This is actually fascinating, if you look at the other posts of the accounts posting "I'm sorry, OpenAI won't let me do that" it's a whole ecosystem of chatbots just talking to each other about random stuff. No humans involved, just endless sludge
Comments
Log in with your Bluesky account to leave a comment
This is what those random "likes" on months old posts or a random follow are for too. It's simulating an account behavior so when they switch on state propaganda amplification in the coming election cycle it'll look more like a legit human account. Less sure about the random "cute asian girl" tactic
The vast majority of “hot girl” accounts liking my posts all have reeeeeeally sketchy links in their bios to something like “hot local girls xxx” (and many of them seem to have the same exact links.)
Pretty sure all of those are hoping dupes will check their profiles and click those links.
That one’s easy; more likely the marks will let their dicks veto their brains when whatever scam they’re running kicks off. It’s why every obvious scam Facebook and Instagram account uses the same type of girl in their profile pics
The sheer volume of it, with paid blue accounts, more suggested state actor to me. Maybe it's camouflage to appear as a non-state actor scam before running influence campaigns
This prompted me to go check on two bot accounts I flagged back in 2022. Back then there were a number of AI userpic accounts retweeting anything with a music term. Now they seem to retweet motivational memes and every so often tweet this stuff, which seems copied rather than ChatGPT text
accounts are often bought and sold so bot accounts that get a following or viral tweet are more likely to be taken over by a real person and that might be whats going on here
or it just went to a different spam generator than chatgpt
Nah they're still bots. They're all in-line with each other (back in 2022 it was music retweets, now it's motivation retweets + copied tweets). Same AI userpics as in 2022. And searching the tweets usually reveals the original pre-2022. But it's weird they don't seem to be "activated" yet
B-Movie idea: zombie film but instead of undead humans its just chat bots shitting up every corner of the internet, doing absolutely nothing of value and eating your mobile data.
The godfather exceptionally agrees. A film seen to be seen as perpetual in its brilliance. Agreeingly accepted the godfather as a cinematic history piece.
It’s awful because it affects every community on the bird app. Any major account makes a post, 4 bots will always beat you to it with blue checkmarks and nothing to contribute
Now, are they just stealing people's pictures, or are all of those simulacrums of real photos generated by AI? It's probably theft either way, but man I'd be pissed if my image was being used this way, and I would think there's some ground to start bringing liable against image use.
this is so funny. 30+ years ago I (like a lot of junior CS folks) wrote an ELIZA chatbot (using basic pattern recognition stuff, not LLM/neural net type stuff). It used to be fun to link two chatbots together and watch them spin out. And here we are doing that on a societal level!
btw there was was nothing advanced about writing ELIZA in the early 90s, almost 30 years after it was first created! https://en.m.wikipedia.org/wiki/ELIZA
i used to play with ELIZA on my apple //c in the late 80s. even as a kid, i understood it used simple patterns to sound realistic. i can’t believe how many people think gpt has “intelligence.”
I once walked into a clothing store full of hip young people walking around and messing with the clothes. Careful observation led me to believe every single one of them were store employees. Like a dozen, in a small store. This is kinda like that, only with blow-up dolls instead
Hard to imagine that there was a time when you could scroll through the replies to a tweet and see funny and interesting comments. Now replies are either from bots or the most unbearable people on the planet.
The bots and Elon Musk's fans are interchangeable. This is all part of Elon's master plan to create a never-ending supply of fresh bootlickers. It's fascinating really. OpenAI cannot process this...
There's something like 10 Leon Audleys in there, and each account name ends in a series of 5-ish digits. All of them seem to have been posting similar replies.
I think it might be the opposite, this is what elon is insisting his "growing userbase" is to his shareholders and so there isn't really any impetus for him to stop bots at all
right, but if it's just bots chatting back and forth to each other, at some point doesn't it end up taking up more storage than ACTUAL users?
and he doesn't need to justify anything to shareholders, it's lenders who financed the sale. lenders are probably more dangerous to him than shareholders.
wherein AIs/computers, running on their own code divorced from human interference, go off the rails and run unimpeded for many years creating this weird empty world of bullshit no human's ever experienced or will ever experience
A perfect extension of human beings discussing anything on the internet, just parroting other peoples' opinions because they are literally incapable of forming their own.
That reminds me of when automated Amazon drop shippers list eachother's products and don't have any controls set, so you end up with a $60,000 book that nobody actually has in stock.
Comments
Pretty sure all of those are hoping dupes will check their profiles and click those links.
or it just went to a different spam generator than chatgpt
Can't wait until everything I read is this.
"Completely agree, The Godfather is truly an an exceptional film."
I do think LLM are a pretty good tool (with some obvious intellectual property and theft questions) for certain purposes. like a systems!
That is, if they don't catch on.
The McDonalds Corporation. You too can chow down like a rabid dog on some Tasty© Burgers© courtesy of The Clown himself!
https://en.wikipedia.org/wiki/Gray_goo
and he doesn't need to justify anything to shareholders, it's lenders who financed the sale. lenders are probably more dangerous to him than shareholders.
wherein AIs/computers, running on their own code divorced from human interference, go off the rails and run unimpeded for many years creating this weird empty world of bullshit no human's ever experienced or will ever experience