If they have any kind of archive of past mod decisions then they can just dump all that into a neural net. And then they get to look all sexy in their upcoming IPO because they are using ⭐️⭐️⭐️ARTIFICIAL INTELLIGENCE⭐️⭐️⭐️ like all the ⭐️⭐️⭐️SEXIEST⭐️⭐️⭐️ companies!!!1! No more of those annoying unpaid volunteers to get uppity any more!
I, for one, do not welcome our new AI moderation overlords, and will probably be done with Reddit if this happens. But I just know someone there has to be pushing for this.
It would take a considerable while to train an in-house LLM AI for moderation purposes, and even if it was trained honestly it would at least be consistent whereas you can get away with breaking mod rules as long as your meme made them chuckle, etc.
Who cares, AI will become a part of Lemmy too and I’m just done with reddit anyways.
Oh I’m not saying these putative AI mods would be any better than existing Reddit mods. They’d probably be even more arbitrary and capricious and unappealable, and I sure don’t expect them to be anywhere near fully-trained. But they’d be owned by Reddit, who would now be ⭐️⭐️⭐️AN AI COMPANY⭐️⭐️⭐️, which is the replacement for “the blockchain” as a thing you vaguely mention your company using if you want rich, dumb investors to wet their pants and throw tons of money at you.
even assuming reddit can ship an “AI moderation bot” that kinda works in the upcoming week (which will be the feat of the year), quality will start going down long before it can moderate at a good enough capacity to work well.
My bet? AI.
If they have any kind of archive of past mod decisions then they can just dump all that into a neural net. And then they get to look all sexy in their upcoming IPO because they are using ⭐️⭐️⭐️ARTIFICIAL INTELLIGENCE⭐️⭐️⭐️ like all the ⭐️⭐️⭐️SEXIEST⭐️⭐️⭐️ companies!!!1! No more of those annoying unpaid volunteers to get uppity any more!
I, for one, do not welcome our new AI moderation overlords, and will probably be done with Reddit if this happens. But I just know someone there has to be pushing for this.
It would take a considerable while to train an in-house LLM AI for moderation purposes, and even if it was trained honestly it would at least be consistent whereas you can get away with breaking mod rules as long as your meme made them chuckle, etc.
Who cares, AI will become a part of Lemmy too and I’m just done with reddit anyways.
Oh I’m not saying these putative AI mods would be any better than existing Reddit mods. They’d probably be even more arbitrary and capricious and unappealable, and I sure don’t expect them to be anywhere near fully-trained. But they’d be owned by Reddit, who would now be ⭐️⭐️⭐️AN AI COMPANY⭐️⭐️⭐️, which is the replacement for “the blockchain” as a thing you vaguely mention your company using if you want rich, dumb investors to wet their pants and throw tons of money at you.
The word you’re looking for is “buzzword”.
That’s the neat part, it doesn’t have to be good.
even assuming reddit can ship an “AI moderation bot” that kinda works in the upcoming week (which will be the feat of the year), quality will start going down long before it can moderate at a good enough capacity to work well.