Living With the Machines
Brands Become an Inference Engine
Today, in what I call the Age of Entanglement, brands are more than just identities we create. They have become inference engines that both people and machines learn to complete.
Living With the Machines: Brands Become an Inference Engine
The recent Chanel v Kamad Reworked case looks like a typical luxury trademark dispute. Kamad Reworked, a smaller company, sold jewellery made from Chanel parts like buttons and buckles. Chanel objected, and the Paris court sided with Chanel.
That could be the end of the story—a big brand simply protecting its trademarks.
But I believe there’s something more interesting happening, which becomes clear as we look closer.
This case shows that a brand is no longer just a product, a logo, or even a carefully managed story. Instead, a brand is becoming more like an inference engine—a network of signals that leads people and smart machines to find meaning in its parts.
This matters because the impact goes far beyond just fashion legal cases.
They reach into AI, identity, interfaces, consumer choice, and the entanglement we’re living through.
First, the explainer
Before we continue, here’s a quick explanation.
Kamad Reworked had been selling jewellery made from bits bearing Chanel signs, including their famous interlocking CC monogram and Chanel word marks. The argument was, in summary, that these were authentic components from genuine Chanel items, reworked into something new.
But the court was not convinced.
It said that even if genuine Chanel elements had been used, incorporating them into newly created jewellery resulted in a different product. Not one that had been placed on the market by Chanel in that form or with Chanel’s consent.
In other words, the usual rule that allows you to resell an authentic branded item did not apply to turning branded parts into something new.
The court found that the Chanel marks remained dominant on the finished pieces and would still be understood by the relevant public as signs of commercial origin. That mattered because the jewellery could still cause consumers to attribute the pieces to Chanel, regardless of who had physically assembled them.
Disclaimers did not help Kamad’s case.
References to Kamad’s own authorship, and even certificates presented as evidence of authenticity, were not enough to dispel the association with Chanel. In some respects, those disclaimers only made sense because the association with Chanel was already there.
So, on one level, this case is about trademarks, upcycling, and how far reuse can go.
But it’s the next level that interests me.
What the court was really protecting.
The court was not just protecting a physical object. It was protecting something more—the leap.
It was protecting that leap of understanding.
This leap happens when someone sees a familiar sign and instantly connects the dots:
CC, therefore Chanel, therefore luxury, therefore craft, therefore status, therefore trust.
That leap is not the product itself.
It is an inference.
And that is what makes this so revealing for the Age of Entanglement.
Inference is also how machine systems operate.
Recommendation engines, search systems, AI models, fraud detection, and shopping agents do not wait for the full picture. They use partial signals to guess the most likely next meaning.
More and more, brands have to work the same way.
A brand is not just there to represent itself, but to trigger a familiar response in whoever or whatever encounters it.
From brand identity to brand inference
For a long time, most brand thinking was built around expression.
How does the brand look?
What does it say?
What does it stand for?
How consistently can it present itself across touchpoints?
That approach still matters, but it’s no longer enough.
I argue it’s because a brand does not just exist in its designed surfaces anymore.
A brand now lives in the signals that last, spread, and get mixed after the product leaves the company. These include search results, metadata, marketplace listings, authentication records, resale histories, cultural references, interface rankings, AI summaries, logo fragments, product categories, and reputation layers.
The product is still part of the brand.
It is just not the whole story now.
In the old model, the brand was something the company expressed.
In the newer model, the brand becomes something consumers and systems can retrieve, infer, and reconstruct.
That is why I believe ‘inference engine’ is a helpful way to describe brand identity today.
A strong brand is no longer simply a sign system.
It is a system that helps others know how to complete the meaning.
This matters because the same thing is happening to us, both as people and as consumers.
It matters because the same thing is happening to us, both as humans and as consumers.
One central feature of my thinking about the Age of Entanglement is that people and brands are both becoming readable networks.
We become legible through data traces, behaviours, biometrics, purchase histories, preferences, search terms, relationship graphs, and all the other fragments systems use to model us.
The brand becomes legible through marks, provenance, style codes, retail signals, media references, verification systems, and all the other fragments people and platforms use to reconstruct meaning.
Neither the person nor the brand is reducible to those fragments.
But both get acted upon through them.
This is why I talk about the ‘legible self’ and the ‘wild self.’ Systems mostly interact with our legible side—the part that can be read, scored, sorted, and predicted. But people are always more than just that.
Brands are going through a similar shift.
Let’s call this the Signal Brand—a networked brand that exists as a web of cues readable by both machines and people.
The Chanel case makes that visible in legal form.
The court was saying: these signals still carry Chanel strongly enough that other people will reconstruct Chanel from the parts, and that reconstruction matters.
Which brings us to a bigger shift in thinking.
This is where the idea gets even more interesting.
If a brand is an inference engine, it’s no longer best understood as a static object.
It is not just something that waits to be recognised. It actively helps whoever encounters it to complete its meaning.
That has consequences, and they matter in three ways.
First, it changes what a brand is.
A brand is now less like a fixed thing and more like a process—an ongoing pattern of associations that can be triggered by pieces, not just the whole.
Second, it changes what brand ownership means.
Now, the question is not just who owns the trademark, but who gets to shape the meaning people take from it.
Who gets to decide how far a network of meaning can travel, and who may recombine its parts to trigger the same completion elsewhere?
Third, it changes what consumers are doing when they choose.
They are not simply choosing between products.
A brand acts as a shortcut for making decisions. It simplifies things and helps people quickly decide if something is desirable, safe, expensive, authentic, or meant for them.
The brand functions as a shortcut for judgment. It compresses complexity. It helps people know quickly under uncertainty: this is desirable, this is safe, this is expensive, this is authentic, this is for someone like me. academic.oup
That can be useful.
It can also be powerful.
Because once brands become engines of probable meaning, they start shaping not just preference but perception.
Now, people and brands share a similar logic, which is important for how smart interfaces work.
One reason this case feels important to me is that it makes brands and humans look structurally similar.
Both are now understood through networks of signals.
Both can be reconstructed from fragments.
Both can trigger inferences that become actionable before the whole person or whole object is ever present.
That matters for Intelligent Interfaces, where these inference systems meet.
Today’s interfaces are more than just screens or pathways. They are active environments where systems read both a person’s legible self and a brand’s signal identity at the same time.
It’s matching them.
Ranking them.
Filtering them.
Structuring what becomes visible, plausible, or desirable.
This is why a recommendation engine is never simply recommending.
It is connecting the brand’s inference engine with the person’s inference engine.
The result shapes what a consumer sees, what feels right, and what is interpreted as the obvious next step.
The consumer side of this follows naturally.
From the consumer perspective, this means brand choice is becoming less direct and more inferential.
You may think you are choosing between products.
But more and more, you are choosing between options that systems have already interpreted, based on what they know about you and the brand.
Which brands show up?
Which ones look trustworthy?
Which ones get positioned as premium?
Which ones are likely to be recommended by an AI shopping agent?
Which ones get treated as authentic, safe, or socially endorsed?
This is not advertising any more.
It’s infrastructural.
If brands act as inference engines within machine-driven environments, then consumer freedom depends less on having choices and more on whether the environment is structured fairly from the start.
That is a very different challenge from twentieth-century brand building.
The brand side follows the same logic.
For brands, the challenge is equally significant.
It is no longer enough to ask, What do we want to say?
Now, the real question is: What do our signals make people and machines conclude?
That means thinking beyond campaign surfaces and into operational identity.
What exists in structured form?
What can an AI system reliably retrieve?
What attributes travel across resale markets, marketplaces, third-party listings, and interface layers?
What fragments of the brand remain legible when detached from the original product or message?
What stories get reconstructed from those fragments?
This is why machine-readable branding is important—not just for technical reasons like schema and search visibility.
It matters in a philosophical sense, too.
Because the more a brand becomes readable to systems, the more it becomes governable through inference.
And as brands become more controlled by inference, they also take on more responsibility for the meanings they help create.
A Way Forward.
So what do we do with this?
I think we need to recognise that brands, people, and systems are now connected in environments where inference is always happening.
And then design accordingly.
For brands, this means not just telling stories, but actively shaping how people see and understand reality.
For platforms and interface designers, it means recognising that recommending, ranking, and retrieving are not neutral actions. They are meaningful decisions that shape outcomes.
For people, it means learning to live with machines without reducing ourselves—or our world—to only the parts that systems can easily process.
If interfaces are now where people and brands are interpreted into action, we need better ways to make those processes clear.
We need systems where people can clearly see how a brand is shown, why a recommendation appears, what signals drive the match, and where they still have the chance to pause, change direction, or say no.
Because the real danger is not that machines will understand everything.
The real risk is that we might start treating these likely completions as if they are the whole truth.
The entangled future of brand identity.
The Chanel case is being reported as a dispute about luxury, upcycling, and trademark boundaries.
But I believe this case shows that what is at stake now is not just who owns objects, but who controls the meanings we draw from them.
It is not just about protecting products, but about protecting the networks of meaning around them.
That matters because the same transformation is happening everywhere.
To brands.
To people.
To culture.
To choice.
We are moving from a world of separate identities to one where identities are mixed, readable, and can be combined in new ways.
In this new world, a brand is not just what it claims to be. And as individuals, we are rarely just what systems say we are.
It is what humans and machines have learned to conclude from the parts.
That’s what an inference engine is.
Learning to live with this reality—by thinking critically, staying aware, and not giving up all control to the system—may be one of the main challenges of our entangled future.
NOTE: This argument relies on the Chanel/Kamad ruling, which involved Chanel-branded components remade into jewellery that the Paris court treated as a new product, and found disclaimers and authenticity-related materials insufficient to avoid likely consumer association with Chanel. thefashionlaw. ie-forum. seoteric.
I hope you enjoy the article. If you’d like to talk more about these ideas, please feel free to reach out.
I’m also opening up spots for conferences, in-house presentations, and in-depth one-on-one calls, starting in September. DM me.
Thanks for reading UNCX | The Age of Entanglement.
Subscribe for free - you get all new posts and support my work.