Identity from physics
The wearer’s tissue vibration is upstream evidence, not an after-the-fact voiceprint guess.
Batteryless is memorable. The durable market wedge is a reliable wearer channel, visible recording, local ownership, no charging ritual on the body, and standard exports.
“When I join an in-person or hybrid conversation, capture what happened, know which words were mine, let everyone understand that recording is active, and return useful notes without making me maintain another fragile cloud-bound gadget.”
The wearer’s tissue vibration is upstream evidence, not an after-the-fact voiceprint guess.
No charge cycles, thermal event, firmware update, or storage breach on the neck.
The desk puck can display recording state to everyone, use better microphones, dissipate heat, and be serviced.
Audio and transcripts remain useful without a subscription or vendor. This directly addresses category shutdown risk.
As of July 2026, Plaud markets NotePin as a wearable recorder with necklace/pin/clip/wrist options, two MEMS microphones, up to 20 hours of continuous recording, speaker labels, and a listed price around $179 on its US product page. It is a conventional rechargeable ambient recorder, not a passive throat channel. Plaud product page.
Limitless says Meta acquired the company; it stopped selling new Pendants on December 5, 2025 and plans support for existing Pendant customers through 2026. Its official notice also describes regional service shutdowns and data export/deletion. This is unusually strong evidence that product longevity and data portability are features, not fine print. Limitless official notice.
Start with opt-in, scheduled, in-person team meetings in controlled offices where a reader puck can sit visibly on the table. This avoids pretending the first reader is pocket-sized, creates a natural consent moment, and values wearer attribution. Do not begin with healthcare, legal depositions, children, covert journalism, or continuous public capture.
| Question | Evidence needed before pricing |
|---|---|
| Will users pay a hardware premium for passive jewelry? | Blind concept tests, working demo, conjoint study, return-reason analysis |
| Can the reader reach consumer BOM? | Architecture selection, RFIC availability, PA/filter/antenna cost, certification and yield model |
| Does local AI avoid cloud COGS without hurting quality? | Representative latency/accuracy benchmarks on target compute |
| Will jewelry sizing explode SKU count? | Anthropometric fit study and modular band strategy |
| Is passive-tag replacement a revenue or trust problem? | Failure rate, cleanability, personalization demand, warranty and recycling study |
The published physics is not, by itself, a moat. Defensibility may come from a comfortable repeatable contact system; body-loaded resonator/reader co-design; a calibration corpus collected ethically across users and motion; robust alignment and uncertainty; compliance know-how; local-first trust; industrial design; manufacturing yield; and protected original inventions. See the FTO starter map.
No marketing claim ships unless its row in the claim ledger names a reproducible test, dataset, owner, date, uncertainty, and current evidence. The full product specification is in PRD.md; the staged business/engineering sequence is in roadmap.md.