Additionally additionally, as it says in the article, reddit sells its content to OpenAI, and in order to do that, they need it to be organic. If you feed AI output back info AI, it makes the output worse, not better. So, there’s another financial incentive to ensure that the people talking on reddit are actually human.
Comment on Reddit will tighten verification to keep out human-like AI bots
doodledup@lemmy.world 1 week agoI know you enjoy hating on Reddit but what you’re saying doesn’t make any sense. Reddit is an advertising business. Bots don’t view advertising but adds infrastructure cost. Additionally, humans are way better at engagement with the platform and people than bots.
zarkanian@sh.itjust.works 1 week ago
glimse@lemmy.world 1 week ago
If openai kept buying reddit data after 2022, they are idiots. They let the cat out of the bag.
Ledericas@lemm.ee 6 days ago
they are selling it to both OPENAI and google for thier AI too.
atrielienz@lemmy.world 1 week ago
I would like to remind you of two things. The first is that reddit user to have a mod tool called “BotDefense”. It’s shutdown in July of 2023 directly lead to a major uptick in Spam Bots.
The second is that part of the ad revenue is “impressions”. Impressions are just an account (bot or human) “viewing” the ad and they do not require a click-through. The platform hosting the ad still gets paid for those.
FreedomAdvocate@lemmy.net.au 6 days ago
If you feed AI output back info AI, it makes the output worse, not better.
Yeh but feeding reddit user output into AI is part of the reason why AI is so confidently incorrect so often.
glimse@lemmy.world 1 week ago
Bot activity brings “engagement” which generated more page views from actual humans. Literally the only downside of this announcement is the optics of having bots in the first place.
The bots were already indistinguishable from humans on reddit, do you really think that this recent scrutiny is going to lead to fewer bots? Or might it actually lead to better bots?