We wrote the paper on how to deslop your language model: https://arxiv.org/abs/2510.15061
It looks like a method of fabricating more convincing slop?
I think the Kagi feature is about promoting real, human-produced content.
SlopStop: Community-driven AI slop detection in Kagi Search
71–80 of 271 posts
Re: SlopStop: Community-driven AI slop detection in Kagi Search
#72> Our review team takes it from there How does this work? Kagi pays for hordes of reviewers? Do the reviewers use state of the art tools to assist in confirming slop, or is this another case of outsourcing moderation to sweat shops in poor countries? How does this scale?
Hey, Kagi ML lead here.
> Kagi pays for hordes of reviewers? Is this another case of outsourcing moderation to sweat shops in poor countries?
No, we're simply not paying for review of content at the moment, nor is it planned.
We'll scale human review as needed with long time kagi users in our discord we already trust
> Do the reviewers use state of the art tools to assist in confirming slop
Mostly this, yes.
For images/videos/sound, diffusion and GANs leave visible artifacts. There's a bit of issues with edge cases like high resolution images that have been JPEG compressed to hell, but even with those the framing of AI images tends to be pretty consistent.
> How does this scale?
By doing rollups to the source. Going after domains / youtube channels / etc.
Mixed with automation. We're aiming to have a bias towards false negatives -- eg. it's less harmful to let slop through than to mistakenly label real content.
Re: SlopStop: Community-driven AI slop detection in Kagi Search
#73Though I'm still pissed at Kagi about their collaboration with Yandex, this particular kind of fight against AI slop has always striked me as a bit of Don Quixote vs windmill. AI slop eventually will get as good as your average blogger. Even now if you put an effort into prompting and context building, you can achieve 100% human like results. I am terrified of AI generated content taking over and consuming search eng…
> Even now if you put an effort into prompting and context building, you can achieve 100% human like results. Are we personally comfortable with such an approach? For example, if you discover your favorite blogger doing this.
I don't care one bit if the content is interesting, useful, and accurate.
The issue with AI slop isn't with how it's written. It's the fact that it's wrong, and that the author hasn't bothered to check it. If I read a post and find that it's nonsense I can guarantee that I won't be trusting that blog again. At some point there'll become a point where my belief in the accuracy of blogs in general is undermined to the point where I shift to only bothering with bloggers I already trust. That is when blogging dies, because new bloggers will find it impossible to find an audience (assuming people think as I do, which is a big assumption to be fair.)
AI has the power to completely undo all trust people have in content that's published online, and do even more damage than advertising, reviews, and spam have already done. Guarding against that is probably worthwhile.
Re: SlopStop: Community-driven AI slop detection in Kagi Search
#74Given the overwhelming amounts of slop that have been plaguing search results, it’s about damn time. It’s bad enough that I don’t even down rank all of them, just the worst ones that are most prevalent in the search results and skip over the rest.
Yes, a fun fact about slop text is that it's very low perplexity text (basically: it's statistically likely text from an LLM's point of view) so most algorithms that rank will tend to have a bias towards preferring this text.
Since even classical machine learning uses BERT based embeddings on the backend this problem is likely wider scale than it seems if a search engine isn't proactively filtering it out
Re: SlopStop: Community-driven AI slop detection in Kagi Search
#75Where does SEO end and AI slop begin?
We have rules of thumb and we'll have a more technical blog post on this in ~2 weeks.
You can break the AI / slop into a 4 corner matrix:
1. Not AI & Not Slop (eg. good!)
2. Not AI & slop (eg. SEO spam -- we already punished that for a long time)
3. AI & not Slop (eg. high effort AI driven content -- example would be youtuber Neuralviz)
4. AI & Slop (eg. most of the AI garbage out there)
#3 is the one that tends to pose issues for people. Our position is that if the content *has a human accountable for it* and *took significant effort to produce* then it's liable to be in #3. For now we're just labelling AI versus not, and we're adapting our strategy to deal with category #3 as we learn more.
Re: SlopStop: Community-driven AI slop detection in Kagi Search
#76So we have two universes. One is pushing generated content up our throats - from social media to operating systems - and another universe where people actively decide not to have anything to do with it.
I wonder where the obstinacy on the part of certain CEOs come from. It's clear that although such content does have its fans (mostly grouped in communities), people at large just hate arificially-generated content. We had our moment, it was fun, it is no more, but these guys seem obsessed in promoting it.
Re: SlopStop: Community-driven AI slop detection in Kagi Search
#77We wrote the paper on how to deslop your language model: https://arxiv.org/abs/2510.15061
Slop is about thoughtless use of a model to generate output. Output from your paper's model would still qualify as slop in our book.
Even if your model scored extremely high perplexity on an LLM evaluation we'd likely still tag it as slop because most of our text slop detection is using sidechannel signals to parse out how it was used rather than just using an LLM's statistical properties on the text.
Re: SlopStop: Community-driven AI slop detection in Kagi Search
#78So we have two universes. One is pushing generated content up our throats - from social media to operating systems - and another universe where people actively decide not to have anything to do with it. I wonder where the obstinacy on the part of certain CEOs come from. It's clear that although such content does have its fans (mostly grouped in communities), people at large just hate arificially-generated content. We…
> I wonder where the obstinacy on the part of certain CEOs come from.
I can tell you: their board, mostly. Few of whom ever used LLMs seriousl. But they react to wall street and that signal was clear in the last few years
Re: SlopStop: Community-driven AI slop detection in Kagi Search
#79Earlier quoted context omitted.
If your wife can't detect that you told your secretary to buy something nice, should she care?
This is an absurd comparison - you (presumably) made a commitment to your wife. There is no such commitment on a public blog?
Is it that absurd?
We have many expectations in society which often aren't formalized into a stated commitment. Is it really unreasonable to have some commitment towards society to these less formally stated expectations? And is expecting communication presented as being human to human to actually be from a human unreasonable for such an expectation? I think not.
If you were to find out that the people replying to you were actually bots designed to keep you busy and engaged, feeling a bit betrayed by that seems entirely expected. Even though at no point did those people commit to you that they weren't bots.
Letting someone know they are engaging with a bot seems like basic respect, and I think society benefits from having such a level of basic respect for each other.
It is a bit like the spouse who says "well I never made a specific commitment that I would be the one picking the gift". I wouldn't like a society where the only commitments are those we formally agree to.
Re: SlopStop: Community-driven AI slop detection in Kagi Search
#80This is so, so exciting. I hope HN takes inspiration and adds a similar flag. :)