Contested and overstated claims (the hype cycle)

BCI attracts extraordinary claims, and a mature reading of the field requires distinguishing demonstrated results from aspirations. Recurring patterns of overstatement include reporting offline accuracy as if it were online usable control; framing a single-subject or single-session result as general; equating channel count with capability; presenting cooperative, calibrated decoding as involuntary mind-reading; and roadmaps that extrapolate one metric, such as channels, while ignoring binding limits like write fidelity, biocompatibility, and the bandwidth of behavior. Company demonstrations, press releases, and preprints deserve the same scrutiny as peer-reviewed, replicated, controlled results.

Honest scholarship separates three categories: established, meaning replicated, controlled, and in humans where claimed; plausible, meaning demonstrated in animals or narrow settings with a understood mechanism; and speculative, meaning no demonstrated pathway. Much public discourse silently promotes speculative claims to established. The corrective is unglamorous but essential: preregistration, online metrics, clinical-grade endpoints, adversarial review, and explicit statements of what a result does not show.