, but because it creates permission around a capability that was already there.

That is a rather different kind of value from many others.

There is another detail in the story of Le Smoking that I keep coming back to. Saint Laurent’s haute couture clients were not particularly enthusiastic at first. The younger Rive Gauche clientele was much more receptive.

The same object appeared to two audiences at roughly the same historical moment, and one saw something awkward while the other saw something that made sense. That seems more useful to me than being brave enough to break the rules.

Sometimes the old audience sees a category error because the new category does not belong to its world.

The emerging audience simply sees itself.

The Mondrian dresses are probably the easiest Saint Laurent pieces to recognize, and that familiarity creates another retrospective trap. We see Mondrian, then we see the dress, and the connection feels almost immediate.

Painting becomes fashion.

I-n-s-p-i-r-a-t-i-o-n.

But inspiration is one of those words that explains almost nothing once you look closely.

Saint Laurent was already working on his 1965 autumn-winter collection when a book on Mondrian changed its direction. He could easily have taken the obvious route: reproduce the colours, borrow the grid, print something recognisably Mondrian onto fabric.

What happened technically was more demanding. The areas of colour were made from separate pieces of wool jersey assembled into the garment. The black lines were not simply graphic decoration laid over the surface; they became part of the construction.

And here the difference becomes much more appealing to me, because Mondrian’s painting and Saint Laurent’s dress do not live under the same conditions.

A painting is flat. A body has volume. A painting stays where it is placed. A dress bends, turns, and walks. Mondrian controls the exact relationship between every rectangle and line; a garment has to survive shoulders, breasts, hips, gravity, and motion.

So Saint Laurent’s real problem was not "How can I reproduce Mondrian?"

We borrow across disciplines constantly now. Education borrows from games. Marketing borrows from cinema. Business borrows from biology. Digital products borrow from conversation. Workshops borrow from theatre, laboratories, television shows, board games and rituals.

And very often I can see the source immediately. That is not always a compliment.

A course becomes “Netflix-like” because it has episodes. A workshop becomes a game because there are cards on the table. A presentation borrows from cinema because someone has divided it into three acts. A visual identity becomes “Bauhaus-inspired” because circles and primary colours have appeared.

The reference has travelled. The thinking behind it may not have.

I have become increasingly suspicious of inspiration for exactly this reason. It is very easy to collect references and much harder to say what you are actually extracting from them.

With Mondrian, it might be geometry, proportion, rhythm, asymmetry, reduction or the relationship between colour and structure.

With Picasso, the answer changes. Saint Laurent spoke about him very differently:

There is no single “Picasso system” waiting to be transferred in the same way. Saint Laurent was responding to multiplicity, abundance, force, the movement between periods and visual languages. The source was still painting, but the variable worth taking from it had changed.

This is where cross-disciplinary work either becomes genuinely productive or collapses into decoration.

Take jazz. Saying that a product is “inspired by jazz” sounds good and tells me almost nothing.
Do you mean improvisation? A strong structure that makes improvisation possible? Musicians listening and responding to one another in real time? Variation around a theme? The alternation between taking space and leaving it? Virtuosity? Syncopation?

Those are not different words for the same idea. They would produce completely different designs.

The same applies to games. If I say that I want to bring game design into learning, what exactly am I bringing? Competition? Agency? Feedback? Rules? Uncertainty? A safe space for failure? The pleasure of mastering a system? Or design methodology?

I don’t need all of them. I need to know which one helps with the problem I am trying to solve.

Saint Laurent had to perform the Mondrian translation again when he worked on costumes for Roland Petit’s Notre-Dame de Paris. What survived on a fashion silhouette had to work differently on a dancer. Movement changed. Distance from the audience changed. Materials changed.

His formulation was wonderfully uncomplicated:

I like emprunté here.

Borrowing is not copying. Something leaves one environment and has to function inside another. The new environment gets a vote.

If I take something from a laboratory into a workshop, perhaps the least valuable part is the aesthetic of white coats, grids and test tubes. Maybe what I actually want is the behaviour underneath: formulate a hypothesis, create conditions, observe what happens, revise.

At that point the finished workshop may look nothing like a laboratory.

Good. The reference has stopped being an image and started becoming a way to think.

This distinction is becoming more important as generative AI makes visual and stylistic borrowing almost frictionless. We can request a website with the feel of a luxury magazine, an illustration reminiscent of a particular artistic movement or a learning experience “like Netflix” and get recognisable results almost immediately.

Surface resemblance is becoming cheap.

So I find myself asking a more demanding question of references now: Why this one? What property am I after? What effect does it create? Can I separate that effect from the visual form that originally carried it?

One little exercise works surprisingly well. Remove the famous name. Not “I’m inspired by Picasso.” Try, "From Picasso, for this particular problem, I want to borrow… ."

If the sentence cannot be completed, I probably have a moodboard. If it can, I may have something I can actually work with. And the better the extraction, the less the result may resemble Picasso in the end.

In January 1971, Saint Laurent presented a collection that was attacked so fiercely that it acquired its own name: La collection du scandale.

Short dresses, platform shoes, square shoulders, heavy makeup, references to the 1930s and 40s and, uncomfortably for many critics, to the Paris of the Occupation.

The obvious contemporary reading is that Saint Laurent was ahead of the trend.

I initially read it that way too. Then I found Paloma Picasso. She had already been buying old clothes at flea markets and wearing them in combinations that caught Saint Laurent’s attention. Her appearance was described as a choc esthétique for him. The collection did not emerge from nowhere. He had seen a behaviour that was already happening and decided to push it much further.

That changes the mechanism. He did not predict the retro current out of thin air. He noticed something while it was still small enough to be dismissed. Someone was already doing it.

This distinction feels important because we like the mythology of the visionary. It is much more dramatic to say that someone saw the future than to say that they looked more carefully at the present. But looking carefully may be the rarer skill.

One woman buying old clothes at flea markets is easy to classify as personal eccentricity. A couture collection organised around the same logic turns the behaviour into a cultural proposition. It enlarges the signal until other people have to respond to it.

They can hate it, which many did.
They can imitate it.
They can argue about what it means.
But it is visible now.

That is when I started seeing the 1971 collection less as a story about prediction and more as a story about amplification.

A lot of change does not begin in the centre of a system. It begins with people doing things badly, strangely or unofficially at the edges. They combine objects that were not designed to go together. They use tools for purposes nobody intended. They abandon parts of a process but keep the rest. They develop workarounds because the official route is inconvenient. They start using a new vocabulary before the old profession has decided whether it approves.

At first, these behaviours rarely look strategic. Mostly they look messy. And strategy has a natural preference for patterns large enough to defend with evidence. How many people? How fast is it growing? What is the market size? Can we segment it?

Those are reasonable questions. I would not want to run a business entirely on aesthetic shocks at flea markets. But by the time a behaviour is clean enough to appear in a trend report, somebody has usually been living inside it for quite a while.

This is where art, fashion and subcultures become valuable as research material. Not because artists have supernatural access to tomorrow, but because they are sometimes willing to take seriously a phenomenon before it becomes statistically respectable.

I would formulate it more cautiously now:

That changes how I look at weak signals in product work too.

People take information from your carefully designed platform and move it into WhatsApp. The first reaction may be that they are using the product incorrectly. But what if the workaround tells us that the conversation is actually part of the value, and our product architecture has treated it as peripheral?

People buy a course, skip large sections and spend their time discussing a few ideas with peers. We can try to increase completion. Or we can at least ask whether completion is measuring the thing we think it is measuring.

People working with expertise increasingly combine consulting, teaching, facilitation, media, digital tools and product design. Traditional positioning advice often treats this as a clarity problem: choose one thing.

Sometimes it is. Sometimes the work has moved and the category has not.

This is where I have to resist another seductive shortcut. Not every anomaly is a weak signal. Sometimes people use a product incorrectly because the interface is bad. Sometimes a strange behaviour disappears because it was simply strange. Controversy does not prove that an idea is ahead of its time.

But I would rather inspect the anomaly before correcting it.

• What are people doing that they are not supposed to do?
• What are they combining that our categories say should remain separate?
• What do experienced professionals dismiss as amateurish that newcomers do without hesitation?
• Where are people building workarounds rather than asking permission for a better system?

Paloma Picasso did not arrive with a deck entitled Emerging Retro Consumption Patterns 1971–1975. She arrived wearing clothes. Saint Laurent noticed.

Sometimes foresight is less about seeing further than everybody else and more about looking where everybody else has decided there is nothing to see.

Marguerite Duras described Saint Laurent’s women in a way that stopped me:

The image is wonderfully concrete. No persona. No segment. A woman moving.

She is in the street, on the metro, shopping, working, crossing from one social environment into another. Fashion now has to travel through that life with her.

The more I thought about Rive Gauche, the less I saw it simply as a story about accessibility.

That was my first reading: couture is expensive, Saint Laurent wants to reach more women, ready-to-wear provides the solution.

But “more women” hides the more consequential change. These women were not just couture customers with smaller budgets. Their lives were organised differently.

Saint Laurent himself put his frustration rather less elegantly:

At first the word milliardaires attracts all the attention. I think the more revealing word may be robes.

What if the new customer was not waiting for a cheaper version of the old thing?

Rive Gauche was developed as its own line rather than simply as reduced couture, and that makes the case much more relevant to intellectual products than the usual “premium versus accessible” story. I see this mistake constantly when expertise is productised.

There is a good two-hour consultation. It works. The expert wants something more scalable, so the logic seems obvious: record the knowledge, divide it into modules, remove the live interaction, lower the price. Now we have a course.

Except the person taking that course is not simply the consulting client with less money.
They do not have you in the room to tell them which part matters. They cannot interrupt when something is unclear. They may need a diagnostic before they need an explanation. They may need examples, feedback, a peer, an external deadline, or a decision tool at precisely the moment a decision appears.

If the conditions of use change, the design problem changes with them. This is why “make it scalable” can become such a strange instruction. We take an experience whose value depends partly on judgement, interaction or adaptation, remove those things because they are difficult to scale, and assume the remaining information is the same product in a more efficient form.

Sometimes it is. Sometimes we have scaled away the reason it worked.

A different audience is not necessarily a lower price point. It may be another context, and context can require another product. The same knowledge can become a book, a diagnostic, a game, a workshop, a conversation, a chatbot or a field guide. If knowledge were the whole value, these forms would be mostly interchangeable. They are obviously not. Every format quietly contains a theory of the person using it.

A book assumes that I can stay alone with the material for a while. A workshop assumes I can give you a particular part of my day. A chatbot makes almost the opposite assumption: perhaps I do not want to follow your sequence at all; perhaps I want to ask the next question that matters to me.

Once I started looking at formats this way, the idea of “the same content in another format” became much harder to believe. The format is already making decisions about my life.

This is why Duras’s sentence feels so contemporary to me. She does not tell us what the customer likes. She tells us where the customer has gone. A much better design question.

Where is the person when the problem appears? What happened just before? What will happen five minutes later? Who else is involved? What tool is already open? How much attention is realistically available?

More awkwardly, what are people doing today that the category still treats as abnormal?

The learner has already left the classroom in many contexts. Learning now happens while working, searching, experimenting, talking to peers, watching fragments of content, asking AI, getting stuck and returning for another piece. This does not make classrooms obsolete. It just means the classroom no longer owns the definition of learning.

The buyer has done something similar to the funnel. We can map awareness, consideration and purchase as neatly as we like. Meanwhile, an actual person can read an article, disappear for eight months, discover somebody else, ask a friend, use a free tool, reject the entire category, see you again through an unrelated post and suddenly buy the expensive offer.

The model is not useless. The person has simply refused to live inside it.

And I think something similar is happening to expertise itself. The person who writes, teaches, consults, facilitates, creates tools and works across several disciplines can look incoherent inside a category built around one profession and one service.

The standard response is "simplify". Often that is excellent advice. But occasionally what looks like an unclear professional identity is an early form of a professional category that does not yet have a comfortable name.

Saint Laurent surrounded himself with women who were difficult to reduce to a static idea of femininity: Betty Catroux, Loulou de La Falaise, Paloma Picasso, Catherine Deneuve. Duras wrote another line about him that I like even more:

That could almost be a definition of a certain kind of design. Not asking the user to describe the future perfectly. Watching closely enough to notice what they are already becoming.

People often change behaviour before they can explain the change. They invent an awkward workaround. They combine tools that do not belong together. They start calling themselves something slightly ridiculous because the existing title keeps cutting away important parts of what they do.

Those behaviours may disappear. Or they may be telling us that the salon is empty and we are still designing the furniture. There is no virtue in chasing every new habit. Some things are worth preserving even when they are inconvenient. Old formats are not bad because they are old.

The question is simpler: does the environment that made this product sensible still exist?If the answer is yes, keep designing for it. If the user is already halfway across Paris, changing the curtains may not help.

Saint Laurent returned to the same garments for decades: the smoking, the peacoat, the safari jacket, the trench coat, the trouser suit, the jumpsuit.

This initially looks odd if we equate creativity with constant novelty. Once you have done the tuxedo, why do the tuxedo again? But perhaps he was not repeating an object. He was continuing a sentence.

The garments changed from collection to collection, but certain operations kept returning: masculine codes moved onto female bodies, utility became elegant, familiar forms were shifted just enough to acquire another meaning. The objects varied; the way of working with them was recognisable.

That is much closer to grammar than to visual consistency.

Saint Laurent’s famous line is:

It is easy to turn that into a quotation about timeless elegance and move on. I think there is a more practical reading.

A language can produce an enormous number of sentences without changing its grammar every morning. Its identity does not come from repeating the same sentence, but from the relationships that remain possible between elements.

A creative signature can work like that too. We usually start with its most visible parts: colours, typography, tone, illustration style, recurring formats. Those things help enormously with recognition, but they are also the easiest things to imitate.

A colour can be copied. A layout can be reproduced. Tone of voice, once one of the more elusive parts of authorship, can now be approximated by AI in seconds. The deeper signature is more inconvenient because it sits in decisions.

This is one of the reasons I am more interested in somebody’s repeated moves than in their repeated visual forms.

Imagine an expert who writes essays, runs workshops, creates games and builds strategic tools. The outputs may have very little in common visually. But perhaps all of them begin in the same place: take something familiar, expose the assumption underneath it, bring in material from another discipline and turn the result into something people can manipulate. That is already a signature.

This also explains why some creators remain recognisable when they change medium. A filmmaker makes photographs and something of the same mind remains visible. A novelist writes an essay and the same way of looking keeps appearing. The medium changes more than the author does.

For intellectual products, this gives me a test I find more practical than “Is the branding consistent?”

Remove the logo.
Remove the colours.
Remove the name of the framework.
Remove the favourite format.
What is still recognisably yours?

Maybe you systematically turn abstractions into objects people can play with. Maybe you refuse to separate business problems from human behaviour. Maybe you work through contradictions rather than forcing a clean choice too quickly. Maybe you repeatedly pull material from art, games or science into places where nobody expects it.

Those are harder to show on a brand board. They are also much harder to copy.

And there is another consequence. A visual identity mainly helps an object remain recognisable once it exists. A grammar can help you create an object that does not exist yet.

If I know how I tend to think, what kinds of relationships I look for, what I refuse, what I repeatedly question and what kinds of transformations I value, I have something that can travel from an article into a workshop, from a workshop into a digital tool, from a tool into a completely new product form.

The visual expression can change without the work losing its centre.

Pierre Bergé said of Saint Laurent:

An œuvre is not a pile of things that look alike. It is what starts becoming visible when enough different things have emerged from a persistent way of looking.

I find this reassuring because we are often encouraged to define a signature before we have enough work to know what ours is. Choose three adjectives. Pick the colours. Define the tone. Create a recognisable format and repeat it. Some of that is useful. But the deeper signature may need to be discovered backwards.

You make things. A question keeps returning. You notice that certain combinations attract you again and again. Different problems produce strangely similar intellectual moves. Something survives changes of topic, client and medium. At some point you can see the pattern. Then you can strengthen it deliberately.

So I would replace the question What should I repeat so people recognise me? with another one:

The answer will probably be less visible. It may also be where the real authorship lives.

There is a romantic version of the title of this series. I know the first publication is a kind of a trial; I experiment with a format.

Art sees first because artists are unusually sensitive creatures who somehow feel what is coming before everybody else. I like the romance of it. I just don’t think we need it.

Saint Laurent did not have to predict that women’s lives were changing. They were already changing. He did not invent Paloma Picasso’s flea-market experiments. He saw them. He did not create the tension between masculine and feminine codes. He found a way to make that tension visible. His way. That is enough.

A forecast tries to tell us what will happen later. Art can take something that is already happening now but is still scattered and difficult to read and increase its resolution.

One unusual single behaviour is easy to ignore. A collection, film, novel, photograph or song can concentrate the same tension until it becomes something we can look at. Once it has a form, we can disagree with it. We can reject it, desire it, laugh at it, imitate it, become offended by it. But before any of those reactions, somebody has made it visible.

Business operates under a different constraint. If I am investing money, hiring people or manufacturing thousands of goods, I would quite like evidence that more than three eccentric people want the thing. So business asks sensible questions. How large is the market? How fast is it growing? Is the behaviour repeatable? Can we measure demand?

Art is allowed to have a lower threshold of suspicion. A filmmaker can become preoccupied with a behaviour she has only seen a handful of times. A novelist can spend years inside a contradiction for which society has not produced a label. A photographer can point the camera at people nobody considers a segment. Most of these observations will not become “the future.”

That is fine. Looking early necessarily produces false positives.

What interests me is the timing difference. By the time something is established enough to become a trend, somebody has already been doing it. By the time a new professional category has a name, people have usually been working inside it for years. By the time a new learning behaviour appears in research, learners have already been inventing strange workarounds around the old system. The report does valuable work. It validates and organises what is happening. But it usually arrives after the first signs.

Art often leaves us material from the awkward period before the pattern becomes easy to defend. That is why I would not replace research with art. I would use them at different moments.

Data becomes powerful once we know what we are trying to measure. Art can be useful while we are still working out what might deserve our attention. And this feels particularly relevant now.

We have become extraordinarily good at generation. I can generate twenty product directions, fifty names, several visual worlds and a pile of frameworks before lunch.The number of possible ideas is no longer the problem it once was. So the question moves upstream.

What did you notice?

What behaviour is still too small to appear on a strategy slide? What does your existing framework keep throwing away as noise? Where are people improvising because the official solution does not fit? What contradiction are they simply living with while we keep trying to “solve” it?

Professional systems have a habit of treating anomalies as mistakes.
A customer uses the product incorrectly, so we improve onboarding.
Learners skip the content, so we work on completion rates.
People combine two offers we intended to keep separate, so we clarify the architecture.
A professional does not fit a category, so we tell them to specialize.

Sometimes the system is right. But sometimes the anomaly contains better information than the system.

Perhaps the customer has found a better use. Perhaps completion is not the right measure of learning. Perhaps two services belong together because the actual problem refuses to respect our categories. Perhaps the professional identity looks incoherent because the job description was built for work that is disappearing.

I don’t think the answer is to celebrate every anomaly. That simply replaces one dogma with another. I want to leave it unresolved for slightly longer.

Art is good at that. A novel does not have to turn a contradiction into recommendations at the end of the chapter. A film can show behaviour without giving us five actions to take on Monday morning. A fashion collection can exaggerate a code simply to see what happens when people encounter it.

Professional thinking is often less patient. We name. Categorise. Build the quadrant. Extract the framework. At some point the framework becomes so convenient that we begin seeing reality through it instead of using it to see reality more clearly.

Which brings me back to the material that started this whole article. I could have kept the ten business lessons from Yves Saint Laurent.

• Break the rules.
• Know your audience.
• Find your signature.
• Take inspiration from other disciplines.

They are not false. They are simply the least specific things we can say about what happened.

• “Break the rules” removes the particular rule.
• “Know your audience” removes the woman running through the metro and the Bourse.
• “Find your signature” removes decades of repetition, revision, and accumulated choices.
• “Take inspiration from art” removes the problem of getting Mondrian off a flat canvas and onto a body that moves.

The more portable we make the lesson, the more of the original case we throw away. And perhaps that is exactly the material worth keeping.

I want to stay with the particular case for a little longer and ask what operation is happening there. What existed before? What actually moved? What remained intact? What changed meaning?

Then I can try the move somewhere else. A fashion collection from 1966 will not tell me what product to launch next year. But it can make me notice that ownership is a design variable. Mondrian will not design a learning experience for me. But the journey from canvas to garment can show me what real translation between disciplines asks of us.

Paloma Picasso cannot tell me which weak signal deserves attention in my field. But she can remind me where to look: among behaviours that are still too small, too awkward or too incorrect to have a respectable category.

That is what I mean by Art Sees First. Not that art knows first. Knowing comes later.

Someone notices something. Someone gives it a form. Other people react to it. The signal grows, mutates or disappears. Researchers measure it. Business organizes around it. Eventually somebody finds a name. And by then it often feels as though it had always been obvious.

Art leaves us traces from before it was obvious.

I think we should learn to read them.

There is a bit of irony in finishing an article that began by rejecting “ten lessons from Yves Saint Laurent” with a method extracted from Yves Saint Laurent.

So I’m not going to give you six steps. I would keep one habit.

Look before you invent.

The next time something feels as if it needs a new solution, I would first look at what is already there and ask whether the object is really the problem.

• Could the ownership change?
• Could the context change?
• Could I take an operating principle from somewhere else without taking its appearance?
• Is somebody at the edge already behaving differently?
• Is an apparently confused combination actually telling me that two categories should meet?

Saint Laurent repeatedly worked with materials that were already available: a tuxedo, a painting, a military garment, an old dress found at a flea market, a woman whose life had already moved beyond the world fashion had been designed for. His originality was not always in the ingredients. It was in the relationship he saw between them.

That feels increasingly relevant in a world where generating another object, image, framework or product concept is becoming almost effortless. Our instinct is to respond by generating more.

I’m not sure that is where the advantage is going to be. When generation becomes abundant, attention becomes expensive and weaker.

So perhaps the question worth keeping from Saint Laurent is not What can I create that nobody has created before? It is a quieter one.