• Hey Guest,

    We wanted to share a quick update with the community.

    Our public expense ledger is now live, allowing anyone to see how donations are used to support the ongoing operation of the site.

    👉 View the ledger here

    Over the past year, increased regulatory pressure in multiple regions like UK OFCOM and Australia's eSafety has led to higher operational costs, including infrastructure, security, and the need to work with more specialized service providers to keep the site online and stable.

    If you value the community and would like to help support its continued operation, donations are greatly appreciated. If you wish to donate via Bank Transfer or other options, please open a ticket.

    Donate via cryptocurrency:

    Bitcoin (BTC):
    Ethereum (ETH):
    Monero (XMR):
N

noname223

Archangel
Aug 18, 2020
7,394
A friend of mine is really pissed that everything ends up in the algorithm. I think Sanctioned Suicide posts are probably part of the training data. And there were jokes the Grokipedia article (I once posted it in a thread) was only so positive because we subtly influenced it. Because we were in the training data.

Personally, everywhere where I use AI chatbots privately and where it is possible to turn it off. I disabled the option that my data ends up in the training data. I think this is the only sane choice. Though, I think on SaSu. Everything that I post on the surface web will inevitably end up in the training data. I think probably also the posts in The Sanctuary. Personally, I really doubt the data of Sanctioned Suicide was safe from that.
 
Last edited:
  • Like
Reactions: katagiri83 and Eriktf
farukon

farukon

New Member
Aug 27, 2026
1
I don't want any of my words to train chatbots ever anywhere
 
  • Like
Reactions: rorosbones
sulkygirl

sulkygirl

Member
Nov 2, 2024
30
I really don't like the idea of an AI being influenced by SaSu. Naturally, there's a lot of pro-suicide rhetoric and an abundance of information on how to die here. All stuff I wouldn't want in the hands of any artificial intelligence. A lot of people here are old enough to make a decision but an easily manipulatable child who got their hands on a chatbot wouldn't be. And AI would likely mangle and misrely instructions on how to go which would be quite dangerous. But as you said, anything on the surface web is inevitably gonna become training data so there's really nothing that can be done.
 
B

BirdWatcher

Member
May 17, 2026
32
I'm not sure SaSu gets scraped for AI training. On the one hand AI labs are desperate to find more human made content to absorb, but on the other hand they are also paranoid of their AIs leaving some very tight boundaries of "politically correct", "corporate friendly" (or whatever you wanna call it) generations. There is a reason reddit content features heavy in AI training material, but 4chan much less so.

I guess the admins of this forum might know if they are getting weird activity from bot crawlers.