I knew IceWhiz. You are correct that he (or rather "they") eventually was kicked from the site. But he/they operated on the site for years and was the biggest PITA you can imagine. He must have single-handedly scared away two dozen honest contributors with his BS. It is very, very easy to game the rules on Wikipedia. Wars of attrition goes on for years. Normal people don't waste their time. IceWhiz and his meat puppets have endless patience and all the time in the world.
Right. The fact that someone so terrible got 99% approval and only one anonymous investigator was able to stop them makes me think that it's likely a lot of other terrible admins who didn't have an anonymous investigator go after them probably go through the process.
And the times I've brought up the fact that Wikipedia can be unreliable before, I've had numerous editors come in and claim that wasn't true and that people could rely on the claims they find in Wikipedia. This runs counter to the claim that Wikipedia editors know about these influence campaigns and openly fight about them. A lot of the active and vocal editors are openly dismissing such concerns.
Wikipedia isn't supposed to be a source, so "reliable" here has to mean "reliably presenting a full range of notable sources". No editor should be saying you can rely on claims found in Wikipedia, except in the sense of relying that the claims are in the sources.
(Except the claim as stated isn't always in the source anyway. Best to check.)
I found Molly White's video here really useful for helping me understand the Reliable Sources policy: https://blog.mollywhite.net/become-a-wikipedian-transcript/
> The way we determine reliability is typically based on the reputation for editorial oversight, and for factchecking and corrections. For example, if you have a reference book that is published by a reputable publisher that has an editorial board and that has edited the book for accuracy, if you know of a newspaper that has, again, an editorial team that is reviewing articles and issuing corrections if there are any errors, those are probably reliable sources.
Yeah. Also, if a specific source is used a lot, it often gets put on a discussion where people vote on how reliable it is. If it's considered unreliable, the use of it will be banned.
https://en.m.wikipedia.org/wiki/Wikipedia:Reliable_sources/P...
Love the use of the "we" word here. :) What is counted as a reliable source is voted on on one of Wikipedia's meta pages. So "reliability" is not based on any factual circumstances, but on whether the vote is won or lost. And you can trivially game that using sock/meat puppetry. Notwithstanding, White's claimed policy heavily favors Western media giants such as The New York Times and The Washington Post which many editors know about. However, the actual information they publish are often much less accurate than what is published in specialized trade magazines or even activist blogs.
> So "reliability" is not based on any factual circumstances, but on whether the vote is won or lost.
Those voters (with the exception if bad actors) are working on the basis of "factual circumstances", which they debate extensively before voting.
They sure do, it's still those who amass the most votes who gets to decide. And it leads to clownish ridiculous results. ADL is listed three times as green, yellow, and red. Comment says "There is consensus that the ADL is a generally unreliable source for the Israeli–Palestinian conflict, due to significant evidence that the ADL acts as a pro-Israeli advocacy group and has repeatedly published false and misleading statements as fact, un-retracted, regarding the Israeli–Palestinian conflict." So an organization that has "repeatedly" been caught spreading false and misleading statements is still a reliable source. LOL
These debates are done in public, and we see how bad they are. It's wonderful that they're done in public, though.
I wonder if there's room in using AI to gather past edits of someone, as part of vetting, and use the sentiment analysis to check how neutral their biases are.
Neutrality != necessarily accurate or useful. And the most neutral thing to say is nothing.
And most LLMs probably have Wikipedia as a significant part of their training corpus, so there is a big ouroboros issue too.