Interface Bandwidth Gap

A vat needs to read from and write to a large fraction of the cortex at once. Today's best implants sample about a thousand channels; the brain has about 86 billion neurons. Slide the channel count and watch the gap, the doublings it would take to close, and why writing is the harder half.

▣ Real-science companion · Audit §4.4
CHANNELS vs NEURONS — count, log scale 10⁰10²10⁴10⁶10⁸10¹⁰ 8.6×10¹⁰ neurons 10³ channels ~7.9 orders of magnitude Reading is hard; writing coherent input back to cortex without seizure or distortion is harder, and unsolved at scale.
Interface parameters
~10³
Fraction of brain sampled
Gap to neurons
Doublings to close

The human brain has about 8.6×10¹⁰ neurons. The Neuralink N1 samples 1,024 electrodes across 64 threads; the Synchron Stentrode fewer. That is roughly one channel per 10⁸ neurons — characterizing an 86-billion-instrument orchestra from about a thousand microphones. Closing the gap to even one channel per neuron is seven to eight orders of magnitude, and the "doublings to close" figure assumes each doubling of channel count is a discrete engineering step.

Reading and writing are not symmetric. Recording neural signals is hard; writing coherent, high-fidelity sensory input back into the cortex — continuously, without inducing seizure, adaptation, or perceptual distortion — is substantially harder and unsolved at scale. A vat has to write continuously. As a reference, the human optic nerve carries on the order of 10 Mbit/s; a full-sensory feed is higher. No known physical law forbids closing the gap; the barrier is materials, heat, and biocompatibility.

Grounded in the Feasibility Audit, §4.4. Pairs with the Acceleration Ceiling (sensory augmentation sits on this interface) and the Feasibility Calculator.

Further reading: Koch, K. et al. (2006). "How much the eye tells the brain." Current Biology 16(14), 1428–1434. Neuralink PRIME study (2024+) — N1, 1,024 electrodes / 64 threads.

v1.0 — Audit §4.4 · Code MIT · Prose CC BY 4.0 · No PII, no network calls at runtime. · Sitemap