NEURO-INFRASTRUCTURE // COGNITIVE ROI RADIUS
REVISION LOG — 4 ENTRIESMOST RECENT: 2026-09-16
- 2026-09-16CORRECTION — participant count reduced from 26 to 21. The 26 figure was not supported by the primary source. Neuralink's own updates page carries no announcement after 28 January 2026, and that announcement states 21 participants; the only material we could find asserting a 26th patient was a consumer-electronics commerce blog, which does not meet our sourcing bar for a clinical figure. The error was internally visible the whole time: this dossier's own deep-analysis block has said 21 (January 2026) since revision 3, so the page carried both numbers at once. Also corrected: the zero-serious-adverse-events record is now labelled self-reported and not independently verified, per our own discard rule on evidence owned end to end by the party being graded. The E·C·I scores are UNCHANGED (55 · 60 · 55) and so is the grade — nothing about restoration-versus-enhancement moved. A participant count is not the claim this dossier grades; it is context, and context still has to be true. Found by a scheduled drift check, not by a reader.
- 2026-07-04Deployed full deep-analysis (two-layer neural architecture + enhancement controversy panel). Clarified: PRIME at 21 participants (Jan 2026); VOICE speech trial counted separately.
- 2026-07-03Added cumulative-implant trend chart across PRIME + VOICE.
- 2026-05-18Initial dossier issued. Enhancement-in-healthy claim graded NOT DEMONSTRATED.
Every entry above was written when the change was made, not reconstructed afterwards. Where we were wrong, the wrong version is named.
Invasive brain-computer interface has crossed from theory into surgical reality — but strictly as medical restoration, not enhancement. Neuralink's most recent public announcement, dated 28 January 2026, puts the programme at 21 participants, spanning the PRIME (motor) and VOICE (speech) studies with expansion into the UK, UAE and Canada. The company reports no serious device-related adverse events; that record is self-reported and has not been independently replicated. The accessible layer for a non-pathological operator remains non-invasive EEG: focus and attention telemetry, not cortical control.
There are two distinct neural vectors, and conflating them is the category error of the entire space. Layer one is the invasive brain-computer interface — surgical, clinical, built to restore lost function. Layer two is the non-invasive operator layer — EEG telemetry and low-current stimulation, available now, built for measurement and modest training.
One restores agency to the paralyzed. The other converts 'focus' from a feeling into a signal you can track and train against. Neither, as of 2026, has been shown to add raw capability to an already-healthy brain. This dossier holds that line.
Invasive BCI — The Clinical FrontierTIER A (RESTORATION)
Neuralink's PRIME study reached 21 participants (January 2026), spanning quadriplegia from cervical spinal-cord injury and ALS. The N1 implant (1,024 electrodes) sits in the motor cortex. Participant P-18 (Jon L. Noble) reported seamless control of a MacBook by thought within weeks. A separate VOICE trial is decoding neural signals into phonemes → real-time synthesized speech for patients with severe communication impairment. Neuralink reports zero serious device-related adverse events to date.
Every implant to date restores a function that was lost — it does not enhance one that is intact. There is no peer-reviewed evidence that invasive BCI raises cognition in a healthy operator. 'Neural enhancement' as a consumer promise is, as of 2026, entirely unvetted.
Implantable BCI is clinical-trial only — restoration of agency for the paralyzed, not an elective upgrade for the intact. Restoration ≠ enhancement. Treat any pitch that blurs the two as marketing, not science.
The Operator Layer — Non-InvasiveCONSUMER / TIER B
Dry-electrode consumer EEG converts attention and focus into a tracked signal and supports neurofeedback training of focus baselines. This is measurement, not cortical control. It is a legitimate telemetry tool — it turns a subjective state into a number you can train against. Enhancement claims beyond that remain aspirational.
A 2024 meta-analysis (6 RCTs, 323 participants) found transcranial direct-current stimulation combined with cognitive training significantly improved working memory in healthy older adults — at 2 mA, across 10 or more sessions. The effect is real but modest, dose-dependent, and best-evidenced in aging populations paired with training.
Robust enhancement in young, healthy, high-performing operators is not established. Most positive signals come from older adults or are inseparable from the training they are paired with. Home-device marketing consistently outruns the evidence.
The frontier promise is cognitive enhancement. The evidence does not answer with a yes or a no — it answers by population. Who you are decides what the data supports.
tDCS paired with cognitive training produced statistically significant working-memory gains in healthy older adults (2024 meta-analysis, 6 RCTs, 323 participants). Real, dose-dependent, and modest — strongest at 2 mA over 10+ sessions.
tDCS + training meta-analysis — Front. Aging Neurosci. (2024) ↗Enhancement in already-high performers lacks robust RCT support; documented effects are inconsistent and small. The population most targeted by marketing is the one with the weakest evidence.
Neuralink PRIME/VOICE — clinical restoration only ↗EEG reliably tracks attention and supports focus training. It raises awareness of a state — it does not raise raw capability. A tracked signal is the deliverable, not a cognitive overclock.
Consumer EEG — attention telemetry (operator-accessible) ↗The honest read: stimulation and neurofeedback are measurement and modest-training tools, strongest in aging or deficit populations. For the healthy operator, today's yield is a tracked signal and a small, training-dependent edge — not a cognitive overclock. OCCABUZZ grades clinical restoration at high confidence and enhancement-in-healthy as unproven. Calibration over certainty.
The neural asset is not upgraded by a gadget — it is measured, trained, and, when damaged, restored. The operator who tracks focus as a signal and trains against it compounds a real but modest edge. The one who buys 'brain enhancement' on a marketing promise buys the promise, not the outcome. Restoration is here now. Enhancement is a projection — and OCCABUZZ will grade it the moment the human RCTs land, not before.
- ›Consumer EEG focus wearable (dry-electrode headset/earbuds) — attention & flow telemetry, available now.
- ›Neurofeedback training apps paired to the EEG stream — trainable focus baselines.
- ›Neuralink-class implants — clinical trial only; restoration of function, not cognitive upgrade.
Today the realistic yield is measurement: an EEG wearable converts 'focus' from a feeling into a tracked signal you can train against. Implantable BCI restores agency to the paralyzed — it does not yet add capability to the intact.
Certainty ends at restoration. No peer-reviewed evidence shows invasive BCI raising cognition in a healthy operator. Treat 'neural enhancement' marketing as unvetted.
Peer-reviewed trials — sample size, effect size, stage.
Applies to healthy apex, not only to clinical deficit.
Survives a demanding calendar. Zero executive friction.
Deployable with named caveats. Context-dependent.
The score is a geometric mean — a single failed layer collapses it. Excellence in two cannot rescue a gap in the third. That is why hype scores low and proven, feasible, broadly-applicable work scores high. The restraint is the product.