The human data layer — the signal AI can't yet read

01Every AI leap has been a leap in perception

It is easy to narrate the last decade of artificial intelligence as a story about models getting bigger. The more useful story is about machines getting new senses. First they learned to read text. Then to see images. Then to hear and produce speech. Each new sense opened an entire category of systems that could not have existed the day before.

The next sense may be continuous physiological context outside the clinic and laboratory. Systems that can access reliable cardiac, muscular, neural, thermal, and movement signals over longer periods could support new forms of monitoring, research, and human–machine interaction. The open question is whether those signals can be captured with enough fidelity, comfort, durability, and governance to be useful in everyday settings.

AI can only act on what it can sense of us.

02The body remains difficult to read continuously

The body produces many signals at once: cardiac rhythm, muscle activation, temperature, respiration, movement, and more. High-quality measurement of those signals often remains episodic or constrained to specialized settings.

Consumer devices have extended selected measurements into daily life, but generally from narrower sensing locations and modalities. The unresolved gap is not that the body is unread. It is that continuous, context-rich physiological measurement remains difficult to achieve without adding friction to everyday life.

03The interface has to fit everyday life

For physiological measurement to extend beyond scheduled tests, the interface must be comfortable, repeatable, and compatible with ordinary activity. Textiles are one candidate because they can maintain distributed contact with the body while remaining familiar and conformable.

In specific tested constructions, textile electrodes have produced meaningful signals. One study recorded ECG comparable in signal fidelity to gel electrodes from knitted CEF constructions, holding r² = 0.93 after 30 wash/dry cycles. In a separate pediatric pilot, a SKIIN™ implementation measured heart rate with 3.6–3.8% NRMSE against reference ECG in twenty children.

The signal is not only cardiac. In separate studies, textile electrodes correlated with gel electrodes for muscle activity — 31 of 40 candidate materials at p<0.001 — and seamless-knit fabric electrodes resolved a decrease in alpha-band brain activity during mental effort. Each result is specific to its construction and population, but together they suggest one approach can reach more than a single physiological signal.

Those results do not establish a universal textile-sensing platform. They show that carefully engineered textile interfaces can produce useful physiological measurements under defined conditions.

04Toward an interoperable layer

No single physiological signal defines the body, and no single implementation will serve every use. The longer-term opportunity is to coordinate multiple sensing and response functions through interoperable textile, electronic, connectivity, and software layers.

MyantX's thesis is that these systems could eventually provide a governed interface through which clinical, research, and connected-product applications access the signals appropriate to their intended use. That remains an architectural direction, not a general-purpose foundation available today.

05A layer that could both sense and respond

Separate MyantX research and development programs have demonstrated sensing and actuation in textile form factors. Textile electrodes have captured electrophysiological signals. Other textile systems have delivered neuromuscular or functional electrical stimulation, while knitted structures have been developed for controlled heating.

In one closed-loop study, textile electrodes were used to regulate grasp force in one participant with quadriplegia to under 15% steady-state error. That result demonstrates a specific sense-and-response configuration. It does not establish that every sensing textile can also deliver therapy, or that the same construction performs every function.

The broader opportunity is to coordinate sensing, interpretation, and response within application-specific systems.

06What could be built on it

The practical challenge is to translate individual demonstrations into governed, application-specific systems that people can use consistently and that partners can validate for a defined purpose.

MyantX's longer-term thesis is that textiles and other soft interfaces could become an important physical interface between people and intelligent systems. The work today is to establish the materials, sensing methods, architecture, evidence, and manufacturing pathways required to make that direction credible.