#iocaine — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #iocaine, aggregated by home.social.
-
Oh, looks like Anthropic (Claude) gave up again (they did before for a short time) on crawling my sites after ~29,728,157 requests. Nice :ablobcatcoffee:
Looks like they enjoyed my poisonous maze a lot. Sad to see them go. They could eat a lot more poisoned garbage. I'm ready.
```
nam_shub_of_enki_request_distribution{group="Claude",outcome="garbage"} 29728157
``` -
hehe, just looked at iocaine's Prometheus metrics :blobcatgiggle:
```
nam_shub_of_enki_ruleset_hits{outcome="garbage",ruleset="bad-browser"} 1337
``` -
I just finished a short introduction post about @algernon's project iocaine: the deadliest poison known to AI :blobcatsunglasses: (not man :blobcatgiggle:)
Within the post I explain what iocaine is, how it's related to AI and LLMs and of course why I use it to fight AI/LLM companies and to poison their crawlers and training sets.
Using, configuring and watching iocaine was also a way for me to shifting my dystopic/pessimistic thoughts :blobcat_thisisfine: into fun and joy :ablobcatattention:
And yes, I hate AI and LLMs and yes, I'm really fine by becoming an obsolete developer by not using it :ablobcatheart:
https://lukasrotermund.de/posts/fighting-ai-and-llms-with-iocaine/
-
:ablobcatheartsqueeze: I have been running iocaine on my server for a week now. During this time, 7,076,701 requests have passed through iocaine, 3,312,318 of which were identified as AI crawlers/bots. 3,741,577 requests came from crawlers/bots that got stuck in iocaine's deadly maze, consuming an infinite amount of poisoned garbage. Furthermore, 972 crawlers/bots were detected that were routed into the maze via major browsers.
All of this is managed by iocaine with just ~80 MB of memory and ~0.1% direct CPU usage. Now that’s what I call efficient! Well done, @algernon.
Let's fight back against AI crawlers and bots. Thanks to projects like iocaine, this is entirely possible, not just theory :blobcat_thisisfine:
#iocaine #ai #llm #FckAI #FckLLMs #selfhosting #crawler #bots
-
@ramsey sorry to hear that! You could use a self-hosted, iocaine-protected forgejo with git send email patches. That's what I have thought about for the last couple of weeks. If centralised platforms aren't helpful here, I'll have to help myself.
-
It's crazy how drastically iocaine's default request handler, QMK, reduced the load on my forgejo container. It's now around 10 per minute instead of over 1,000 per minute! I hope all of these AI bots and crawlers get lost in iocaine's poisonous, infinite maze of garbage :blobcatgiggle: :ablobcatwave:
evil hehe
-
I have just installed iocaine 3.2.0 by @algernon and have already started successfully serving poisoned garbage to the AI agents. I love it! I especially like how simple the setup was, and how easy it was to expand my existing Caddyfile. My monthly donation is set up too. What a great project!
-
I came across iocaine from @algernon a few minutes ago because of the DDoS attack on Codeberg. Anything that stops, annoys, or destroys AI has my support. I also had to deal with annoying, aggressive crawlers recently and I banned ~150,000 IPs.
I'll set it up in Caddy tonight. Let's see how it performs. If it works, I'll definitely make a donation.
-
And I do have a blog beside the house construction one too. I do post there, seldom, but today I got something techie to write about even.
Got one more server of mine protected by iocaine, good by AI crap bots.
-
Crazy
#Vivaldi on #Android even supports #Javascript #bookmarkletsGuess its ~ByeBye~ Brave
Maybe #iocaine wil, leave me alone now too
#repost •acws #acws -
Eat this ChatGPT! :blobcatgooglytrash:
Iocaine is doing good work serving Markov-chain generated garbage to ChatGPT. :blobcathappy:
-
It's been wild watching the crazy array of obvious AI bots working their way deeper into my iocaine tarpit.
They are taking the greatest care to not burden my system (rate is holding at about 5 pages per minute, on average, for the last 30 days straight). The site they are chewing through is statically generated, so it can handle a lot more.
It's just such a weird feeling watching them, like when a toddler thinks you can't see them because *their* head is covered.
-
Iocaine: The deadliest poison known to AI
https://git.madhouse-project.org/iocaine/iocaine
#HackerNews #Iocaine #AI #poison #technology #deadliest #innovation #cybersecurity
-
#GPTBot macht nach wie vor ca. 20% der Zugriffe dieser Mastodon-Instanz aus, aber der Crawler bekommt nur noch von #Iocaine generierten Unsinn. Das reduziert die Datenmenge, die wir an ihn ausliefern, drastisch und zerstört die Qualität unseres Datensatzes für ihn vollkommen.
Es hilft uns also Kosten zu sparen, verschlechtert die LLM und macht auch noch diebische Freude! Win-Win-Win! :KritischerTreffer:
-
This is -ing unbelievable:
In the 17 hours running my "Discworld Ólyfjan" Iocaine, GPTBot has download the same 84 pages over 10000 times. They don't even change!And Google has it on the search index: "Ólyfjan" [name of any discworld character]
has results.HEX, the Bursar, even the troll Brick would be more intelligent than that...
-
One of the things that annoys me the most is that the scraper that went furthest into the tarpit (83 links deep) is also the one who comes back reading the same pages again and again:
{host="olyfjan.blomi.is",user_agent="Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; GPTBot/1.2; +https://openai.com/gptbot)",user_agent_group="GPTBot"} has sent 6991 GET requests, for the same 84 pages, downloading 22779416 bytes.
-
It looks like several scrapers have found my Discworld tar pit. 😈
Today's stats so far:
GPTBot/1.2: 12405406 Bytes
Googlebot/2.1: 1391937 Bytes
ClaudeBot/1.0: 6359 Bytes
Amazonbot/0.1: 4622 Bytes
AhrefsBot/7.0: 1414 Bytes -
TIL about an AI poisoning tarpit called "iocaine" by @algernon which generates loads of garbage sites from just two files of sentences and words.
I have all of Sir Terry Pratchett's books as epub, so the path was clear... -
es wird der Tag kommen an dem Manipulation von AI Suchmaschinen illegal wird, bzw. Deutsch als Störer verfolgt wird.
Es ist mitunter ja eine moralische Frage, wenn ich weiss das die KI/AI Crawler ungefragt und unerwünscht sich alles zusammen klaubern : darf ich deren content dann vergiften? Und wenn die LLM auch irgendwo was gutes hat, handel ich dann unmoralisch?
poisoning crawler :
https://git.madhouse-project.org/algernon/iocaine -
For those who run webservers and are a bit sick of LLM crawlers, you may find iocaine https://git.madhouse-project.org/algernon/iocaine/ and nepenthes https://zadzmo.org/code/nepenthes/ interesting.
#LLM #iocaine #web #AI #nepenthes #tarpit #WebServer #nginx #caddy
