reddit is about to become 100% AI generated content unless you pay for it. imagine actually paying for access to listen to real redditors.
One of my guilty pleasure is to read those relationship threads. Most are so stupid it ridiculous how can you not laugh?
@xuya to be fair to reddit's ownership, the site is unusable in it's current state, outside of just a newsfeed to see what's popular. everything is controlled and fake, so much of what's on there now is AI generated, and the onboarding experience is a nightmare (since your account is unusable on all the major subreddits until it accumulates enough age + karma). i tried making a new account after all mine got banned, and it's useless.
my assumption is that this is being done to stop AI bots, since reddit is getting progressively more useless as an AI training set because of how much it's been polluted with AI output.
my assumption is that this is being done to stop AI bots, since reddit is getting progressively more useless as an AI training set because of how much it's been polluted with AI output.
@xuya all the relationship subs are ground zero for AI content. reading the top posts on them are like a constant stream of ragebait specifically tailored at reddit's audience. like yes, every day there are 10+ people posting perfectly enraging stories to reddit, mmmhrmm. yes, absolutely genuine.
@HWABAG not for much longer. i assume the accidental banning of all their nsfw subs last week was a trial run for once they go pay for play. credit card processors are some of the most prudish organizations around.
apparently at some point during the day or nights they have to take the sick out of their mouth, and that’s when the choker comes in handy
>"I'm trying to be alternative and unique by being a carbon clonal copy of everyone else doing the same stupid shit that I am, Mom, God!"
They didn't have a father figure around to teach them how to tie a rope or else they wouldn't be gay homosexual faggots in the first place.
@DEERBLOOD @xuya @Owl @grey it’s called training. they’re getting ready to wear the big one
@Owl @xuya @DEERBLOOD @grey yeah and then we can remind them of their true name and cast them into hell
@DEERBLOOD @Owl @xuya @grey AUTUMN YOUR NAME IS MICHAEL STOP IT YOUR PARENTS MISS YOU YOU DDONT NEED TO LIVE IN THE ALPACA COMMUNE LARPING AS A FARMER
@xuya let's see what's up there right now:
> my husband made porn of me
> i started having sex with my husband at 16
> my husband cancelled our honeymoon
> my lesbian wife's brother is a sex offender
> my gf was a slut before getting together
> my ex-wife is acting like a cunt (okay, this one's believable)
and it's like that every day lol. i've seen posts there where they leave the AI boilerplate in the post by accident. this happens because of the karma minimums that are enforced all around the site, if you're going to sell an account it needs to have enough karma that the buyers can use it, so posting engineered AI slop is an easy way to prepare it for sale.
> my husband made porn of me
> i started having sex with my husband at 16
> my husband cancelled our honeymoon
> my lesbian wife's brother is a sex offender
> my gf was a slut before getting together
> my ex-wife is acting like a cunt (okay, this one's believable)
and it's like that every day lol. i've seen posts there where they leave the AI boilerplate in the post by accident. this happens because of the karma minimums that are enforced all around the site, if you're going to sell an account it needs to have enough karma that the buyers can use it, so posting engineered AI slop is an easy way to prepare it for sale.
True. A good father will prepare his trans children for their future. But they need to hurry. They won't have much time.
@grey
>People might, big "might", finally leave Reddit wholesale and start making forums again
Who am I kidding? Normalfags and the sophomoric techbros will keep using shitty programs and sites like Dicksore. The internet is dead.
By the way, this is why Reddit devs have been doing their damnedest to preventing FOSS third-party frontends from working.
>People might, big "might", finally leave Reddit wholesale and start making forums again
Who am I kidding? Normalfags and the sophomoric techbros will keep using shitty programs and sites like Dicksore. The internet is dead.
By the way, this is why Reddit devs have been doing their damnedest to preventing FOSS third-party frontends from working.
@grey With the CIA no longer funding Reddit, they don't have a sustainable business model. This is excellent news.
> imagine actually paying for access to listen to real redditors.
Regular people won't. Even advertisers won't, and they shit buckets of money at anything that has a non-negative ROI: https://www.techspot.com/news/78713-reddit-users-almost-worthless-compared-other-social-networks.html .
reddit-usaid.png
reddit_groomer.png
reddit_is_fucking_terrible_at_programming.png
> imagine actually paying for access to listen to real redditors.
Regular people won't. Even advertisers won't, and they shit buckets of money at anything that has a non-negative ROI: https://www.techspot.com/news/78713-reddit-users-almost-worthless-compared-other-social-networks.html .
reddit-usaid.png
reddit_groomer.png
reddit_is_fucking_terrible_at_programming.png



@waltercool @grey If Lemmy ever gets its shit together enough to start federating properly, it might end up useful.
Already happened.
One of my mutuals ran a based instance called wolfballs. The instance was one of 2 on Lemmy worth talking to, and to this day despite having been down for years most Lemmy instances block both by default.
One of my mutuals ran a based instance called wolfballs. The instance was one of 2 on Lemmy worth talking to, and to this day despite having been down for years most Lemmy instances block both by default.
- replies
- 1
- announces
- 0
- likes
- 2
@sj_zero @waltercool @grey
> oh the mastodongs are walling themselves off
...Good?
It's not like the people that prefer hugboxes are fun to talk to or have anything interesting to say.
> oh the mastodongs are walling themselves off
...Good?
It's not like the people that prefer hugboxes are fun to talk to or have anything interesting to say.
You're not wrong.
Sometimes I head over and it's like "oh yeah, these aren't actually people, they're just preprogrammed dialog trees"
It's like talking to a less advanced version of eliza
Sometimes I head over and it's like "oh yeah, these aren't actually people, they're just preprogrammed dialog trees"
It's like talking to a less advanced version of eliza
@sj_zero @waltercool @grey
> It's like talking to a less advanced version of eliza
Well, I'm gonna quote the same thing I quoted in the
https://blog.freespeechextremist.com/blog/on-human-bots.html post:
> [Richard] Wallace's theory of A.I. is no theory at all. It's not that he doesn't believe in artificial intelligence, per se; rather, he doesn't much believe in intelligence, period. In a way that oddly befits a contest sponsored by a bunch of Skinnerians, Wallace's ALICE program is based strictly on a stimulus-response model. You type something in, if the program recognizes what you typed, it picks a clever, appropriate, "canned" answer.
> There is no representation of knowledge, no common-sense reasoning, no inference engine to mimic human thought. Just a very long list of canned answers, from which it picks the best option. Basically, it's Eliza on steroids.
> Conversations with ALICE are "stateless"; that is, the program doesn't remember what you say from one conversational exchange to the next. Basically it's not listening to a word you say, it's not learning a thing about you, and it has no idea what any of its own utterances mean. It's merely a machine designed to formulate answers that will keep you talking. And this strategy works, Wallace says, because that's what people are: mindless robots who don't listen to each other but merely regurgitate canned answers.
I think talking to them isn't just missing out on some boredom; I think it actually hurts your ability to think. You ever have a discussion with someone and it was clear they were not listening and were picking their responses out of a list of canned rebuttals? I mean, learned helplessness results from a feature of the brain: information that is not relevant is discarded, and information that is not actionable is not relevant. So you talk to those people long enough and all the thoughts that might occur to you during a normal conversation get quieter, the information that is relevant during a stupid internet argument gets louder: I think "stupid internet argument" is a symptom of a (contagious) cognitive disease.
> It's like talking to a less advanced version of eliza
Well, I'm gonna quote the same thing I quoted in the
https://blog.freespeechextremist.com/blog/on-human-bots.html post:
> [Richard] Wallace's theory of A.I. is no theory at all. It's not that he doesn't believe in artificial intelligence, per se; rather, he doesn't much believe in intelligence, period. In a way that oddly befits a contest sponsored by a bunch of Skinnerians, Wallace's ALICE program is based strictly on a stimulus-response model. You type something in, if the program recognizes what you typed, it picks a clever, appropriate, "canned" answer.
> There is no representation of knowledge, no common-sense reasoning, no inference engine to mimic human thought. Just a very long list of canned answers, from which it picks the best option. Basically, it's Eliza on steroids.
> Conversations with ALICE are "stateless"; that is, the program doesn't remember what you say from one conversational exchange to the next. Basically it's not listening to a word you say, it's not learning a thing about you, and it has no idea what any of its own utterances mean. It's merely a machine designed to formulate answers that will keep you talking. And this strategy works, Wallace says, because that's what people are: mindless robots who don't listen to each other but merely regurgitate canned answers.
I think talking to them isn't just missing out on some boredom; I think it actually hurts your ability to think. You ever have a discussion with someone and it was clear they were not listening and were picking their responses out of a list of canned rebuttals? I mean, learned helplessness results from a feature of the brain: information that is not relevant is discarded, and information that is not actionable is not relevant. So you talk to those people long enough and all the thoughts that might occur to you during a normal conversation get quieter, the information that is relevant during a stupid internet argument gets louder: I think "stupid internet argument" is a symptom of a (contagious) cognitive disease.
@p >read the news article
@grey That is hilarious *but* consider that they're comparing it to Facebook; it might be technically correct to say that the average intelligence of any place that people post is higher than on Facebook.