Comment by anileated

I can’t stress this enough, for anyone who (like myself) for some reason didn’t think this was possible: Twitter’s community notes are no different from tweets in their ability to spread misinformation.

What makes them worse than tweets (at least the original immutable tweets) at combatting intentional disinformation is that Birdwatch notes 1) are visually formatted to communicate an impression of absolute ground truth, and 2) don’t reflect any controversy or edit history.

A trending tweet with a false “correction” (such as the one that claimed a video of Xinjang police brutality against an Uyghur was showing Taiwanese police) would be viewed by millions of people, 99% of whom would read that note completely uncritically, before it would get corrected. The people who are equipped to recognize the lie, and who care to fight to get it corrected, are few compared to the army of internet trolls spreading that lie.

Eventually the note may get rewritten—but at that point the tweet is no longer trending; the operation was a success and no one even knows that it took place. Since the note gets rewritten with no history maintained, the only evidence of the original malicious, false correction would have to be in that updated note, which is obviously an unpopular choice because it makes the new note harder to read and makes future readers do extra work to untangle what happened there.

(Incidentally, one of the things that could reliably be used to combat bias—labeling tweets from accounts that are known to be associated with governments or such—was nuked by Elon right away.)