[kemotep]

I guess the trick is determining the definition of “bot”. Is it how many “users that are software programs” vs “users that are disingenuously engaging with the platform to spread or amplify specific kinds of content that are either programs or real people”?

Mastodon and Twitter likely have a similar percentage of total active users that are the former. I would speculate that Mastodon may even have more bots as a percentage of active accounts than Twitter because it is a smaller platform. Think news posting accounts that just link an article the second it is published.

Twitter, being a much larger mainstream platform probably has significantly higher percentage of users being the later kind of bots compared to Mastodon because they would have more of an impact on Twitter and Mastodon admins are more likely to just ban or defederate from instances that spread content that they don’t want to engage with.

[mariusor]

I don't know about twitter, but on the fediverse the good faith bot accounts are clearly marked as such.

[MuffinFlavored]

> engaging with the platform to spread or amplify specific kinds of content that are either programs or real people

If you were Twitter/X, how would you improve their current (probably state of the art) "malicious-content" bot detection + mitigation?

It sounds like it ain't just captchas and IP banning?

[kemotep]

They don’t because it would be against their financial interests to do so.

Advertising, blue checks, admitting that they have fewer genuine monthly active users than they currently report all negatively affect their bottom line. And any moderation solution that could work would be decried as censorship and people who do genuinely engage could leave the platform.