[citation needed]

I've changed my mind arguing online, and so have plenty of others I know.

Rhetorical Use Of "[citation needed]" Considered Harmful. (By me.)

Given that your response refers to evidence that is more than sufficient to flat out invalidate the excessively general claim in question a citation wouldn't even help, never mind 'be necessary' - it'd just make the source cited look bad too.

One of many preferred alternatives to "[citation needed]" in such cases is "No, that's bullshit!"

(I like citations. I like asking for citations. I sometimes accept demands for citations to be appropriate. I would like citation requests to be statements that make it likely for citations to be provided when they are useful. This makes usages as a rhetorical device my enemy.)

I have too, but only after discovering rationalism. Original line was meant to apply to general internet.

And in the broadest strokes that statement is objectively false - I was deconverted from christianity via online argument. However the arguing was repeated confrontations over several weeks, and then the deconversion didn't take place until months later - after I'd completely lost touch with my opponent. I've never been able to contact him and thank him properly.

But for the average person, any single argument session will not noticeably change their position. I feel that adding any caveats to the original statement makes the emotional weight swing far away enough from 0 that the caveat makes the statement less true in practice. Similar to how you are better off saying "You have no chance of winning the lottery, ever" than you are saying "There is an infinitesimal chance you'll win the lottery". One cannot activate few enough neurons to properly convey the chances of winning when thinking about the chances of winning, and thus accepting "You'll never win" is closer to the truth when run on human brains.