I already don't see them.

My operating assumption is that bot problems are only get worse. Going off the very unscientific metric of "how hard captchas have gotten", it certainly feels like the arms race between bots and captchas is going in favor of the bots. As we get closer to human-ai parity, I'd imagine the ability to spam the Internet with near-impunity is low-hanging fruit.

I know bots give off other signatures in the way they traffic the web... but how hard are these to dupe as well? Is there anything else my assumption misses? These are genuine rather than rhetorical question: I have no expertise in this area and am assuming something about my assumption is wrong.

I do think stopping bots is a hard problem, but twitter seems to be doing much worse at it than other sites do and I don't think you can explain that with the arms race.

@jacksonpolack Fair. I dunno if a macro-view is entirely appropriate for what seems like a localized problem. Twitter could make some short term progress even if there's reason to be grim about the long-term problem.