Submitted 8 months ago by L4s@lemmy.world [bot] to technology@lemmy.world
Reddit Signs AI Content Licensing Deal Ahead of IPO::Reddit Inc. has signed a contract allowing a company to train its artificial intelligence models on the social media platform’s content, according to people familiar with the matter, as it nears the potential launch of its long-awaited initial public offering.
Long-awaited, said no one. Is AI going to fabricate even more of the bullshit on reddit then?
Is this dead Internet theory?
You want to see dead internet theory, go browse top all time on any subreddit that used to allow submissions from gfycat. It’s a wasteland.
The shitshow that is Reddit IPO is long-awaited.
Problem is Reddit content and votes aren’t all human so unless they kept a record of which parts are just chatbots and which votes were faked its not exactly useful to train on in a pure sense.
Considering the disinformation wars and botnets between the big countries its hard to even get a idea of what people really think and what is bullshit and what isn’t.
In any case I’m glad reddit has fucked themselves. This small corner of sanity is a bastion in a shit blizzard.
I’ve been on reddit, I don’t know that I would like to use a LLM trained on much of the content there (excluding tech/DIY space)
gpt3/4 are already trained on reddit data. Not reddit data exclusively, but there’s a lot of it in there.
Reddit is actually pretty decent for training llms. Funny enough an ai finetuned on 4chan does better in intelegence benchmarks.
Now I wish I could remember what the nonsense I replaced all of my content with before I deleted my account.
I used some text telling Spez he was a greedy little pigboy and to train his AI with that, if I recall correctly.
Aww man, why did I not think of doing that!?
I used text from a random syllable generator intended for constructed languages.
They say it’s $60 million on an annualized basis. I wonder who’d pay that, given that you can probably scrape it for free.
Maybe it’s the AI act in the EU. That might cause trouble in that regard. The US is seeing a lot of rent-seeker PR, too, of course. That might cause some to hedge their bets.
Maybe some people had not realized that yet, but limiting fair use does not just benefit the traditional media corporations but also the likes of Reddit, Facebook, Apple, etc. Making “robots.txt” legally binding would only benefit the tech companies.
Hmmm anyone remember when Andrew Yang was running for president and said that data was the new oil and that people should own the content they put on social media?
I deleted all my posts before closing my accounts back when they were breaking third-party apps, although I’m sure they probably kept a private log of all posts specifically for this purpose.
To be honest, I expect AI companies are scraping Lemmy and other places for training data anyway, but I’d rather Reddit specifically not make any money off my posts.
So what’s the best way to scrub your reddit comments and posts?
They keep a history of all your comments and edits. Deleting them will work for the companies that are scraping it for free though, but it also brings up the value of Reddit’s private database.
I guess I’ll go the GDPR route in this case
I’ve been using Power Delete Suite for years. It runs as a browser bookmark, so doesn’t need API, etc. I’ve got it deleting everything older than 3 months each time I run it.
I made enough Reddit comments that they could probably make a solid imitation of me
kowcop@aussie.zone 8 months ago
Seems pretty clear why the apis were shut down for apps
Copernican@lemmy.world 8 months ago
They were transparent about it. AI and gatekeeping the user generated comments was the deciding factor to close the API and that’s what they told the public.
6daemonbag@lemmy.dbzer0.com 8 months ago
IIRC that was not the case. They very publicly blamed 3rd party apps, which was both disingenuous and not transparent.
kowcop@aussie.zone 8 months ago
I can’t remember if the word at the time was that they were trying to stop the calls from affecting performance or they wanted the juicy data all for themselves
KairuByte@lemmy.dbzer0.com 8 months ago
That is absolutely not the case. They stated a lot of different reasons, ranging from “these freeloading third party developers are making money off our hard work and should be paying” to “we’ve been doing this for free and it costs us a lot of money.”
What you’re thinking of, is the fact that everyone was well aware of the truth, and the fact that they were just butt hurt about the fact that AI was being trained on the data and they didn’t get a cut.
So they did the same thing, and just fucked everyone over.
dgmib@lemmy.world 8 months ago
If it was just about monetizing scraping for AI models, they could have easily had different pricing for AI uses than they did for 3rd party apps.
If it was about the lost revenue from the lack of ads on third party apps, they only needed to give existing 3rd party apps a longer period of time to transition their business models. 3rd party app users would have been paying way more than Reddit was losing from the lack of ads.
No Reddit wanted to kill off the third party apps. They used the AI scraping as an excuse to shut them down. They wanted to force people onto their shitty app.
I don’t know what their actual reasoning for that is, but there’re basically two possibilities I can think of:
Their executive team and board of directors is ridiculously incompetent.
Their shitty 1st party app is harvesting significantly more data about you than the 3rd party apps did, and they can sell that data for more than the $2-5 per user per month they would be getting if they gave the 3rd party apps time to transition to a paid business model.
BleatingZombie@lemmy.world 8 months ago
I don’t think that’s true. If I remember correctly it was just obvious what they were trying to do. They were never transparent