You are right about this.
The people who understand and know how to use AI for writing are the one who can edit and make the written content sound more human.
It is only learners of AI who don't edit i.e. copy from AI and post directly paste to social media are the ones whose content can be detected by AI detectors.
I have personally used AI for crafting social media posts(for some accounts I manage) especially tweets and I do a little edits.
I'm sure 80% of Crypto Twitter uses AI for their posts but we have no proof hehe.
The crazy part is that even if somebody did copy and paste AI writing, some AI detectors still may fail to detect it. Then there are people who did not even use AI and detectors mat mistake it as AI. There’s no certainty in it at all and being able to call out some people, but not others just makes it totally illegitimate to contest someone’s work unless you have definitive proof. And yeah, most of crypto Twitter is absolutely insane and I’m sure majority of those people are using AI to profit big time. We see how well Twitter is doing compared to Hive, sadly. Embracing AI, at least to some extent, is essential to be successful in the future as a social media platform.
We'd use AI on Hive but it'd be considered not original content!
Anyway as we said it's still difficult to detect AI written content and worse still original content is being flagged as AI.