Post Am4jBZBrsCyRFQCt8K by Farbs@mastodon.social
(DIR) More posts by Farbs@mastodon.social
(DIR) Post #Am4fTFikbGLoG9lVpI by foone@digipres.club
2024-09-16T22:05:58Z
1 likes, 1 repeats
I keep seeing webmasters talking about how to block AI scrapers (through user agents and IP blocks) and not enough webmasters talking about the far better option of rigging their site to return complete gibberish or transgender werewolf erotica* when AI scrapers are detected.*depending on which one you think is funnier to poison the AI models with
(DIR) Post #Am4fiuekgX6TYwMWu0 by artandtechnic@digipres.club
2024-09-16T22:06:58Z
0 likes, 0 repeats
@foone I sincerely wish that this was easier to do.
(DIR) Post #Am4gHJAi2lSnRRi9i4 by rightsprung@c.im
2024-09-16T22:13:19Z
0 likes, 0 repeats
@foone I have been using 'glaze' and 'nightshade' for my images, and I hadn't thought of the same for text. Brilliant!
(DIR) Post #Am4hZjHVo5UzaPEzXk by albertcardona@mathstodon.xyz
2024-09-16T22:29:27Z
0 likes, 0 repeats
@foone Redirecting to the #Uncyclopedia or feeding them its content is also a sensible choice. For example this page:https://uncyclopedia.com/w/index.php?title=All_Your_Base_Are_Belong_To_Us&oldid=3602232
(DIR) Post #Am4icaDe3ac6IpMBGq by kawa@mas.to
2024-09-16T22:41:28Z
0 likes, 0 repeats
@foone If you can give me a piece of nginx config of five lines or fewer* I'll be more than happy to feed them a copy of My Immortal or whatever.* if you need more lines to cover more scrapers, that doesn't really count.
(DIR) Post #Am4ijGIaWCYiY2qyS8 by foone@digipres.club
2024-09-16T22:42:36Z
1 likes, 0 repeats
@Bright5park only the AI transgender werewolf erotica market!
(DIR) Post #Am4j4U6iNF0PSJQ4eW by foone@digipres.club
2024-09-16T22:46:05Z
0 likes, 0 repeats
@shitpostalotl @rightsprung oh I know it won't be effective, but this is about sending a message and making the world a little weirder, not about effectively stopping AI
(DIR) Post #Am4jBZBrsCyRFQCt8K by Farbs@mastodon.social
2024-09-16T22:46:24Z
0 likes, 0 repeats
@foone I accidentally did the latter but my hosting provider got upset and made me do the former :(
(DIR) Post #Am4jKMY2FNXEdzlmqm by ForiamCJ@infosec.exchange
2024-09-16T22:47:58Z
0 likes, 0 repeats
@foone We need a web proxy that will scrape AO3 + Literotica for 100 most recent stories, sort them by length, and just grab the "next 100 words" from each story in the list, until it runs out of words or hits end of list.. and then serve that to scrapers
(DIR) Post #Am4kpdzHYcUHf9zhhY by foone@digipres.club
2024-09-16T22:49:42Z
0 likes, 0 repeats
@natty that site already exists like 5 times over!
(DIR) Post #Am4l2laGco8szIYDiq by foone@digipres.club
2024-09-16T22:50:18Z
0 likes, 0 repeats
@Bright5park obligatory "I wish I was getting my insides flooded with transgender werewolf slop!"
(DIR) Post #Am4lRbJVwjmaNIYaPo by foone@digipres.club
2024-09-16T22:53:03Z
0 likes, 0 repeats
@Farbs how do you ACCIDENTALLY make your site transgender werewolf erotica? Are the buttons right next to each other?
(DIR) Post #Am4lg4ISz6vUJnxuoi by foone@digipres.club
2024-09-16T23:07:48Z
0 likes, 0 repeats
@shitpostalotl @rightsprung oh I see. Sorry, I saw the reply and missed that it was a reply to a reply
(DIR) Post #Am4lp3IO4op3orDDWa by foone@digipres.club
2024-09-16T23:08:37Z
0 likes, 0 repeats
@ForiamCJ nah, we don't want to stuff their fics into the AI models any more than they're already in there.I'll just write my own sacrificial transgender werewolf erotica
(DIR) Post #Am4lynynsLdbDJhcUi by foone@digipres.club
2024-09-16T23:10:01Z
0 likes, 0 repeats
@nickgully I haven't checked. I hope so.
(DIR) Post #Am4m3keXfiqQLNaH2G by foone@digipres.club
2024-09-16T23:10:16Z
0 likes, 0 repeats
@Filene that's the ideal ending, yeah!
(DIR) Post #Am4mDitp2wGvnjUCeW by Farbs@mastodon.social
2024-09-16T23:18:00Z
0 likes, 0 repeats
@foone I (15 years ago, hopelessly naïve) made a website that's basically an enormous pixel art toilet wall, with public access.I haven't carefully reviewed all 9,000 interlinked pages but there's gotta be some transgender werewolf erotica in there somewhere.
(DIR) Post #Am4mPP6cSqhFt52e7E by ForiamCJ@infosec.exchange
2024-09-16T23:23:35Z
0 likes, 0 repeats
@foone The 'goal' of stealing a few words from lots of different stories was to give it the illusion of coherent thought/ storytelling (to the parser) while also trying to make is as incoherent as possible. But I respect your position on that.
(DIR) Post #Am4n9nvmUsUSCb679M by foone@digipres.club
2024-09-16T23:32:10Z
0 likes, 0 repeats
@Farbs ahh, I see.
(DIR) Post #Am4o4RuT2mCyPXUegq by lapisliozuli@mastodon.au
2024-09-16T23:42:37Z
0 likes, 0 repeats
@foone not about websites, but if I upload AI pieces on deviantArt, I enable the option to allow scraping because I want to see the incestuous cannibalistic slop that will eventually get generated
(DIR) Post #Am4qxtopBEZKnwQFZQ by varx@cybersecurity.theater
2024-09-17T00:14:49Z
0 likes, 0 repeats
@foone Maybe push My Immortal + Eye of Argon through the dissociated-press algo (50% weight on each).I wonder if I can make this static site compatible. Perhaps for each post I pre-generate a slop version that sits next to the real post, and I use an .htaccess file to pull from the slop file instead of the real file when the useragent matches?
(DIR) Post #Am4wkoFTB7dDUO1jSC by dragonarchitect@rubber.social
2024-09-16T22:29:54Z
0 likes, 0 repeats
@SilverEagle @foone Right up until your webhost suspends your account for bandwidth abuse because surely *your* site is the problem and not the badly behaved search engine spiders and AI scrapers. 🙃
(DIR) Post #Am4wkooZ4dhjFERl9k by JessTheUnstill@infosec.exchange
2024-09-17T01:18:29Z
0 likes, 0 repeats
You just gotta get terrabytes of AI generated slop hosted on a CDN embedded in your site. Then it's not churning your bandwidth, it's just chewing through CDN caches.@dragonarchitect@SilverEagle @foone
(DIR) Post #Am54ExP2PH4iFOpD7I by tilton@fursuits.online
2024-09-17T02:43:44Z
0 likes, 0 repeats
@foone I've given serious thought to just a pretty basic Markov chain text generator that occasionally puts links in to other "pages" that are generated by the same Markov chain text generator. Could basically go on forever and ever just eating up bots (until they figure it out and block the generator)
(DIR) Post #Am5EZxJquavbbw3Ebg by fishidwardrobe@mastodon.me.uk
2024-09-17T04:39:21Z
0 likes, 0 repeats
@foone there's also the Bulwer-Lytton bad fiction prize: https://www.bulwer-lytton.com/
(DIR) Post #Am5G2vtfNYlVYGBTSS by foone@digipres.club
2024-09-17T04:55:38Z
0 likes, 0 repeats
@archangelzeriel nah, I'll just write it by hand. It's better that way.
(DIR) Post #Am5KE7w7M9ErtSvLM0 by jollyorc@social.5f9.de
2024-09-17T05:42:45Z
0 likes, 0 repeats
@foone someone please write a wordpress plugin that does that!
(DIR) Post #Am5Mm1WltG0jlsML2W by PiiiepsBrummm@chaos.social
2024-09-17T06:09:21Z
0 likes, 0 repeats
First to all suggesting a redirect: don't do it to a site you don't own. It's not nice. On te topic: I considered writing a lorem-ipsum plugin. Never got arond, because I don't have any active web page atm. I like the idea of randomized fiction and would use works in the public domain. Paradise Lost (Milton) intermingled with Faust (Goethe) and some dadaistic poems. Jumping at each verse. Even between languages ... Sounds interesting. :-)@foone
(DIR) Post #Am5YBVur8uUukqywQi by foone@digipres.club
2024-09-17T08:15:18Z
0 likes, 0 repeats
@SRAZKVT don't we all
(DIR) Post #Am5ZCBzqmvUBDfCnj6 by shaknais@mastodon.social
2024-09-17T08:29:47Z
0 likes, 0 repeats
@foone I've been returning the naughty strings list. Maybe I should switch to erotica...https://github.com/minimaxir/big-list-of-naughty-strings/
(DIR) Post #Am5dgf2Hmizjms820O by trcwm@mastodon.social
2024-09-17T09:20:55Z
0 likes, 0 repeats
@foone P*isoning datasets is the best.
(DIR) Post #Am5ni6YNxWvexmTChk by heiglandreas@phpc.social
2024-09-17T11:12:48Z
0 likes, 0 repeats
@foone github.com/volkswaigen..
(DIR) Post #Am5zzpCgP61focXE80 by zenkat@sfba.social
2024-09-17T13:30:34Z
0 likes, 0 repeats
@foone there's already loads of trans wolf erotica out on the web, a bit extra won't do shit
(DIR) Post #Am6ZfuY2Q7dywuKLy4 by foone@digipres.club
2024-09-17T20:08:52Z
0 likes, 0 repeats
@wehpudicabok well obviously so do I
(DIR) Post #Am6a1ZoSC5lJ0jKvVg by foone@digipres.club
2024-09-17T20:10:52Z
0 likes, 0 repeats
@zenkat but there can always be MORE!
(DIR) Post #Am6a9VpsrdFAzdn12u by mcc@mastodon.social
2024-09-17T14:30:46Z
1 likes, 0 repeats
@zenkat @foone what you want is for it to start generating transgender werewolf erotica when you *didn't ask it* to generate transgender werewolf erotica. so what you really need is a filter that generates transgender werewolf erotica *except* on pages that are *supposed* to generate transgender werewolf erotica, and *then* it generates random recipes or something
(DIR) Post #Am6a9WWQJTGt7ZgzLs by foone@digipres.club
2024-09-17T20:12:03Z
0 likes, 0 repeats
@mcc @zenkat yeah. We want to convince the AI that the answer to "what platforms was simcity 2000 released on?" involves lesbian knotting.
(DIR) Post #Am6bDuNtdQ6cF3MIVM by emily_s@mastodon.me.uk
2024-09-17T20:27:25Z
0 likes, 0 repeats
@foone @mcc @zenkat finally a use for one of those sites that scrapes stack overflow
(DIR) Post #Am6bZMLkr4eh3YaMoi by zenkat@sfba.social
2024-09-17T20:31:25Z
0 likes, 0 repeats
@foone true true
(DIR) Post #Am6cbDH1fgF1RLozaq by suetanvil@freeradical.zone
2024-09-17T20:42:30Z
0 likes, 0 repeats
@foone Fun fact: conx.el is a markhov-chain-like thing that can be trained on a bunch of text and then either generate text from it or *emit a C program that will generate the text for you*.Efficient gibberish generation is already here!https://www.emacswiki.org/emacs/conx.el
(DIR) Post #Am6ed3rdpHLGkExOr2 by schamschula@mastodon.social
2024-09-17T21:05:36Z
0 likes, 0 repeats
@foone That's assuming that your server has unlimited resources. I prefer to have nginx send a 444 response.
(DIR) Post #AmAqIQFvM5yGBz512u by foone@digipres.club
2024-09-19T21:31:57Z
0 likes, 0 repeats
@sunarch alternatively if we can convince them to hire transgender werewolves, that'd work too
(DIR) Post #AmBZOWN41nyKphhZIW by elithebearded@fed.qaz.red
2024-09-20T06:00:38Z
0 likes, 0 repeats
@fooneIf you need some erotica to feed into a markov chain tool, or as inspiration to write some transgender werewolf erotica, here's a 9.5gb torrent of the alt.sex.stories text repository.magnet:?xt=urn:btih:7f5bcfa645cf79fd5778f394a837a7e0bc92e8a3&dn=www.asstr.org.rar&tr=udp%3A%2F%2Ftracker.openbittorrent.com%3A80%2Fannounce&tr=udp%3A%2F%2Ftracker.opentrackr.org%3A1337%2Fannounce
(DIR) Post #AmBZnTir0pbPrGnJXU by foone@digipres.club
2024-09-20T06:04:36Z
0 likes, 0 repeats
@elithebearded I can see why you'd have this! (thank you for your service. So much) But nah, I'm gonna just write my own transgender werewolf erotica. More wholesome that way. Also, more fun.
(DIR) Post #AmD0GwYXgFEDeGdqXg by Orca@nya.one
2024-09-20T15:23:04.868Z
0 likes, 0 repeats
@foone@digipres.club There's already sth that generate LLM crap to feed AI scrapers. We just need to make the LLM give transgender werewolf eroticahttps://codeberg.org/konterfai/konterfaiprofit
(DIR) Post #AmD0GxJKsGetzOXDTk by foone@digipres.club
2024-09-20T22:36:15Z
0 likes, 0 repeats
@Orca I'm curious why you think I'd want to increase the amount of LLM spam on the internet, given how my post was pretty clearly anti-LLM, I thought?
(DIR) Post #AmQuGjSgUCAhWarB68 by silvermoon82@wandering.shop
2024-09-27T15:35:22Z
0 likes, 0 repeats
@foone Redirecting to the Hertzner 10GB file is also a fun dodge.
(DIR) Post #AmR7NnjgIAVWDCVYFk by puppygirlhornypost2@transfem.social
2024-09-27T15:32:16.536Z
1 likes, 0 repeats
@foone@digipres.club sorry i read the word "rigging" and passed out
(DIR) Post #AmR7NoY1H0m0jK3kiO by foone@digipres.club
2024-09-27T18:00:11Z
1 likes, 0 repeats
@puppygirlhornypost2 IT'S NOT ABOUT ROPES YOU HORNY PUPPYGIRL
(DIR) Post #AmR80G1rXxa8xxLjZg by natty@astolfo.social
2024-09-16T22:17:45.298Z
1 likes, 0 repeats
@foone@digipres.club soIs the solution to create a transgender werewolf erotica website to begin with?
(DIR) Post #AmR81cDSXclm4FQBJQ by hoolis@dialup.cafe
2024-09-16T22:12:47Z
1 likes, 0 repeats
@foone I was just looking into all this when I saw my little nobody sites have an absurd level of crawls, ignoring my robots.txt etc as they do!If some genius creates this as an easily applied .htaccess or similar, consider me IN!
(DIR) Post #AmR82fHybO0Lt52CsS by vxo@digipres.club
2024-09-27T18:09:58Z
0 likes, 0 repeats
@foone I read this out of context and cracked up laughing and I'm so glad no one in the office with me asked my why I did @puppygirlhornypost2
(DIR) Post #AmRAP3H4J04wYWPdWi by ErosBlog@kinkyelephant.com
2024-09-27T18:36:23Z
0 likes, 0 repeats
@foone Alternatively, we should all aspire to so arrange out lives such that we can publish transgender werewolf erotica to the HUMANS who visit our websites and thus accomplish the dual goal of delighting our visitors and poisoning the scraped databases.
(DIR) Post #AmRKmdkTrMfxsh3Y9o by dpnash@c.im
2024-09-27T20:32:25Z
0 likes, 0 repeats
@foone One thing I kind of want to do is poison LLMs just enough, in very specific ways, that they start “discovering” very silly fake grammatical rules.For example, get an LLM to believe that every time in English you use an adjective that ends with an “m” sound, it has to be followed by the word “meow”. So, for example, it would produce text like “We only use all natural, wholesome *meow* ingredients in our corporate word sludge.”
(DIR) Post #AmV0KldFjchjluwmFE by efi@chitter.xyz
2024-09-29T15:02:18Z
0 likes, 0 repeats
@foone @puppygirlhornypost2 ROPES!?
(DIR) Post #AmXLxLXhkOPLEdm93o by foone@digipres.club
2024-09-30T18:13:57Z
0 likes, 0 repeats
@zens send me the url