Email or username:

Password:

Forgot your password?
All posts Softwarewolf's posts Post Back to profile
Softwarewolf

uBlock Origin filter list for sites that contain AI-generated content for uBlock Origin. Useful for scrubbing AI-generated bullshit from Google, DDG, and Bing image search pages.

Edit: I didn't make this! Just seemed useful.

github.com/laylavish/uBlockOri

15 March at 15:36 | Open on chitter.xyz
39 comments
Baloo Uriza

@faoluin Oh man, is there one for Pinterest?

15 March at 15:43 | Open on social.tulsa.ok.us
Softwarewolf

@BalooUriza Don't think so, not in this list anyway. But that would be very nice.

15 March at 16:51 | Open on chitter.xyz
mathew

@faoluin @BalooUriza uBlacklist will let you put an end to Pinterest garbage results, as well as the MSN spam in DuckDuckGo.

iorate.github.io/ublacklist/do

16 March at 17:53 | Open on universeodon.com
Jaime Herazo

@BalooUriza
Just adding "-site:pinterest.*" to searches should do
@faoluin

17 March at 17:47 | Open on mastodon.ie
Lufty Foxyloon

@faoluin Oooooh~ :blobfoxaww: Just installed it! Veeery useful~

15 March at 17:33 | Open on blimps.xyz
Tea :verifiedtransmasc: :verifiedambiamorous:
Softwarewolf

RIP my notifications, gonna have to mute this. But glad to see Fedi hates LLM pollution as much as I do.

15 March at 18:36 | Open on chitter.xyz
Essem :skeeter:

@faoluin is there a list that blocks the sites themselves instead of just the search results on specific engines?

15 March at 18:46 | Open on wetdry.world
Softwarewolf

@esm Don't know, but I assume you could do something like that with a Pi Hole.

15 March at 18:52 | Open on chitter.xyz
β™ΏπŸ”ž Many the Mousetacular

@esm @faoluin This one needs submissions but is aimed at blocking specifically websites that use AI art: github.com/NoraCodes/diffusion

16 March at 14:15 | Open on tenforward.social
Fiona

@esm@wetdry.world @faoluin@chitter.xyz You could probably generate that pretty easily from this list, just use the URLs it's matching hrefs against. Might just need slightly different handling for plain domain vs. (partial) path, at least for uBO. Might even be able to do that with 2 per-line regex search-and-replace actions… ​:neocat_think:​

16 March at 14:55 | Open on woem.men
noodlejetski :verified_gay:

@esm @faoluin uBlacklist addon lets you do that, but that would probably require a list in another format

17 March at 18:58 | Open on masto.ai
Fiona

@faoluin@chitter.xyz I could use something like that for those spam farms that copy random code snippets and pretend that's programming help. When searching for somewhat obscure topics there's often more of that than actually relevant results… ​:neocat_box:​

16 March at 15:03 | Open on woem.men
Kevin Karhan
Jo Jitsu
Joachim Ziebs

@faoluin Would it be possible to add the list to Pihole?

16 March at 17:47 | Open on blabber.rocks
jbaggs

@TexJoachim @faoluin This list filters on more than just sites, so you're not going to be able to get 100% there with DNS filtering.

16 March at 19:36 | Open on infosec.exchange
RooneyMcNibNug

@TexJoachim @faoluin I just did a

$ perl -lne 'print $1 while /"(.*?)"/g' list.txt > root_ai_urls.txt

against that list to produce a nice blocklist with all of the main URLs.

You could use something like adblockplus syntax w/those if you want.

I also did a full enumeration of subdomains on that list (code in the comment of one of the files below) for a more verbose URLs output.

I put it all in this repo for now - do whatever you want with this info, of course: github.com/RooneyMcNibNug/piho

@TexJoachim @faoluin I just did a

$ perl -lne 'print $1 while /"(.*?)"/g' list.txt > root_ai_urls.txt

against that list to produce a nice blocklist with all of the main URLs.

You could use something like adblockplus syntax w/those if you want.

I also did a full enumeration of subdomains on that list (code in the comment of one of the files below) for a more verbose URLs output.

17 March at 16:16 | Open on mastodon.social
billy joe bowers is tired.πŸ‡ΊπŸ‡¦
183231bcb

@faoluin@chitter.xyz How does this work? Do humans have to add every spam site to the blocklist manually? Is that sustainable given how quickly bots can generate spam? If it's automated or partially automated, how can it accurately determine what sites have generated spam?

16 March at 21:06 | Open on transfem.social
Nonya Bidniss
Colin Cogle :verified:

@faoluin Things I didn't know I needed. Thank you for sharing!

17 March at 2:37 | Open on mastodon.colincogle.name
KateYagi

@faoluin You can comma seperate the sites a filter can apply to?! That's huge, I really need to give the docs a full read-over sometime.

17 March at 10:02 | Open on denden.world
Martin Vogel
gittaca
MaxTheFox

@faoluin Nice, always wanted something like this.

I see a bit less of it in Google today than like 2 weeks ago but it's still there and kinda irritating.

17 March at 14:11 | Open on spacey.space
nintegge
pystykorva
Jessica "Jenny" O'Hare πŸ‡

@faoluin @secfurry this needs to be ootb on ublock if not already.

17 March at 15:56 | Open on mastodon.xyz
Dawn TΓ₯ke πŸŒ™:sparkletrans:

@faoluin
Added at work. We'll see how it goes! Wonder if it works on startpage.

17 March at 18:00 | Open on tech.lgbt
Rathmox

@faoluin I read the blocklist, and it's really looks bad.
It's blocks a lot of websites, yes, also blocks user accounts using AI from social media.
But it also blocks massively used websites such as shutterstock because "have a lot of AI "art"/heavily support AI "art", but also have some authentic artwork".
But the worst part to me is that it blocks, without telling users, crypto / NFT websites.
Don't get me wrong, I see why people hate these, but it's not linked to AI generated content at all. (even if there are obviously people doing that).

@faoluin I read the blocklist, and it's really looks bad.
It's blocks a lot of websites, yes, also blocks user accounts using AI from social media.
But it also blocks massively used websites such as shutterstock because "have a lot of AI "art"/heavily support AI "art", but also have some authentic artwork".
But the worst part to me is that it blocks, without telling users, crypto / NFT websites.
Don't get me wrong, I see why people hate these, but it's not linked to AI generated content at all. (even...

17 March at 18:36 | Open on shelter.moe
Rathmox

@faoluin But at the end, it won't help, 1 account blocked, 10 reappear and this list will be useless.
But I also fear that they block users because they got caught once using AI, even if they apologized.

17 March at 18:38 | Open on shelter.moe
πŸ‘»β„’ Argentinian gourd shipment 🐊

@faoluin Great find, thanks for sharing πŸ’š

17 March at 20:18 | Open on elekk.xyz
Deus

@faoluin That's a great list, actually. Got to know of AI sites I wasn't aware of or could play with..lol

But yeah, will share with friends who don't like AI stuff and might want it.

17 March at 22:05 | Open on charcha.cc
The Doctor

@faoluin I'm going to try to turn this into a search bot filter.

17 March at 22:20 | Open on hackers.town
lavender

@faoluin this is beautiful thank you for sharing

18 March at 5:58 | Open on urusai.social
Go Up