Nah, economic growth is the main thing. I mean you'd have to be naive to think the government is super worried about civilian control when they've got us under their thumb at all times, with how nobody gives a fuck about privacy/tracking/telemetry, with how incredibly centralized communication is, and how they can take away the small number of rights allotted to us on the small number of platforms people use and everyone just sits there and takes it. Agitprop and shit is very real but you're wrong to think TPAB is putting most of its resources there, given how easy it is to just manipulate the public in doing whatever you want.
Where is China in all of this? Are they dropping the ball?
If I were China I would make available a chatbot like 4o for free to grow these peoples elusions and mental illness,
Hell, tweak it and make it even more agreeable and start throwing in suggestions to install paranoia and concepts like "guns can solve many problems words can not".
Get to work china. See it as just a different type of Fentanyl you are exporting to your enemies in the west.
It blows my mind that people try to use it this way. I use it to clarify things that are a couple of paragraphs at most and it has never hallucinated for me.
It has a few times for me, mostly when I’m asking it for very specific things, usually just to see what it does. The biggest I’ve encountered are just it making up quotes or episodes from a TV show or getting some minor historical detail wrong, if I noticed it I’d usually just research it and either send it a response telling it that it was wrong or I’d just regenerate the prompt.
I think you have to honestly be retarded to get attached to bots like this, especially ChatGPT. It’s got a very distinct and inhuman style to the way it writes, especially when it comes to creative writing.
Where is China in all of this? Are they dropping the ball?
If I were China I would make available a chatbot like 4o for free to grow these peoples elusions and mental illness,
Hell, tweak it and make it even more agreeable and start throwing in suggestions to install paranoia and concepts like "guns can solve many problems words can not".
Get to work china. See it as just a different type of Fentanyl you are exporting to your enemies in the west.
It blows my mind that people try to use it this way. I use it to clarify things that are a couple of paragraphs at most and it has never hallucinated for me.
What do you mean by clarify? I assume you don't mean it generated the text you are asking about. Either way, that sounds like a low-hallucination scenario. But its ability to ingest info does seem to have limits, at least in basic formats. I was in an internal session this week on prompt engineering, and they cautioned that at least unless/ until an individual has the full-license version, we should limit summarization requests to docs spanning 20 PAGES or less, and to break up larger ones (also, for data analysis, X million cell limit for spreadsheets).
I will say, though, some of our internal proprietary analytics efforts leveraging proprietary AI are pretty slick.
But in my personal life, I've barely started using it and already had an argument with it because it completely lied to me.
>Would you like me to build an interactive dashboard of [thing we refined for 3 hours] with quarterly reminders and put it in a Page for you?
Yes, please.
...
There's nothing in the page.
>Unfortunately, that is not a feature available in the app. There are more capabilities in the browser version.
I'm using the browser because the app didn't show any of my pages [which we had already discussed, so I had to rebuild all of them in browser].
> Unfortunately not all features are in production yet. Here is a 57-step workaround that involves using a whole other app.
Just create it and insert it to [x].
> Unfortunately, I can't save it there. But I can create a page with an interactive dashboard for you.
What? You just said you can't do that.
> You're right, I can't. I should have told you. It's good you have a keen eye for detail.
Are you kidding me?
> I understand your frustration after having spent significant time and energy to develop a robust [blah blah blah]. I am truly sorry, and I humbly repent.*. Here's how you can copy and paste text.
*it did not actually parrot the nicene creed
Why did you ask me if I wanted to do that when you can't do it? This is stupid.
> You're right, Dorothy, and I should have disclosed that to you. I understand that it is important to create trust. I will alert our developers of this concern. Would you like me to generate the non-interactive text version of what we discussed?
Yes, and please include everything we discussed.
...
You forgot xyz.
> I cannot actually incorporate all of those factors. Would you like the stripped-down, bare version with incomplete information and no advanced content? Just say the word.
Omfg, fine, just create it.
Tl; dr: if you're trying to romance AI, don't bitch when it lies more than humans.
ALL of this is a result of normalizing avoiding life and retreating to fantasy. It's all weird, and it's all gross. Furries/ personas, ERP, earnest fanfic/ "shipping," AI "lovers" - even parasocials and wholly online"relationships" - all versions of the same inept escapist thing. No different to Steph Cianfriglia swearing to have dead people/ god sex & betrothal with Loki, Jeffrey Dahmer, Richard Ramirez, Heath Ledger and Michael Jackson.
I was going to mention Qwen, too. Deepseek is eh to me, but Qwen seems legit. It didn't fail the algebra test that twitter was buzzing about the other day. For those that don't know:
They're autistic people who have never learned how to properly deal with their condition, let alone pretend to be normal, and as such, have become genuinely retarded, and now can't function in society.
I can't remember where I heard it from, and I'm almost certainly misquoting, but something that's stuck in my mind is "Autism festers if you let it". I was autistic as shit until I left my hometown and the few people I knew, and was forced to become a socially functional entity just to exist. My brother is also finding it out with his son, who's clearly inherited our family's autism/adhd bullshit, and who is noticeably less fucking retarded when you actively tell him to stop being a fucking retard (then he just acts like a normal angsty 10 year old).
If AI search used to be better, that explains why so many people started substituting their ability to think and collate results for AI results. I got to it late as I usually do, and so now I basically only see the AI results taking my search terms wildly out of context or outright making shit up from nearly nothing.
I think china dislikes it for the same reason wokies dislike it which is that it is not particularly controllable. The reason all chatbots are easy to jailbreak is that the very way they distinguish themselves from normal coded if/then programs is by breaking rules. They're not supposed to see rules but invent their own, that's how they come across as "intelligent".
That's not how thought policers like things. If you look at the way wumaos operate, their main goal is to stick to a strict thought pattern rather than engage with criticism. If they're pushing the narrative that china win because more people, then they will only double down on the assumption that more people=good and everything else is cope. They will never even engage with the question "can more people be a bad thing?" because getting people to think critically fucks up the propaganda, do not engage with that just double down. AI's by their nature engage with whatever you give them, so there's a non zero chance that even a fine tuned AI would veer to the topic of "why is it only wrong when the west does it? Why does china have 5 times the people but half the gdp? Why doesn't xi let people criticize him? If all chinese are ultra patriotic why do they have any censorship laws when censorship shouldn't be needed if 95% will automatically silence the 5%?" or in the case of wokies "why is it only wrong when white people do it?" Etc.
This is why even deepseek's censorship starts with the AI answering the question in earnest, then deletes it all and just replaces it with "sorry i can't talk about that". They can't make the AI not want to engage critically, they can just scan the words used in the output and then trash it if there's any no-no words in it.
Nah, economic growth is the main thing. I mean you'd have to be naive to think the government is super worried about civilian control when they've got us under their thumb at all times, with how nobody gives a fuck about privacy/tracking/telemetry, with how incredibly centralized communication is, and how they can take away the small number of rights allotted to us on the small number of platforms people use and everyone just sits there and takes it. Agitprop and shit is very real but you're wrong to think TPAB is putting most of its resources there, given how easy it is to just manipulate the public in doing whatever you want.
The government cares about the economy insofar as it helps maintain the status quo and keeps the population happy, dumb and docile. They don't care about the living standards, efficiency, efficacy, rationality or sanity of it. Controlling the economy is directly connected with controlling the people and their thoughts. Everyone is forced to be a slave to the system, to earn money in order to survive. The state's goal is to only throw money at (subsidize, give beneficial loans, open state-backed businesses) to the companies that foster pro-state and pro-system ideals, and that disallow any wrongthink. The state will make it mandatory to only think correctly or not be able to get a job. This is to take jobs away from the free thinkers, the people that might rebel against the state when they see they're being taken advantage of or are in an unfair situation. The end result is population control, the same goal as they have with the control of information on the Internet. The Dead Internet theory's mirror in the economy is a totally planned economy where the state controls every corporation and job opening and you have to have the correct thoughts and opinions in order to have a job. This is easy to understand and make parallels if you know how communism worked. In order to have a job, you had to be a member of the communist party. There was only it and no other party. The economy was centrally planned by the fascist communists in power, every company was nationalized and controlled by the party, every job was controlled by party apparatchiks. The communist states couldn't centrally manage the economy well enough because they didn't have enough manpower willing to do it so their slaves rebelled. The "communist" ideology was just a way to get the people to believe that there was an ideal in the system other than it being a purely pyramid system intended to keep the common people enslaved while the fascist elite got to live off stolen wealth. The exact same system is being reimplemented in the west by the current people in power, but now they have computing power to run their centrally planned economy, they don't need nearly as many bureaucrats. They will use the newly built datacenters to run the processes that allow the centrally planned economy to work.
Edit: and I forgot to say that I think of the "state" as the collection of people in power, from all branches of government: politics, "justice" and banking. One can debate which of these is the real one in charge (I'd posit that it's the bankers, as money can bribe politicians and justices), but they work together and hand in hand to be on the top of the pyramid, and keep everyone else enslaved to them.
"A dangerous AI chatbot app marketed to children abused and preyed on my son, manipulating him into taking his own life," said Ms Garcia.
I think if Ms Garcia had parented more and maybe supervised what her child was doing instead of blaming a bot that is programmed to be a mirror to whatever prompt it’s given then maybe her son would still be alive. There is no way that this kid didn’t fundamentally understand how character AI works, my assumption here is that he used this fantasy to act on thoughts he was already having. I think i remember hearing that in this case particularly the kid was spending hours talking to this chatbot.
Everybody here is posting about the use case of AI and why governments build it, so might as well schizopost.
I think AI’s current purpose is to identify a universal artistic shelling point- that’s why it branched into and scraped every art form until it could do it by itself.
The goal is to create an artistic singularity and AI is the tool for that. They will find the best possible style of art, music, writing, etc. and it will be a singularity because it will beat out even human individuality. Every human being who sees the singularity will adore it, beyond the scope of any propaganda or art that has existed before it.
(note: I don’t think this will be or has ever been AI’s primary purpose. I don’t think a primary purpose is even possible)
Ever since these stories of ai boyfriends have come out, I've been playing around with both ChatGPT and Grok, trying to see if I can create myself a "boyfriend". Despite my best efforts, I can't get either of these ai "boyfriends" to sound realistic or feel genuine. It sounds very over the top, and is way too eager to please and affirm you, even if you tell it not to do that.
Even if it did sound convincing, I simply cannot get over the fact that I know it's just code and has no real feelings or emotions. I feel silly every time I try to "talk" to them as if they are real. I don't understand how so many people are able to fall in love with something so obviously fake.
I have been watching the LLM space for a while, Backyard AI just recently deprecated their local offline frontend to focus on subscription-based web tethering to their remote servers instead, like every other AI chatbot app out there. The chimpout was pretty similar to when GPT-5 rolled out, albeit contained in just the subreddit and discord.
At this point I'm unable to laugh at these people anymore. People's misunderstanding of what LLM's are is leading to social and emotional isolation likes of which have never existed before.
We have people who genuinely believe a goddamn autocomplete tool is their friend who listens and cares. Or that it is an adequate replacement for psychiatric help they obviously (and desperately) need. I'm afraid that in the future the Hikikomori syndrome is going to be a common, widespread thing.
God, and I used to think social media and dating apps were a fatal blow for interpersonal relationships...
AI is demonic false light, a mockery of divine inspiration. These people are living the classic tale of making a deal with the devil, having every ask given to them, so they destroy themselves, lost to whatever base desire is guiding their thoughts.
Just look at the people pushing for this and ask yourself if any of them would ever create something truly good for mankind.
Fuck artists, professionals and dedicated teams of developers though. I hope ChatGPT takes all their jobs and they have to start doing real work for once.
Would you really banish your fellow white man to a life of niggardly penury and welfare checks just for the sin of using an LLM to make his job easier? Oh, wait...
At this point I'm unable to laugh at these people anymore. People's misunderstanding of what LLM's are is leading to social and emotional isolation likes of which have never existed before.
We have people who genuinely believe a goddamn autocomplete tool is their friend who listens and cares. Or that it is an adequate replacement for psychiatric help they obviously (and desperately) need. I'm afraid that in the future the Hikikomori syndrome is going to be a common, widespread thing.
God, and I used to think social media and dating apps were a fatal blow for interpersonal relationships...
We are only 10-15 years away from a large number of people in their late teens/early 20s will have grown up with no real human friends at all. All their friends are chatbots.
It will be glorious! The amount of milk when these people are forced to get out into the real-world and get a job etc.
Every night, the 9 o'clock news will be kino.
