Who knows what scale they're operating at. The problem with this kind of bot is that you only really notice if they're doing a bad job (theoretically). This might be someone who wrote an LLM bot for a lark, a small-time social media botter testing a variant for fedi deployment, or an established bot trainer with dozens or hundreds of accounts that's field-testing a more aggressive new model. I doubt you could get away with hundreds of bots like this on lemmy, I think the actual user pool is small enough that we'd notice hundreds of bots posting at this volume. but again, I don't really know how I'd detect it if it were less "obviously smells like LLM slop" than this one. In bot detection, as in so many fields, false negatives are a real bitch to account for.