probably if we assume messages between ai and hooman become indistinguishable, would need a way to get over the sybil problem although, and i think this will be a more gated/centralized/hybrid solution than a fully decentarlized one
agreed—do you think thats a reasonable assumption to make rn? re: de/centralized solutions, one can get philosophical rather quickly trying to define what identity even is, and thus, how best to approach verification. curious: what makes you think a hybrid solution will emerge?