1020
Reddit's licensing deal means Google's AI can soon be trained on the best humanity has to offer — completely unhinged posts
(www.businessinsider.com)
this post was submitted on 22 Feb 2024
1020 points (98.7% liked)
Technology
59724 readers
3156 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related content.
- Be excellent to each another!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, to ask if your bot can be added please contact us.
- Check for duplicates before posting, duplicates may be removed
Approved Bots
founded 2 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
As I mentioned before, I use scripts to replace my comments with random excerpts from text in the public domain. I do this multiple times before finally deleting them. The result is that it becomes very difficult for the AI or anyone to figure out what is a legitimate comment and what is a line from Lady Chatterley's Lover or a scientific paper of the ecological impact from the Japanese whaling industry. It's easier to just filter out my username from their data sets.
They have almost definitely archived data and around the time of the API bullshit, made sure they didn't delete those archives. They have that content if they want to use it.
I've done the "switch, switch, switch, delete" at least twice a year for most of the twelve years I was there. The idea was to pollute the data, not delete it. Even if you started during the API bullshit, you still would have had plenty of time to corrupt your data enough. Remember, the idea is to make it so that it is difficult to tell what is a legitimate comment and what are excerpts from random text.
Most people don't reread their past comments and edit them. They could simply ignore any edits after the average time a person would notice a typo or something needing clarification, say anywhere between 5 minutes and 24 hours, or just ignore all edits. So your effort is wasted and you're still training the AI.