AI uBlock Blacklist
github.com202 points by rdmuser 15 hours ago
202 points by rdmuser 15 hours ago
I like the idea and even considered contributing to the list, but this stopped me:
> NAQ (Never Asked Questions)
> My website is on your list!
> Cry about it.
That's quite a suspicious attitude. Clearly the maintainer believes he is infallible. I understand the emotions behind this, but this is not how a public blacklist should be maintained.
> but this is not how a public blacklist should be maintained.
Cry about it.
There's nothing in that repo that even pretends to be flawless, impartial or anything else. The sheer amount of mental denial of service that having to deal with SEO slopshitters opening issues saying that they promise their substack is totally written by hand makes this an impossible task.
Ban first, ask questions later. If you find that some rules are unfair, edit them yourself, for your personal usage.
I agree.
I find it a bit ironic that this site regularly talks about banning whole countries and IP ranges on our servers, then acts shocked when users do the same. The fact that somebody went to the effort to create and share this shows how poorly the public sees the web.
The reality we face is "Check your AdBlocker" is the new "Check your spam folder" and we should adjust accordingly.
Yuuup. My personal website has been inaccessible to a few friends, they thought my server was down. It turned out they had some blocklist (not related to AI) installed on their PiHole, and for whatever reason my website was on that list. It is, in fact, to this day, because my request to unblock it went completely unanswered. I still don't know why the website is on the list.
Go to the Adguard GitHub (or use the extension) and report it. And get all your friends to switch to Adguard extension and Adguard Home (Pi Hole alternative) as blockers.
Easylist and its sublist are notorious for being poorly maintained and ignoring issues opened against it. Adguard is much more active in maintaining its lists. Especially Adguard its language blocklists have much, much less breakage and missed ads than Easylist.
>> And get all your friends to switch to Adguard extension and Adguard Home (Pi Hole alternative) as blockers.
Nice of you to slip this "easy" step into your advice. Give me a break!
Perhaps it got hacked and was hosting malware without you being aware? They are pretty good at hiding it from the site owner (showing the original website to you, but not to others).
The server is and has been clean the whole time. I don't even run WordPress or anything similar on that server that would be a common hacking target. If it was hacked, I'm pretty sure Google Safe Browsing would be the first to flag the site, not some random PiHole list.
I would add that with this attitude and how new this initiative is, there's very little chance it will still be updated 5 years from now. Really this sort of thing needs to come from Easylist or similar, who have a track record of maintaining these for years.
I don't understand the need for the author to commit the rest of his life to this or start a foundation. It is a good list for now and if its never updated again, that seems fine.
in that case they should just contribute to one of the existing, more established lists. We don't need n+1 standards...
Probably because there's about the same chance of them being innocent as the "Help I was wrongfully banned by VAC :(((" posts in the Counterstrike community.
Reminder that false positives are not only possible but likely. I remember one instance where you could get people banned by sending them a specific string of characters over chat. Anticheat was scanning the entire contents of RAM looking for it.
These days anticheat software is likely to snap at anything. Who knows what they think of the development tools Hacker News users are likely to have on their computers? They really hate virtual machines for example. There's no telling how they'd react to a debugger or profiler.
Yeah that's what the people love to say on the Steam forums when they've gotten busted in one of its many ban waves.
Not any more it seems: https://github.com/alvi-se/ai-ublock-blacklist/commit/7ebaa7...
If the website is not AI slop, presumably they would remove it from the list.
Also seems a bit hypocritical given the screed about how such a list is necessary because the AI content might output hallucinations or damaging content without review.
But if it’s the author’s blocklist that is wrong, unverified, and causing harm to others? Cry about it.
The false positive problem gets worse over time too. Domains get sold, sites pivot, old content gets removed. A blocklist with no removal process and a "cry about it" attitude in the FAQ is basically a one-way reputational blackhole. At minimum it needs an expiry or re-review mechanism. Even browser safe browsing lists re-check URLs periodically.
AI-generated content vs human-generated content is merging as such a fast pace that such a list doesn't seem like a scalable general solution
Ublock Origin also already has an “AI widget” blocklist you can enable. Literally the only extension that keeps me on Firefox because of how useless it is on Chromium.
A new more grounded list focused on specifically blocking content farms and similar low quality sites.
A nice alternative to this very broad anti ai list: https://github.com/laylavish/uBlockOrigin-HUGE-AI-Blocklist
Edit: Oh I should mention I found it through reddit and there is some good discussion there where they describe how they find stuff etc: https://www.reddit.com/r/uBlockOrigin/comments/1r9uo3j/autom...
So there is a spreadsheet of websites. That is very interesting. There was an article here sometime ago about a media group who have so many super SEOd websites. They all have common footer text. I searched and added as many as I could find in uBlacklist. I have a gist listing them and how I searched for them. You might find that useful.
Edit: https://gist.github.com/SMUsamaShah/6573b27441d99a0a0c792431...
The broad list seems to just be a hater list. It's not trying to cover cases of deception (passing off AI material as if it's something else), as it includes sites which are very open about what kind of content is on there.
Would you say the same about a block list that blocks anything else? I don't care how obvious an ad is, I don't want to see it. Same with social widgets or cookie consent banners, or newsletter sign-ups.
But I wouldn't call the person that maintains the news letter popup block list as "newsletter hater"
>Would you say the same about a block list that blocks anything else? I don't care how obvious an ad is, I don't want to see it. Same with social widgets or cookie consent banners, or newsletter sign-ups.
He's not complaining that widgets for his favorite social network site is getting blocked, he's complaining that anything vaguely related to social networks are getting banned. Some of the sites on that list are stuff like chatgpt.com, which might be AI related, but clearly doesn't fit the criteria of "AI generated content, for the purposes of cleaning image search engines".
The purpose of the broad list is removing AI-generated content from search results, so that the user doesn't have to wade through (as much) slop to find the human-created content they're looking for.
While I applaud the honesty of sites that are open about their content being AI generated, that type of content is never what I'm looking for when I search, so if they're in my search results it's just more distraction/clutter drowning out whatever I'm actually looking for. Blocking them improves my search experience slightly, even though there is of course still lots of other unwanted results remaining.
Granted, I definitely count as an AI hater (speaking of LLM's specifically). But even if I weren't, I don't think I'd be seeking it out specifically using a search engine; why would I do that when I could just go straight to chatgpt or whatever myself? Search is usually where people go to find real human answers (which is why appending "reddit" to one's searches became so common). So I see this as a utility thing, more than a "I am blocking all this just because I hate it" thing. Although it can be both, certainly.
Edit: removed an off-topic tangent
Hasn't been updated in 5 months
Oh good point I also overlooked that with the anti ai list.
The big anti ai list also seems to be focused on hiding links from ddg/bing/google where this new more focused list just blocks sites. I tend to like block ones vs hiding because they pop up a nice warning no matter where I came from and I can still decide to ignore it if I want so they is more user agency instead of just quietly hiding a unclear chunk of the net from search engines.
Not necessarily disagreeing the whole principle...
> All I hear is skill issue. Imagine needing an AI to write stuff.
Grammarly users (and underrepresented non-English speakers) would complain.
There’s not a single group who’s ever been told skill issue that didn’t complain
Sure, but there also plenty of times “get gud!” is used for gate keeping. Life is on a continuum, man.
Personally I find that I prefer badly written english or auto-translated stuff written in languages foreign to me over ai generated or even just ai polished works I've seen. There is just so much more character, depth and variance there vs ultra ai generic or slop text.
That being said this project seems focused on content farms not people who just need a little help writing so this whole conversation is a bit of a side tangent.
One of my coworkers is EXTREMELY capable but functionally almost illiterate. He’s recently discovered that he can put an idea in Copilot and have it generate an email. So now instead of brief, correct, but difficult to parse emails we receive 20-paragraph, bulleted, formatted OpenAI slop. It’s been a very strange thing to see, like someone getting extraordinarily bad cosmetic surgery.
"One of my coworkers is EXTREMELY capable but functionally almost illiterate."
I cannot imagine what it means. To me it reads like "I know someone who can run very fast but has no legs."
Capable doesn't mean capable of office work though, I could see someone with a language disorder doing electronics and have trouble with words, not numbers. Or someone who has trouble with written words specifically doing most of their learning with classes and videos.
Exactly right. The individual in question produces excellent deliverables within their space. They, the coworker, are very good at receiving inputs, but not very good at outputs (other than their deliverables). In a way, it's like having an offshore worker who speaks almost none of your language but can understand it and produce good work.
Sam, and when you ask them a deeper question about it on a call they usually have no idea. It’s making people very lazy.
I have a similar coworker, but he's not great at prompting, so 10% of the time the AI version of himself makes confident assertions that he did not intend and are clearly not true. Genuinely no idea what I'm supposed to do about it.
Exactly right. He’s good at what he does, except communicating, and people are beginning to associate him with AI slop they don’t have time to read rather than the excellent work he does for them.
Yeah I hate it when people do that and I always call them out on it.
Unfortunately our company is trying to be "AI First" so they'll just point to that and continue their bullshit.
Our company literally promotes AI slop over personally made content even if it's mediocre crap. All they care about is rising usage numbers of things like copilot in office.
I mean, I know it is probably tongue in cheek but that never-asked-question was particularly out of place. Massively generated AI contents are usually not THAT thoughtful anyway.
This specific list from this specific author isn’t worth using since they refuse to remove items from the list if domain ownership changes.
E.g., bought a domain that previously hosted AI content.
E.g., Whitehouse.com used to be a porn site, now it’s not.
If you don't know English and you want to write English anyway, please just use a machine translator.
From experience: If you don't know Danish, please don't ever use machine translators to translate from English. Regardless of what some people may think, they make mistakes, so many mistakes.
I get why it's tempting, good translators are expensive, and few and far between. A friend of my is a professional translator and she's not exactly in need of work, but a lot of customers look at her prices and opt for machine translations instead and the result not always impressive. Errors range from wrong words, bad sentence structure to an inability to correctly translate cultural references.
Right, makes sense for Danes, or other population where English knowledge is basically ubiquitous. But I'm think it might look differently in other places, if the choice is between "Badly translated but I can understand 95% of it" and "In a language I don't understand at all, maybe 1% I could figure out", then the choice might be a bit different.
nope, let the user does the translation, with his own choice of tool and being thus perfectly aware of the shortcomings.
I know that some people translate my French posts to read them. That’s really cool. But I would never post something I didn’t write myself (but I use spellcheking tools. I even sometimes disagree with them)
There are levels to things. In a professional context (including product design and documentation/instructions) don‘t use machine translation[†].
For your personal hobby site or for general online communication, you probably shouldn’t use machine translation, but it is probably useful if have B1 language skills and are checking up on your grammar, vocabulary, etc. As for using LLMs to help you write, I certainly prefer people use the traditional models over LLMs, as the traditional models still require you to think and forces you to actually learn more about the output language.
For reading somebody else’s content in a language you don‘t understand, machine translation is fine up to a point, as long as you are aware that it may not be accurate.
---
† In fact I personally I think EU should mandate translator qualification, and probably would have only 20 years ago when consumer protection was still a thing they pretended to care about.
Why? A model correcting your errors is a powerful tool to learn the language, much better than just writing the phrase in your native language.
I use Grammarly at work (it's mostly to make sure our brand guidelines are kept) and I don't find that it (defaultly) corrects too far into the ai slop territory. It's mostly just making sure your sentence is correct.
Op is going after AI slop bot farms like android authority
I mean, the reason we use grammarly is because we recognize we have a skill issue.
At least we're not yet in the phase where we have a whitelist for the internet.
Love this, I wish there were more and broader categories of sites one could block. You can always temporarily allow sites.
In the enterprise space, there are URL reputation providers. They categorize sites based on different criteria, and network administrators block or warn users based on that information.
In my humble opinion, there needs to be a crowdsourced fund (or ideally governments would take this seriously and fund it on behalf of people) for enabling technologies that allow user friendly internet experiences. Browsers, frameworks, vpn providers, site-reputation, deceptive content, dns-providers, email providers,trusted certificate authorities(no,google and microsoft shouldn't get to police that), nation-state or corporate affiliations,etc... You shouldn't need to setup a pi-hole.
Imagine a $1B/yr non-profit fund for this stuff. if 10M people paid $10/mo that's $1.2B/yr. Proton has $97M revenue in 2024 and 100M total accounts (I don't know how many pay but the spread is roughly $1/user). I really think now is the time to talk about this when so many are wary of US tech giants and looking for other opportunities.
Also need a rule that filters out HN submissions from that Simon Wilson guy.
Why? He posts high-quality content that's interesting if you care about that field. It's not my cup of tea, but it's pretty far from what this list tries to block.
I take it you're talking about the user here with the nick simonw? I find his comments on HN interesting and balanced: don't know why you think he should be filtered out.
Why is apnews.com on the list?
Also check the https://botblock.ai/ , AI extension to detect AI replies on twitter
That's a curious one, Twitter is worthless anyway. Before AI bots proliferated, the change to rank paid accounts high in replies turned it into a de facto entry level $8/month advertising tier.
Glad we're moving in this direction, I've also got a tool that I use to determine if writing is AI using common tropes and reconstruct the OG prompt from it: https://tropes.fyi/aidr
I would rather have a whitelist that adds a nice tag at the end of the link, indicating that overall it has high quality content. This also forces you to periodically check the sites you've whitelisted
Meta question: do you guys feel the adblockers will maybe not be that important in the future? As for myself, I ended up to use just a few websites, but those are reputable and I don't mind a few ads they provide. The only adblock which is still very much needed is one for Youtube.
According to uBlock Origin it blocked 9.5 million requests to ads/third party trackers since I installed it. So yes, it's very much needed.
I feel that blocking, substituting, and even inserting user-defined resources for a website must be a native browser feature.
I dont think this is a sign of the times or the future. I think its just your own personal browsing habits.
I used to run pihole on a Pi and now I directly run unbound, still on a Pi. The difference on a great many sites is night and day: you simply get way fewer ads. And that's just by using a DNS blocklist.
Occasionally I'll get one site that refuses to load because I've got an "adblocker" but most sites do work fine, just with way fewer ads.
I usually now just ask agent, for example Gemini in Antigravity to check certain article or a group of articles, like "check all AI-related article in tldr.tech and tell what is interesting"... I am already a bit lazy to browse myself, and in this process I dont care about ads.
I feel like this is a bit of a sinking ship. I suppose if you want to avoid known sources of slop then this works … but beyond that it’s a bit of a lost cause. It’s like sports betting — once it’s there then there’s no saying who is (ab)using it.
Tragic twist: repo was entirely AI generated
Admirable idea and execution…but it does apply opposing evolutionary/economic pressure for AI-slop to become less detectable over time. AI will learn and adapt.
Metaphorically speaking, it’s the Borg we’re dealing with, not the Klingons. All Janeway did was slow the Borg’s progress.
Cory Doctorow wrote a story ~20 years ago about how the first sentient machines would be spam bots because their job is to pass as human, and anti-spam systems provide competitive evolutionary pressure.
He may not be too far off.
I think that's the one. I was a bit off on the timing, it's not 20 yet. Great read either way.
From the story:
“Spam-filters, actually. Once they became self-modifying, spam-filters and spam-bots got into a war to see which could act more human, and since their failures invoked a human judgement about whether their material were convincingly human, it was like a trillion Turing-tests from which they could learn. From there came the first machine-intelligence algorithms, and then my kind.”
It's actually rather difficult for SoTA models to shift tone without losing performance on various datasets, so not such a one-sided arms race.
What happens if a legitimate site (forums, wiki, etc) gets mass-spammed with slop?
Firefox already feeling more responsive.
flip it, and build green(organic) lists perhaps work towards having sites than dont just, not use AI, but never talk about it it's not just AI, search is a scam, no mojo in the world can extract the contact info for the business next door and the mountains of porncoin, scamulous garbage and hate news taking up a full 50% of whats left, does in fact make a determined effort to greenwall a section of the web something to consider
Come on guys, 2026 and you still using "blacklist". Why not BlockList?
Because changing blacklist to blocklist, master to main, etc. is a meaningless act of virtue signalling.
I'd argue it's not meaningless because the point wasn't to show inclusion but power. Nobody went for master's degrees, "master" as a rank in video games, or anything else.
Reminds me of [1]twitch.tv trying to remove "blind playthrough" as a tag to encourage inclusive language.
1. https://www.reddit.com/r/Twitch/comments/k7dvgw/twitch_remov...
I do use "blocklist" on new project and name my main "trunk" and not "master" but I'll both a) defend other's rights to use terms like blocklists and master and b) call out the virtue signalling ones who are trying to push a political agenda by trying to control thoughts (by attempting to control speech).