Work

The plasticity budget: why the brain's learning, not the implant's bandwidth, sets the pace of brain–machine integration

Year: 2026
Type: preprint
Abstract: Brain–machine interfaces are described by what the device provides: electrodes, channels, bits per second. What makes them work is supplied by the user. Every interface requires a particular brain to reorganise, and that reorganisation draws on resources that are limited and shared with everything else the brain learns and maintains. This article names that requirement the plasticity budget and proposes that it, and not channel count, sets the pace of integration between brains and machines. Three resources are distinguished: the repertoire of activity patterns a neural population can already produce, the cortical territory already committed to other functions, and the capacity to consolidate what is learned. The evidence for each is assessed with its limits, including results that count against simple competition. Two consequences follow. First, reading and writing are asymmetric. In reading, adaptation can be moved from the brain to the decoder, and the field has been doing so. In writing, the brain is the only learner, so the budget binds there first, and it binds hardest where the target cortex has been taken over by another function. Second, the cost is measurable with implants already in use, and it is not being measured. The article specifies a ledger of what to report, seven falsifiable predictions, one of which would overturn the framework, and design rules for interfaces that spend less of the user's plasticity.
Language: English
Open Access status: green
OpenAlex ID: W7234500058
Cites:
Cited by:
Related to: 10