BCI Cognition Enhancement is Possible
lesswrong sequence argues brain-computer interfaces can boost intelligence to solve alignment.
a lesswrong author proposes that brain-computer interfaces could boost human intelligence to a 'moderately superintelligent' level, defined as 3-6 standard deviations beyond current capability in x-risk relevant cognition. this enhanced intelligence could theoretically execute pivotal acts, such as negotiating a global pause on ai development or discovering an algorithmic solution for corrigibility. the sequence argues this approach bypasses the difficulty of aligning artificial superintelligence by instead enhancing human alignment researchers. however, the author observes that existing bci funding is directed almost exclusively at prosthetic applications rather than cognitive enhancement.