SlopStop: Community-driven AI slop detection in Kagi Search

Nice. This is needed at every place where user-generated content gets commented and voted on. Any forum that offers the option to report something as abuse or spam should add "AI slop" as an additional option.

Here's what pattern suppression actually does on a model that's trained to open its writing with "You're absolutely right.":

You're spot-on. You're bang-on. You're dead right. You're 100% correct. I couldn't agree more. I agree completely. That's exactly right. That's absolutely correct. That's on the nose. You hit the nail on the head. Right you are. Very true. Exactly — well said. Precisely so. No argument from me. I'll second that. I'm with you 100%. You've got it exactly. You've hit the mark. Affirmative — that's right. Unquestionably correct. Without a doubt, you're right.

I'm willing to bet money you can easily tag these openers yourself.

This sampling strategy and the elaborate scheme to bake its behavior into the model during the post-training are terribly misguided, because they don't fix the underlying mode collapse. It's formulated as narrowing down the output distribution, but as with many things in LLMs it manifests itself on a much higher semantical level - during the RL (at least using the current methods) the model narrows the many-to-many mapping of high-level ideas that the pretrained model has down to one-to-one or even many-to-one. If you naively suppress repetitive n-grams that are not semantically aware and manually constructed patterns that don't scale, it will just slip out at the first chance, spamming you with minor non-repetitive variations of the same high-level idea.

You'll never have the actual semantic variety unless you fix mode collapse. Referencing n-grams or manually constructed regexes as a source of semantical diversity automatically makes the method invalid, no matter how elaborate your proxy is. I can't believe that after all this time you persist in this and don't see the obvious issue that's been pointed at multiple times.

People don't call it slop because of repetitive patterns they call it slop because it's low-effort, uninsightful, meaningless content cranked out in large volumes

I like the effort, but it's super restrictive. They exclude all of Substack on principle (but weirdly, allow blogspot.com and wordpress.com). They exclude anything that isn't a blog. And they exclude blogs that aren't updated often enough.

The end result is that there's a lot of "small web" stuff that doesn't show up. Looking at my bookmarks, I think 90% of them are in the "small web" category in spirit, but maybe 10% have any chance of appearing on the Kagi list.

That's definitely true, but keep in mind the economics of cranking out AI slop. The whole point is that you tell it "yo ChatGPT, write 1,000 articles about knitting / gardening / electronics and organize them into a website". You then upload it to a server and spend the rest of the day rolling in $100 bills.

If you spend days or weeks fine-tuning prompts to strike the right tone, reviewing the output for accuracy, etc, then pretty much by definition, you're undermining the economic benefits of slopification. And you might accidentally end up producing content that's actually insightful and useful, in which case, you know... maybe that's fine.