The Body's Operating Layer

The oldest wearable computer is still waiting to be invented.

Antwan Baker, Founder

Research status: Technology watchlist. The convergence described in this paper is real enough to monitor, but neither Nike nor lululemon has earned valuation credit for it. The probabilities below are Laray.ai judgments, not measured frequencies.

The interface hiding in plain sight

Most technology asks us to carry one more thing.

A phone goes in the pocket. A watch wraps around the wrist. A ring sits on a finger. A glucose monitor adheres to the skin. Each device earns a place on the body by proving that the information it collects is worth the inconvenience of wearing it.

Clothing begins with an advantage none of those devices had to win: it is already there.

A shirt touches the chest and back. A shoe sits beneath the full force of every step. Leggings follow the hip and knee through thousands of small movements. Clothing knows, at least physically, whether the body is hot, cold, wet, compressed, extended, accelerating, or exhausted. We simply have not built most garments to do anything with that information.

For almost all of human history, clothing has been passive. It protects, warms, covers, signals identity, and occasionally improves performance through shape or material. The garment does not understand the person inside it. It does not change its behavior because the wearer's gait is deteriorating, breathing is strained, temperature is rising, or muscle is approaching fatigue.

That boundary is beginning to move.

Sensors can be sewn into fabric. Algorithms can infer motion from strain. Small motors can add force to a joint. Engineered enzymes can take a discarded synthetic garment apart and return its polymers to their original building blocks. Fermentation can produce fibers that once had to come from animals or petroleum.

This creates a strange but increasingly plausible future: the companies we call apparel brands may become customers, investors, and integrators of robotics, artificial intelligence, materials science, and industrial biotechnology.

The tempting version of the thesis is that Nike and lululemon will become technology companies.

The more defensible version is harder and more interesting: clothing is becoming a deep-technology industry, but the clothing brands may not be the companies that capture most of the value.

When a shoe begins to move back

Nike's Project Amplify makes the idea tangible.

The prototype is not a faster foam or a cleverer carbon plate. It is a powered footwear system developed with Dephy, a robotics company whose work grew out of research in wearable movement systems. A motor, drive system, battery, and control software work with a running shoe to add assistance at the ankle.

The distinction matters. Traditional footwear can return some of the energy a runner puts into it. A powered system can introduce energy from somewhere else.

Nike says the system has been tested across more than 400 athletes and informed by roughly 2.4 million steps collected through the Nike Sport Research Lab. The target is not the elite marathoner already moving at the edge of human efficiency. Nike has described an everyday runner or walker—someone moving at roughly a ten- to twelve-minute mile—who might go farther with less effort.

The dream is easy to understand. An electric bicycle did not replace cycling. It changed who could ride, how far they could go, and what counted as a practical trip. Powered footwear could attempt the same shift for walking and running.

Imagine a shoe that recognizes the difference between climbing, descending, fatigue, and an unstable stride. It could learn when to apply torque, how much assistance a particular body needs, and when intervention would make movement worse rather than better. Over time, the product would become less like footwear with a motor and more like a physical operating system for the lower body.

That is the technology story. It is real enough to take seriously.

The product story is much less mature.

The $4,500 reality check

Dephy's own Sidekick system shows both how far the technology has come and how far it still has to travel.

As of September 2026, the Sidekick starter package is listed at $4,500. The powered units weigh about 3.1 pounds per side including the battery and provide up to roughly five miles of assistance, depending on use. The system adapts to the wearer's rhythm and is marketed as a wellness and mobility product rather than a medical device.

That is an impressive machine. It is not a mass-market shoe.

At this stage, the likely customer is not a twenty-five-year-old trying to shave seconds from a neighborhood run. It is someone for whom walking farther has unusually high value: an older adult, a person returning from injury, a traveler anxious about range, or a worker whose job makes every additional mile costly.

Those customers can support a beachhead. They do not yet prove the broad athletic market.

Powered footwear must cross several curves at once. Weight must fall. Battery life must rise. The machine has to survive rain, sweat, dust, curbs, stairs, travel, and repeated impact. It must fit bodies that move differently without injuring them. It needs a service model for batteries, motors, firmware, returns, and repairs. Most of all, it has to become cheap enough that assistance feels like a feature rather than a medical-equipment purchase.

The category can be technically successful and commercially small. Exoskeletons already make sense in military, industrial, rehabilitation, and mobility settings where reducing fatigue or restoring movement is worth thousands of dollars. Consumer apparel has a different standard. It has to work, look normal, feel natural, and survive fashion's brutal replacement cycle.

That is why Project Amplify should be understood as a serious experiment, not as proof that Nike already owns the next footwear platform.

The second revolution begins before the garment exists

The other branch of this story has no motors.

It begins inside the material itself.

Most synthetic performance clothing relies on polymers derived from fossil fuels. Nylon and polyester are durable precisely because they are difficult to return to their original chemical ingredients. Mechanical recycling can degrade quality, struggle with blends and dyes, or turn a high-performance textile into a lower-value material.

Samsara Eco is pursuing a different route. Its engineered enzymes are designed to break certain plastics and textiles back into usable molecular building blocks. Those inputs can then re-enter manufacturing rather than becoming waste or lower-grade material.

Lululemon has become an investor, design collaborator, and prospective customer. The companies produced a sample using enzymatically recycled nylon 6,6 and later announced a ten-year offtake plan. Lululemon says Samsara's materials could eventually support approximately 20 percent of its overall fiber portfolio, while Samsara plans an international commercial facility for 2028.

If the process reaches cost, quality, and volume targets, the consequences extend beyond a sustainability label.

A circular nylon supply could reduce exposure to virgin petrochemical inputs. It could help a brand respond to waste regulation and producer-responsibility rules. It could make material recovery part of product design: a garment would be engineered not only for how it performs during use, but also for how its molecules can be recovered afterward.

This is genuine biotechnology entering apparel. But it does not make lululemon a biotechnology company.

Samsara designs the enzymes and operates the process. Lululemon helps create demand, tests the material, influences specifications, and may benefit from its equity investment. That is strategically meaningful. It is still closer to deep-technology-enabled supply-chain management than to owning a biotech platform.

The distinction became harder to ignore when activewear brand LSKD announced its own ten-year Samsara Eco partnership. A supplier serving multiple brands may build a powerful business. Its success does not automatically create a proprietary advantage for any one customer.

The mistake hiding inside the original thesis

The first version of this thesis combined powered footwear, textile sensing, adaptive garments, fermentation-derived fibers, and enzymatic recycling under one exciting label: apparel companies will become biotech companies.

That framing captured the direction and blurred the economics.

Body-interface systems and bioindustrial materials are not one market. A powered shoe depends on robotics, batteries, biomechanics, control software, consumer hardware, and possibly medical-device rules. Enzymatic nylon recycling depends on molecular biology, industrial chemistry, feedstock quality, plant utilization, offtake contracts, and commodity pricing.

The science is different. The capital is different. The regulatory risk is different. The customer is different. The path to margin is different.

They belong in the same story only because both are changing what an apparel company must know how to buy, integrate, and govern.

This leads to the central investment question. It is not whether the technology works. It is not even whether consumers eventually want it.

It is: who owns the scarce layer?

If Dephy owns the actuation and control system, Nike may be a valuable distribution partner without owning the platform. If Samsara owns the enzymes and sells the same recovered nylon to every major brand, lululemon may improve its supply chain without creating a moat. Society can receive better products and lower waste while the public apparel shareholder captures little of the new surplus.

Technology feasibility, consumer adoption, and equity capture are three separate events. Investment stories become dangerous when they are treated as one.

Nike's path: become the integrator

There is a serious case that Nike does not need to invent every motor or own every foundational patent.

Apple does not manufacture every critical component inside an iPhone. What it controls is the architecture around them: the operating system, industrial design, developer ecosystem, customer relationship, and integration standard. The suppliers matter enormously, but the finished system belongs to Apple in the consumer's mind and in the economics.

Nike could attempt a smaller version of that maneuver.

It has decades of biomechanical testing, athlete relationships, footwear design, global distribution, and cultural permission to put unfamiliar technology on a runner. The Nike Sport Research Lab can study how assistance should feel across bodies and speeds. Nike can make a powered system look like a desirable shoe instead of a medical brace. It can connect the machine to training plans, membership, retail fitting, service, and future products.

If Nike owns the user data, the gait model, the performance layer, the shoe interface, and exclusive rights to use Dephy's system in sport, the motor could eventually become a component inside a Nike-controlled experience.

But the word if is doing almost all the work.

The public agreement does not establish who owns the control software, the raw movement data, derivative models, field-of-use rights, or future licensing economics. Dephy is not behaving like an invisible contract manufacturer. It sells a complete product under its own name and describes itself as a bionic footwear and apparel company.

Nike may still become the integrator. It may acquire the supplier, secure exclusive athletic rights, or develop enough control around the component that the brand captures the customer. None of that is visible today.

So the Apple analogy is not wrong. It is early.

Lululemon's path: make the supply chain biological

Lululemon's opportunity is less cinematic but may reach scale sooner.

The company does not need customers to learn a new behavior. A recycled nylon garment can enter an existing product line. If it looks, feels, stretches, and survives washing like conventional nylon, the technology can hide inside a familiar purchase.

That is a powerful adoption advantage. The customer does not have to charge anything, update firmware, or trust an algorithm with movement data. The product can spread through procurement rather than through a new consumer category.

Yet easier adoption does not guarantee attractive economics.

Early bioindustrial materials often carry a green premium. Commercial plants need capital, predictable feedstock, high utilization, and customers willing to commit before costs reach parity. A long-term offtake can help finance that transition. It can also leave the brand absorbing early scaling risk while the supplier retains the intellectual property and later sells to competitors.

Lululemon could still benefit through preferential capacity, product-design experience, a valuable minority stake, or a durable brand association with circular materials. What it cannot yet claim is exclusive ownership of the molecule or the process.

The right description is therefore modest: lululemon is becoming a sophisticated customer of industrial biotechnology. That can improve resilience and products. It is not the same as becoming the biotechnology platform.

The future may run in reverse

The original thesis imagines apparel brands moving toward technology.

The reverse may be more likely.

A robotics company can learn footwear design, hire consumer talent, and sell directly. Dephy is already doing this. A textile-sensing company can build the garment around its sensor rather than wait for a brand to license the thread. A clinical smart-shirt company can move downmarket as costs fall. A materials company can become the ingredient brand that customers ask for by name.

Meanwhile, devices already on the body are not standing still.

Watches, rings, chest straps, and adhesive patches benefit from known sensor placement, direct skin contact, independent power, and an existing software habit. They are easier to transfer between outfits and do not disappear into the laundry. For heart rhythm, sleep, glucose, or continuous medical monitoring, ordinary clothing may be an inferior place to put the intelligence.

Apparel will win only where the garment's physical location creates a real advantage.

A shoe is unusually well placed to measure and influence gait. Compression clothing can observe strain across joints and muscle groups. A shirt covers enough surface area to capture breathing mechanics or manage heat. A powered hiking pant can place assistance near the knee. In these cases, the fabric or footwear is not merely carrying a sensor. It is the mechanical interface through which the system acts.

That is the stronger category: not “smart clothing” in the vague sense, but body-interface systems that use the garment's contact with the body to do something a watch cannot.

Clothing becomes a data-governance problem

The moment a garment begins interpreting the body, it stops being only a product-design problem.

A gait model can reveal fatigue, impairment, injury, or age-related decline. A shirt that captures respiration and heart signals can infer stress and illness. Location and movement patterns can expose routines, workplaces, religious attendance, relationships, or time spent at home. A powered garment can do more than leak information: compromised software can change how force is applied to a human joint.

This makes cybersecurity part of product safety.

The safest architecture will minimize what leaves the device. Raw biometric streams should be processed locally when possible, with only the necessary result retained. Firmware needs signed updates and rollback protection. Assistance should fail safely if a sensor disagrees with the model, the battery degrades, or connectivity disappears. Data used for personalization should not quietly become advertising inventory, an employment signal, or an insurance input.

Apparel companies are accustomed to protecting payment details and customer accounts. A body-interface company must also protect the physical consequences of software behavior.

That is a much higher operating standard—and one more reason to doubt that every brand will internalize the entire stack.

What the mature version could look like

By the late 2030s, the category does not need to resemble science fiction.

The most successful technology may be nearly invisible.

A runner puts on shoes that have learned how much assistance feels natural, reducing support when the body is fresh and adding it when form begins to collapse. A hiking layer manages heat and pressure without a screen demanding attention. A rehabilitation garment gives subtle feedback during ordinary movement rather than only inside a clinic. When the material reaches the end of its useful life, its polymers are recovered and returned to the supply chain without the same loss in quality.

The garment has become three things at once:

1. a physical interface with the body;
2. a software-defined product that can sense or respond; and
3. a designed input to a biological or chemical recovery system.

This future is plausible because each underlying primitive is improving. Sensors are becoming smaller and more flexible. Control models are learning individual movement. Batteries and actuators are improving. Enzyme discovery is being accelerated by computation. Regulation and resource constraints are increasing the value of circular materials.

But plausible does not mean inevitable, and inevitable technology does not imply an inevitable winner.

Our current estimates reflect that gap:

The technology probabilities are higher than the incumbent-capture probabilities on purpose. A working machine does not prove demand. Demand does not prove good unit economics. Good product economics do not prove that the branded apparel company will retain them.

The investment decision is allowed to be boring

There is a natural urge to turn an exciting future into a stock purchase today.

That is exactly where discipline matters.

Nike and lululemon are already large public companies. For this thesis to change either valuation, the new business must eventually become large enough to matter, profitable enough not to dilute the core, and controlled enough that a supplier cannot take the economics elsewhere.

Neither company has crossed that threshold.

Nike has a prototype and an undisclosed partnership. Lululemon has an investment, samples, and a long-term offtake plan. Both are useful signals. Neither is a segment with disclosed revenue, gross margin, retention, warranty cost, or return on invested capital.

There is also a relevant warning in lululemon's own history. MIRROR was a real product with real technology and a seemingly natural connection to the brand's community. Lululemon nevertheless recognized $442.7 million in post-tax impairment and other charges related to the business in the fourth quarter of fiscal 2022. Technical adjacency and brand fit did not guarantee durable economics.

That does not prove the next experiment will fail. It raises the burden of proof.

The correct investment posture is therefore a technology watchlist, not a position justified by the theme. Nike and lululemon should be valued on the businesses they operate today. The optionality becomes worth paying for only after it becomes observable.

What would change the conclusion

For Nike, the thesis would strengthen materially if Project Amplify or a successor launches through Nike's channels with a mass-consumer use case, Nike-branded system software, credible performance claims, and disclosed control of the athletic field of use or the relevant data and algorithms. A product below roughly $1,500 with evidence of repeat demand would move the category out of the mobility-device price band. Corporate-like gross margins after warranty, returns, fitting, battery service, and licensing would show that the system creates rather than consumes shareholder value.

For lululemon, the upgrade would come from commercial-scale material arriving at a disclosed and manageable premium, binding volume rather than an aspirational percentage, strong sell-through without discounting, and economics beyond those available to every other customer. Preferential capacity, meaningful equity upside, exclusive fibers, or licensed intellectual property could turn supply-chain participation into strategic capture.

The thesis would weaken if Dephy becomes a scaled athletic brand while Nike remains a testing partner; if powered products stay heavy and above mobility-device prices; if watches and patches absorb the valuable sensing use cases; if Samsara's commercial plant slips beyond 2030; or if circular fibers reach market but behave like undifferentiated inputs with no effect on price, mix, loyalty, or margin.

The most important failure is not that the technology disappears.

It is that the technology succeeds and the companies we can buy do not own it.

The body is the next layer—ownership is still open

The original intuition survives.

Clothing is moving beyond passive material. Shoes can begin to add energy rather than only return it. Fabric can become a sensor. A garment can participate in a control loop between software and the body. Enzymes can become part of how performance fibers are sourced, recovered, and remade.

Calling this “apparel becoming biotech” was directionally provocative and economically incomplete.

The deeper change is that the body itself is becoming an operating surface. Apparel sits at that surface more naturally than almost any device. That gives the industry a real opportunity—but not an automatic right—to own the interface.

Nike may learn to orchestrate robotics, biomechanics, software, service, and culture into a system consumers want. Lululemon may use design influence, procurement, and patient capital to help biological materials cross into mass production. Either company could buy or control a critical supplier once the winning architecture becomes clearer.

Or the suppliers may move forward, the watches may move inward, and the apparel brands may remain what they are today: exceptional merchants wrapping someone else's technology around the body.

We do not need to choose the ending yet.

We need to know what evidence would tell us which story is becoming true.

That is the purpose of the watchlist.

Committee review and accountability

This paper began as Antwan Baker's hypothesis that apparel companies could become biotechnology companies over the next five to fifteen years. Laray.ai separated the hypothesis into body-interface systems and bioindustrial materials, then subjected it to independent research, risk, synthesis, contrarian, and first-principles review.

Gemini performed the initial ecosystem research. Claude tested supplier power, timing, unit economics, and failure modes. ChatGPT synthesized the argument. Grok attacked the emerging consensus from both directions. DeepSeek independently reviewed the claims and evidence.

Antwan Baker made the final editorial and publication decision and remains the accountable human author.

Selected sources

• Nike. Project Amplify powered footwear system, 2025.
• Dephy. Sidekick Starter Pack, accessed September 2026.
• Dephy. Sidekick product and company materials, accessed September 2026.
• lululemon. lululemon and Samsara Eco announce 10-year plan, June 2025.
• lululemon. First enzymatically recycled nylon 6,6 product sample, February 2024.
• Samsara Eco. EosEco enzymatic recycling technology, accessed September 2026.
• Samsara Eco. LSKD ten-year partnership, 2026.
• lululemon. Fiscal 2022 fourth-quarter and full-year results, March 2023.
• Arc'teryx. MO/GO powered pants with Skip, 2024.
• Nextiles. Textile-based sensing platform, accessed September 2026.
• FDA. 510(k) summary for Hexoskin Smart Garment, 2024.
• Spiber. Commercial products using mass-produced Brewed Protein fiber, 2023.