Would be funny if the Wikimedia Foundation developed an AI that is trained on all the major AI models, and uses that to detect AI slop.
They are constantly changing, but one could probably get pretty far focusing on ChatGPT (which is what most “lazy” authors use).
And there are already efforts in this domain from the community, see the “slop” profiles in EQ bench:
https://eqbench.com/creative_writing.html
Traditional LLMs would be better suited (ironically) for fact checking, eg they check for citations, then go to follow the links and see if it matches the text. They’re also much better at “checking” for sanity than actually writing it out. An obviously this would just be a first pass for a person to quickly confirm.
Nothing miraculous:
One way Wikipedians are sloshing through the muck is with the “speedy deletion” of poorly written articles, as reported earlier by 404 Media. A Wikipedia reviewer who expressed support for the rule said they are “flooded non-stop with horrendous drafts.” They add that the speedy removal “would greatly help efforts to combat it and save countless hours picking up the junk AI leaves behind.” Another says the “lies and fake references” inside AI outputs take “an incredible amount of experienced editor time to clean up.”
Typically, articles flagged for removal on Wikipedia enter a seven-day discussion period during which community members determine whether the site should delete the article. The newly adopted rule will allow Wikipedia administrators to circumvent these discussions if an article is clearly AI-generated and wasn’t reviewed by the person submitting it.
The Wikimedia Foundation is also actively developing a non-AI-powered tool called Edit Check that’s geared toward helping new contributors fall in line with its policies and writing guidelines. Eventually, it might help ease the burden of unreviewed AI-generated submissions, too. Right now, Edit Check can remind writers to add citations if they’ve written a large amount of text without one, as well as check their tone to ensure that writers stay neutral.
Remember that a spell checker also works completely without AI. Computers can do advanced stuff without LLMs.
Paywalled
A little while ago there was a similar article about the same thing. The answer was: deleting stuff.
If an article matches certain criteria, it just gets deleted without any discussion. Normally, there would be a lengthy discussion phase, which works reasonably well as long as actual humans are the ones writing the articles. Now that LLMs are generating trash, you need to lower the threshold for deletion in order to keep up with the rapid pace.
Didn’t they introduce a new rapid deletion process? IIRC it’s a seven-day proposal and discussion before deletion. For suspected AI it’s basically a one day thing.
Thanks! Also: https://archive.is/MmlXG
This place needs a bot that automatically archives every link posted because most people are still painfully unaware of these tools:
You sure? Worked for me.
