Reddit's Sale of User Data for AI Training Draws FTC Investigation

It already has been used by LLMs

A lot of users, including myself - though definitely still a small minority - quit Reddit over the increasingly aggressive monetisation attempts; some of us used automated history editors to prevent our posts from being scraped or searched (though it almost certainly will not remove them from back end databases).

We didn't know that this was the specific way our content was going to be used, but we knew it was probably going to be something we wouldn't like.

We have now been proved right, and if you decided, back then, to stay and keep contributing to the multi-billion-dollar site for free, now you know why the Cassandras were raising such a fuss.

Put another way, it’s fine with me if someone wants to use the content that I provided to drive value on a social website.

It’s not OK with me that someone uses the content that I provided to replicate the author of what was written (without permission).

One thing to really understand about artificial intelligence is that it is in someways, essentially a reverse engineering of the person providing the content. I did not authorize a digital twin to be created from the content that I provided. I did not authorize a digital copy of myself, but that is what all of these companies are doing.

A fundamental stripping of the authorship to the point where you can replicate a proxy of the author and not just use the derivatives of that author.

This is a fundamental difference, because no other technology has ever been able to go back up the food chain and replicate the author in a way that allows for not only replication of the content and its relative functions, but a complete photocopying of the ideas and thinking and other mechanisms and patterns of the person that’s creating it.

It’s a profound difference of people have a hard time seeing.n

I gave permission to use the content I provided. I did not give permission for someone to act as me or make a proxy of me based on what I have provided.

I consented for my media to be used, but not me and that’s a huge difference.

Wouldn't all of the populate models and applications (chatgpt, copiolot, etc) have to be burned to the ground and start clean? My understanding was that pretty much every one of these were trained on "stolen" data, i.e. without prior agreement with the authors / content creators.

This seems obviously backwards and a boon to enormous corporations.

If you want to build a search engine, you have to index everything possible, or it won't be able to find that and then it's useless. Transacting with each individual person in the world would only be possible for megacorps -- it's already expensive enough to index everything if you don't have to do that. It's also perverse that someone could have a right to prevent someone else from providing true information about them. If you're standing there making a false claim, you shouldn't have a right to prevent someone else from proving you wrong just because the evidence is from your own past.

But search engines have been ML since even before LLMs, and LLMs are fundamentally the same. How can someone have a default right to deny the public access to facts?

Now, maybe there are some things you want to keep private, and then if you share them with someone you want to bind them to not sharing that information with others, like an NDA. But that's opt-out, not opt-in, and you can't really do that for things that are public.