Reddit said in a filing to the Securities and Exchange Commission that its users’ posts are “a valuable source of conversation data and knowledge” that has been and will continue to be an important mechanism for training AI and large language models. The filing also states that the company believes “we are in the early stages of monetizing our user base,” and proceeds to say that it will continue to sell users’ content to companies that want to train LLMs and that it will also begin “increased use of artificial intelligence in our advertising solutions.”
The long-awaited S-1 filing reveals much of what Reddit users knew and feared: That many of the changes the company has made over the last year in the leadup to an IPO are focused on exerting control over the site, sanitizing parts of the platform, and monetizing user data.
Posting here because of the privacy implications of all this, but I wonder if at some point there should be an “Enshittification” community :-)
I wonder what the risks are to including deleted and pre-edited content in training data. Most of the edits are going to be typos and formatting, do you want 2-3 copies of the same message with typos in them for training data? Similarly, deleted comments are mostly nonsense, unhelpful, duplicate, or highly controversial things.
If someone wants to dig through and find individual users to restore that’s one thing, but I don’t think I’d immediately choose to train off of that other data unless I had to.
It should be very easy to distinguish edits and deletes which were made within a few minutes or hours after writing a comment, from those made months or years later right around the reddit blackout.