In the Age of AI, Fashion Should Make What Humans Truly Need
What will we choose to keep human in a world where artificial intelligence can do more of what we once considered uniquely ours?
That question stayed with me after listening to Bill Gates discuss the next twenty years of AI in an October 2026 interview with The Verge's Decoder. He suggested that some activities—including sports, art, cooking and care—could be designated as human-reserved, even if machines eventually became more capable at performing them.
The point is not that AI cannot participate. It is that human participation may itself be what we value.
It made me think about the company we are building at Perfitt, and about a question that goes beyond size recommendations: What should fashion become when AI can help us design, predict and produce almost anything?
When making more becomes easy, making the right thing matters more
AI promises extraordinary gains in the speed of design, planning, merchandising and production. Brands may be able to explore more concepts, predict demand more frequently, and coordinate increasingly automated operations.
But this creates a choice. We can use those capabilities to flood the market with even more products. Or we can use them to understand people more precisely—and make what is truly appropriate for them, in the quantities actually needed.
These futures are not the same.
Fashion has long struggled with uncertainty. Brands must decide which designs to develop, which shapes and sizes to offer, how many units to manufacture, and where to send them before they can know precisely who will buy and wear them.
When those decisions miss, the costs appear as an uncomfortable fit, an avoidable return, the wrong size sitting in the wrong market, or a product that never finds its owner.
According to the European Environment Agency, an estimated 4–9% of textile products placed on the European market are destroyed before use. Its review also estimates online footwear return rates in Europe at roughly 30%. These are estimates for a particular market, not universal rates, but they illustrate how much friction remains between production and the people products are meant to serve.
From a pattern on the cutting table to an item someone will actually wear, every production decision matters. Photo: byhome work, via Unsplash.
The next definition of quality will be human-specific
The sports example offers another clue about the future. If people continue to place special value on what the human body can do—running, competing, moving, exploring—then the equipment and clothing that support those experiences may matter even more.
Not simply as fashion statements, but as products that genuinely work with an individual's body and movement.
Two people can wear the same labeled shoe size and have very different experiences. Foot length alone does not fully describe width, shape, volume or preference. Nor does the same nominal size ensure that two models will feel the same. In apparel, proportions and motion introduce another layer of complexity.
In that world, high quality may increasingly mean high relevance to the individual: a functional product with the right structure, the right fit, and the right performance characteristics for the person using it.
AI should help deliver that precision. Not by asking everyone to become more alike, but by helping brands recognize that people are different.
Human movement remains the point. Photo: Jorge Alberto Vega Barrera, via Unsplash.
From fitting a shoe to understanding demand
At Perfitt, we started with a concrete problem: helping people understand which footwear will fit them.
Our technology connects information about a person's feet with model-specific information about the shoes they are considering. That may sound like a small part of fashion commerce. In reality, it sits at the meeting point of two things the industry has often treated separately: the body that will use a product, and the product being designed, produced and sold.
Today, we work on fit guidance and measurement experiences across footwear, retail and apparel applications. Our longer-term vision goes further.
Imagine if anonymized and appropriately governed fit information could inform not just a recommendation at checkout, but an earlier design choice. If brands could understand differences in foot and body distributions across markets when planning their size ranges. If product-fit signals could complement sales and stock data when deciding what to manufacture and where to allocate it.
This is not a claim that all of these decisions are already fully automated by Perfitt. It is the direction we believe this data can help unlock—and one that requires careful validation with brands, manufacturers and retailers.
Our 3-Zero Vision: less mismatch, less loss, less unnecessary production
We call this long-term ambition our 3-Zero Vision.
Zero fit-related returns: move toward a world where a mismatch between a person and a product is the exception, not an expected cost.
Zero size-specific inventory loss: better connect products already made with the people and markets that actually need them.
Zero overproduction: use better understanding of real demand to guide designs, size mixes and production quantities before waste is created.
These are ambitions, not claims that fashion can literally eliminate every return, every unsold item or every forecasting error. Their purpose is to change the question driving the industry.
Instead of asking how much more we can make and sell, we should ask how precisely we can make what people need.
AI alone will not guarantee that outcome. Cheaper creation and production could just as easily increase excess. Better data, business incentives and intentional decisions must work together.
A fashion industry made by humans, for humans
Twenty years from now, I believe AI could make possible a very different model of fashion: one in which people create fashion for people, while intelligent systems help determine the appropriate designs, fits and quantities—with far less waste.
At the same time, I expect the human value of sport, exercise and movement to become more visible, not less. As those experiences gain importance, functionality and the fit unique to each person may become defining signals of product quality. We will increasingly look to AI to help us achieve that level of individual precision.
That is the future we are working toward at Perfitt.
Not a world in which technology replaces the reasons people love fashion, but one in which technology allows us to serve those reasons much better.
Fashion will remain human. What changes is how intelligently—and how precisely—we make it.
Perfitt intends to be at the center of building that future.
Steena Lee
CEO, Perfitt