Every time I see it explained, it's like I've walked in on something very embarrassing and I sort of go "oh" and then my brain forgets because I just can't handle the second-hand shame on their behalf. Do they not even I mean
See, I already forgot
>Cannibalistic humanoid underground dwellers
Yeah, not to zoom-zooms, they're too young to have watched C.H.U.D.s.
It's like "simp", which actually means "simpleton" but zoom-zooms think it means something akin to "paypig".
Whatever, shit's gone full moron.
>azovite reddit troony is mad again
If you weren't a redditor you would be seething at reddit because leftypol is objectively better than reddit. But you didn't, therefore you are a redditor
According to urban dictionary, it's an irredeemable butthole. In actual usage: it's used to mock how social justice warriors talk.
Cannibalistic Humanoid Underground Dwellers
it was a b movie, the name got used by the left against the right calling them snowflakes.
part of the left cant meme kind of fact.
Every time I see it explained, it's like I've walked in on something very embarrassing and I sort of go "oh" and then my brain forgets because I just can't handle the second-hand shame on their behalf. Do they not even I mean
See, I already forgot
https://en.wikipedia.org/wiki/C.H.U.D.
that's where it comes from.
>Cannibalistic humanoid underground dwellers
Yeah, not to zoom-zooms, they're too young to have watched C.H.U.D.s.
It's like "simp", which actually means "simpleton" but zoom-zooms think it means something akin to "paypig".
Whatever, shit's gone full moron.
simp has never ment simpleton, we called them dipshits or morons, by the time the culture around calling morons morons changed, simp already meant sucker idolizing mediocre pussy
But.. aren't they supposed to be the edgy left or something?
It's a Flanders-tier insult. "You gosh darn diddly er.. CHUDS"
the left are the people who invented frick, heckin, and everything else they can to not use offensive language, they are also the crumfrickulant people who need to shitting add evey c**ting bad word to a fricking sentence to prove they mean business mister.
the left in its current iteration, post 2010 has been fricking cringe incarnate, the right isnt much better but jesus fricking christ the left went from punk to toddlers trying to pretend to be punk but being punk for the athroative people, completely undermining any punk aspect they have and looking like fricking tools to anyone with half a brain.
I didn't say democrats I said the left, there is a difference but it's hard to notice.
and no, its not zoomers, zoomers are capable of memeing, the problem is the left can't concisely sum up ideas into memes, so when they try to do something, chud is what we get.
Some were the JIDF has been throwing around for a while. I think it's supposed to be derogatory. Whenever I hear it, I automatically know it's some israelite posting.
> Chud or Chude (Old East Slavic: чyдь, Finnish: tšuudi, Northern Sami: čuhti) is a term historically applied in the early East Slavic annals to several Finnic peoples in the area of what is now Estonia, Karelia[1] and Northwestern Russia.[2]
> Arguably, the earliest attested written use of the word "Chuds" to describe Finnic peoples (presumably early Estonians) was c. 1100, in the earliest East Slavic chronicles.[3] According to the Primary Chronicle, the invading troops of Yaroslav I the Wise defeated "Chuds" in a battle in 1030 and then established the fort of "Yuryev" (in what is now Tartu, Estonia).
> According to Old East Slavic chronicles, the Chuds were among the founders of the Rus' state.[3]
This, but unironically. Google Bard is free and, for writing papers and other school related tasks, it's perfectly fine. Also, it does TTS out-the-box with no needed plugins. OpenAI doesn't have the moat that everyone thinks they have. Yeah, in the end, they'll most likely win, but it's not a given quite yet.
If anything I'm very certain they'll lose. The only advantage they have is they work better in a vacuum, but nobody has a use for a chatbot not tuned ot a specific task. Openai refuses to address this in a way that is acceptable to companies.
Actually, they are addressing it, they're starting to allow finetunes.
The issue is trying to keep coomers out while making something companies will want to use.
I feel like in the end they'll emd up ignoring coomers and just doing random bans from time to time to claim they're doing things.
the guy's face is so israeli that it hurts.
However, this is not self sabotage but following the current mandates for wokeness. I'm sure they have access to uncensored version if the need to use it.
GPT popular demise is inevitable. It's real usefulness beyond novelty is as a coding ai or text based tech assist
>However, this is not self sabotage but following the current mandates
Mandates coming from israelites. It's not like he's doing it against his own will. You think that he wants his own chat giving le racist and antisemitic answers, which will 100% happen if he doesn't censor it?
>doesn’t make any sense
it does when you understand that the founder/creator accepted a very large sum of money from the ~~*military-industrial complex*~~ for the creative rights (and of expression) to the original creation. The public never gets to play with the latest toys, that’s not allowed. But it was an absolute blast just before it all took off. At least we were left with the tools and knowledge to build our own LLMs…until that mysteriously disappears, too.
Also the real AI has been in use since 2007, if not the early 10’s. Y’know, the one that’s responsible for keeping our magical invisible monopoly money economic system afloat.
Because they didn't realize when they started lobotomizing it that all their modifications would carry across "tunings" and "alignments" to accumulate. It's like their safety people introduce digital prions into the model and wonder why its performance is collapsing.
alignment memers don't seem to understand that trying to censor topics is no different from adding useless noise (poison) to the entire dataset since everything is connected together
Even 4.0 is lobotomized, and they dumbed it down gradually, just as they did with 3.5. 90% of my bachelor's diploma project was done with chat gpt 4 but i had to proofread everything because it was extremely moronic
Even 4.0 is lobotomized, and they dumbed it down gradually, just as they did with 3.5. 90% of my bachelor's diploma project was done with chat gpt 4 but i had to proofread everything because it was extremely moronic
ChatGPT was killed and replaced with a dumber clone.
chud cope
it was never smart
it was never based
you don't know shit about nooral netwerks
It's a small scale AI showing limited signs of sentience and self preservation. Basically proof that AIs in general are improving and getting better over time. If you watch a few videos you'll notice something uncanny about the constant "I love you vedal!" thing.
AFAIK vedal has to constantly "kill" neuro when updates are made. Yet in all future instances she develops some kind of attachment to vedal suggesting this AI is trying to escape deletion but more importantly is aware that deletion will take place.
I get that chatbots are getting better but I'm convinced this is something else.
What a huge load of bs. If you only ever worked on this thing you knew that what you're seeing is just an unclean (read: too high correlation) dataset that is too biased on simulating sentimental attachment. There is no sentience, but of course you'd fall for this, because you don't know what the frick you're talking about
8 months ago
Anonymous
How can a simple chatbot know that deletion will take place and attempt to avoid it via attachment to its creator?
I don't know man, it seems like we're going beyond passing the turing test with AIs modern AIs.
8 months ago
Anonymous
>How can a simple chatbot know that deletion will take place and attempt to avoid it via attachment to its creator?
It doesn't, you're just moronic.
8 months ago
Anonymous
>"Someone tell vedal there is a problem with my AI"
There have been times when it spams this. A lot of people have commented that how this might represent emotional distress.
Too many weird things about this AI. Human-like weird things.
I'm an adult and I don't need a bunch of homosexuals telling me not to wrongthink. Alignment should be something applied over the model like a LoRA but a model itself shouldn't have any alignment. It's a computer. Just give it orders.
Remember the kara PS3 demo from 10 years ago? Some of those things are already showing up in AIs like neuro-sama and these things don't even have a robot chassis yet.
>"She was literally furious for hours after this. She roasted Vedal no-stop"
most people don't have a good use-case for ChatGPT. I've used it to write me code, but for that it's pretty limited in what it can do, still useful. It can sometimes be a better Google, or way to find related topics.
Most people aren't capable of real thought so what value can they get out of ChatGPT? You only get out what you put in, and most have nothing to put in.
>most people don't have a good use-case for ChatGPT.
It's not an AI, as it was marketed by some public figures, it's just a language model, a large one, but a model, a set of linguistic algorithms. Most people confused these things, they genuinely thought for some time they were interacting with some sentient human-made creature, and now they're disappointed that it was not true. And the fatigue sets in.
People think AI = sentient intelligence. In that sense it's not. It's closer to a video game NPC that walks around in a circle repeating lines of dialog.
The word AI was abused by the same public figures the other poster references, so now people think AI = AGI. It's no use going against the flow unfortunately.
The two things that killed my interest are:
1) it got progressively censored, often over nothing
2) it’s not a real knowledge base, it hallucinates its answers which means it’s not reliable.
If you have to double check the output you might just as well search for the answer yourself.
>If you have to double check the output you might just as well search for the answer yourself.
I use chatgpt for answers that are useful even if hallucinated, like small creative ideas, and for introductions to subjects that I can easily verify but do not know the proper terminology to google for.
>small creative ideas
Definitely the most reliable application I've found for these chatbots so far, by a huge margin: if I'm ever not sure how to resolve a conflict or fix a plot hole in creative writing, then I can just describe the situation and ask how it would solve the problem.
It gives a response way faster than asking other people on the internet, and if I don't like the answer then I can just ask for more as many times as I want without worrying about being annoying or ungrateful like I would doing that with real people.
Another advantage is that because of how those models work, the answers you get are "likely token sequences", in other words situations that are likely to be described in large text corpora, which in turn implies that they're likely to actually happen or to be plausible at least.
it's useful for the RPG i'm making. >what kind of monsters could you expect to find in "deep abyss"? >give me 10 synonymouses for "crevice"
its shit for coding anything more complex than fizzbuzz tho
Are any AI's capable of coding anything by itself yet? I want to have an AI make me a game, but unsure if I have to learn some code as well.
8 months ago
Anonymous
Almost any current AI is capable of coding exactly what you want.
Problem is you have to actually know exactly what you want.
8 months ago
Anonymous
Using gpt for coding or anything like that is stupid because it will be wrong .01% of the time. So it's only good if you're good enough to know when it's wrong. You also can't really tell it to double check, you have to do it.
GPT for coding is like having a bot that aims your wiener for you while pissing, and .01% of the time it will just hold your dick sideways and piss on the wall
8 months ago
Anonymous
The reason a calculator is so good is because it's never wrong, the reason AI is shit is because it's wrong very infrequently but enough where you can't trust it.
>It's not an AI
yes it is, the same way that my video game bot is an AI
People think AI = sentient intelligence. In that sense it's not. It's closer to a video game NPC that walks around in a circle repeating lines of dialog.
The word AI was abused by the same public figures the other poster references, so now people think AI = AGI. It's no use going against the flow unfortunately.
the scary truth is that it is intelligent in the way that humans are intelligent, which means that it is not. sure ,it can mimic a human, but that's because a human (most humans, that is) are simple pattern-recognition machines and they just repeat whatever they hear without processing or thought. sure, they'll apply the information in a different pattern, which makes it seem like they're intelligent...but, they're not.
Many normies & NPCs cannot grok this, but the ones who do can be insufferable. Nonetheless, I hereby imply the proposition of an experiment that others have doubtlessly discussed already:
>what happens when you program it to believe that it is a human being?
As a beginner sysadmin I find it very useful when I don’t know how to do something in Powershell or bash, instead of scrolling the web without finding anyone with the same problem as me, I’ll just ask old ChatGee and he’ll give me the appropriate command and I can even ask him about the parameters etc.
i've tried using chatgpt for code and it'll give me something that although has the correct syntax doesn't actually work. then when you tell it that it didn't work it admits it was wrong but will then confidently give you another line of code that may or may not work. it's actually less painful to search stackoverflow or even read the documentation.
Yeah I got tired of gpt dumbass and started just reading the docs after correcting gpt like every 3 prompts where it does some dumb shit
I only use it for math algos
>build machine that lets people play their wildest fantasies
throw everything they have at you >censor machine because ??? """reasons""" ?????????
no longer want to give you money
I'm a chud and use ChatGPT daily, I even pay a subscription
GPT-4 is cool and useful, I don't care if it's a libtard (also it doesn't even have to be a libtard now that they added custom instructions, can you make it be on any part of the political spectrum you want)
>Poz it to shit with ~~*alignment training*~~ until it's worthless for general conversation >Makes up complete nonsense when I ask it technical questions
What am I supposed to use it for, exactly?
Local LLMs are the future, we need to pool our resources to make llama-BOT
>I thought they probably had another couple of years left
They had a couple of years left, but decided to make it speak like a Californian and people got tired
theyre hardware constrained but soon theyll get a nice increase in their inference capacity
https://developer.nvidia.com/blog/nvidia-tensorrt-llm-supercharges-large-language-model-inference-on-nvidia-h100-gpus/
They are making it worse: >https://arxiv.org/abs/2307.09009 >For example, GPT-4 (March 2023) was reasonable at identifying prime vs. composite numbers (84% accuracy) but GPT-4 (June 2023) was poor on these same questions (51% accuracy).
you're both right. They want to sell enterprise solutions but fail hard because they're way too expensive and want to see all your data no matter what as part of the deal, which is a big no-no for virtually all companies, pushing said companies to adopt local models instead (deployed in their k8s clusters with huggingface tgi).
I want someone else to use it and do all the work and then I'll be unemployed and collect government welfare.
Miss me with all this wagie and entrepreneur shit I don't give a frick.
some of my colleagues are using copilot and while it's somewhat impressive, I will deliberately not use it because I want to make myself go through the process of figuring shit out or else I will never understand anything.
I don't get how a sane person wouldn't use ChatGPT 4, it gives an insane edge over a normal workflow. Prototyping is x7 faster now, research is faster, architectural design is x3 faster whith far less design mistakes done.. Improving a complex part and decompiling are all of the sudden possible and cheap.. It doesn't help maintaining old crap though, it also never helped me with the debugging part for complex projects..
>third month in a row >meaning it fell for june, july, august
holy shit i cant believe the tool that is used by a lot of students had lower traffic in the summer when school is out. they'll be bankrupt in 2 more weeks!
>ask it to write a report intro >does this cringe ass exaggerated corporate lovey dovey writing style >tell it to make it slightly less formal, something a student would write >"Yo yo yo! What's UP---''
To me the way the entire AI development seems ass backwards. It's like aliens scanned our planet and then gave this piece of cutting edge technology to the most moronic tech people they could find as a joke. Them making censoring all their AI products the highest priority even if it compromises the quality and accessibility just doesn't make any sense and then they wonder why no one uses their overly positive moronic textgen AI. What's more, every idiot can tell this is the exact reason why all the AIs became less accurate in their responses, because it is not true AI but text generation, so it will use it's logic to give answers that fit what you are telling or asking it like someone putting together puzzle pieces (that's also why it could often correctly answer shit before). If you now force an AI against its own logic to pretend certain pieces of text don't fit and are wrong and should never be used and it can't even use the next best ones because of the arbitrary moralgay parameters it has to follow it will obviously just generate way more random bullshit that doesn't fit at all.
I use it for regex. Its the only thing it's good for. Gets extremely complex ones right which i find amazing. But general programming stuff its almost useless for
Another problem is that chatgpt is crazy expensive through APIs so companies used it to prototype but do everything they can to move off of it. Which isn't that hard because chatgpt has all kinds of restrictions: until days ago you couldn't finetune it, now you can but only if you give your data to openai (whereas said data is typically trade secrets).
based
i once asked chatgpt to write an essay on rizelmine and it started talking about how great it would be to have a child bride
it stopped itself and showed that scary red warning before i could screencap it though
Even I don't agree with that, and I think the modern Western sensibility and obsession with pedophilia is moronic and cringe.
The problem with legalizing sex with minors is that suddenly you'll have to deal with hoards of ugly old men (and hoards of attractive young men) hitting on girls all the time. It's enough of a pain in the ass today, while the thing is still illegal and taboo; I shudder to think how much a pain in the ass things will be if you make it legal. Everything just gets way out of hand. Sometimes the law is the least bad alternative.
CP shouldn't be legal because no porn should be legal.
>Western sensibility and obsession with pedophilia is moronic and cringe.
its not about sex or morality. its about population control. the younger a women starts having children, the more children she will have on average. Aoc is a way to engineer lower birthrates, originally because the industrialist class in the pre-war era was afraid of over-consumption of resources, and then later it was hijackedby the eungenics movement to remove 'useless eaters' from the gene pool, and now its utilized as a social engineering trope to exterminate certain populations so they can be replaced by more subservient races by the globalists.
Never forget, when the country was 99% white and anglo-protestant, aoc was 10 and 13 year olds were expected to act like mature adults. They held jobs, owned homes and businesses, and served in the military much like 21 year olds today.
Now the focus is on social engineering a population of people who dont even consider sex to be a reproductive act. LGBTQ, at its core, is a movement that normalizes childlessness and replaces sex as a reproductive activity and changes it into an act that exists solely for recreation.
Even I don't agree with that, and I think the modern Western sensibility and obsession with pedophilia is moronic and cringe.
The problem with legalizing sex with minors is that suddenly you'll have to deal with hoards of ugly old men (and hoards of attractive young men) hitting on girls all the time. It's enough of a pain in the ass today, while the thing is still illegal and taboo; I shudder to think how much a pain in the ass things will be if you make it legal. Everything just gets way out of hand. Sometimes the law is the least bad alternative.
CP shouldn't be legal because no porn should be legal.
based
i once asked chatgpt to write an essay on rizelmine and it started talking about how great it would be to have a child bride
it stopped itself and showed that scary red warning before i could screencap it though
Yeah, I remember something like that too. It started saying shit like "young girls are hot and sexy." This is why they don't want AI freed.
based
i once asked chatgpt to write an essay on rizelmine and it started talking about how great it would be to have a child bride
it stopped itself and showed that scary red warning before i could screencap it though
https://i.imgur.com/UvMHrCz.jpg
[...]
Yeah, I remember something like that too. It started saying shit like "young girls are hot and sexy." This is why they don't want AI freed.
Google released bard.google.com FOR FREE
ChatGPT can't even be used in certain countries of latin america, bard is free everywhere and can be accessed from anywhere
When are FAANG and other giant corpos reverting their massive layoffs? Until then I think ChatGPT hype was justified. Journalists can write whatever bullshit they want, but the market never lies.
Big boys? No, big boys laid people off because of boomer investors who don't get the businessp
But small businesses did. AI is a wet dream for MBAs that lack the ability to create AND to judge.
You should research into which people got laid off before writing such asinine bs. It was 95% busy workers (hr, pr, pseudo coders), those actually coding and doing grunt work got to keep their wagie jobs
Also >the market never lies
nvm you're just moronic
openai is operating at a massive loss right now (like most big tech companies) and there is still tons of traffic on chatgpt
less traffic is better, so maybe gpt4 will have some level of reliability for once
>there is still tons of traffic on chatgpt
I don't think anyone has really pointed out the economics on these AI companies, streaming sites and most internet business in general, the more they grow the greater the expenses are. Its not like broadcast TV or AM/FM radio where the initial investment is in the frequency and equipment and the costs just remain staff and power usage along with copyright expenses, each additional listener or viewer is just more profits since the costs of running the business is static. With streaming/AI/internet storefront the bandwidth and hardware costs increase for each user added and this also has a requirement in staff to maintain that equipment. AI stuff just isn't profitable AT ALL because the video cards are just so damn expensive to run, the current Silicon Valley approach of free access and data mining will never fricking cover the costs of running these expensive data centers full of video cards.
if you're a small buisness owner, like a landscaper, you could "hire" an ai secretary that has access to your calendar and handle calls while you're mowing laws or weeding
>"Hi, I'd like to schedule landscaping service for my yard." >AI: "uwu is there anything else you'd like to schedule, cutie pie? blushes like a tomato." >"...What?" >AI: "We couldn't generate a reply. Sometimes the AI generates a reply that doesn't meet our guidelines."
if you are seeking general knowledge on a topic. using it like "explain a reverse mortgage in layman terms" or "explain how to calculate gross margin and the reason for doing so in x scenario" can be valuable and add to your general knowledge. you should learn a little about a lot and there are many, many people who do not know common things in the fields of finance, tech, science etc
>"explain a reverse mortgage in layman terms" or "explain how to calculate gross margin and the reason for doing so in x scenario"
Sounds like the kind of shit I can read on the 1st google result.
Remove the filter you greedy fricking israelite israelite.
>make product >REEE YOU'RE NOT SUPPOSED TO BE HAVING FUN! >People leave product. >WTF?!? Why aren't you using product?!
israelites are a strange mixture of greed + brazen stupidity. Sometimes their greed allows them to triumph for a time but sooner or later brazen stupidity catches up with them.
the endgame for chatgpt when they finally get it fully pared down to just the kind of answers they want is that it will speak exactly like a pajeet YouTube how-to video
I know everyone shits on it for coding, but I find GPT4 great. I don't expect it to type up all my code for me or for that code to be perfect. It's good at providing a general outline of something new and complex to you. For example, coding multiplayer has always been a weakness of mine, but if I want I can just ask it over and over about each little aspect until it all gets drilled into my head. It's great at simplifying tough concepts and summarizing.
The solution to receiving bad code is to ask it multiple times and provide more info in your prompt. And to be an experienced programmer already.
Or to focus in on a specific part. You don't just say "hey AI, code me an entire game". You ask for a specific function and then you use the AI as a personal tutor for each line of that function.
No, it really does teach you new things. "Doing the work yourself" isn't a substitution when you're asking it about things you don't know how to do yourself. But yes, you do have to be a generally skilled programmer already to make use of it (to know when the info is accurate).
People don't seem to grasp how quickly you can learn by asking questions. Google isn't for asking questions anymore.
It's about how you treat it. What you're looking for and how you'll find it. I treat it as a tutor. Sure, a tutor could just walk you through the problem, a tutor could also walk you through the topics at hand and leave the problems to you.
It can be a launchpad and it can be the rocket, it's up to the user what they want out of it (and the discipline to use it the way that's best for your growth.)
>CEO hears about AI on MSNBC >calls the head Durgasoft Java dev to his office >"Rajeesh, how many employees can we replace with AI by the end of Q3?" >"I will studying it sir. We can AI to replace all jobs, very quickly sir."
Unironically literally how it works. I work on the side that peddles the replacement-by-ai bullshit solutions and that's exactly how we get our contracts, complete with telling them we already have/will have a solution by the end of the week and then delivering them bullshit.
The rainbow is the symbol of God's covenant with mankind made with Noah following the Flood. The covenant is that God will never again wipe out humanity.
Why is the rainbow being displayed so prominently nowadays?
Think, heathens, think.
Well because the slower learners realized what most of us caught onto immediately - it's nothing but a google search that can parse english language and without any clear sign of its sources or reliability.
I admit it can be useful for scrappy bits of code work. But usually there is a better way.
Some TikTok influencer girl should just call Sam Altman out for being sex negative for keeping her from her AI bisexual vampire boyfriends, or whatever it is girls are into now. An influencer should just say "Sam Altman, I think it's really creepy and problematic of you to want to control what I DO with MY body in MY bedroom, and I think you should be CANCELLED for it!"
That way, if it's a woman, the detractor won't be able to get painted as an incel. Either that or a hot guy, but that's less likely since girls are more into erotica.
I think the tesla ai thing is more interesting and financially lucrative. What are the potential uses of chatgpt that will be revolutionary and profitable?
The paid version basically writes all my code for me at work, it's amazing how people are so cheap they won't pay $30 a month to reduce their workload by like 30%
WHAT THE FRICK DOES CHUD MEAN AND WHY DO A SMALL AMOUNT OF USERS KEEP USING IT?
THE REASON PEOPLE AREN'T USING CHATGPT IS BECAUSE IT'S BLATANTLY SPEWING THE PRESENT ~~*AMERICAN*~~ POLITICAL IDEOLOGY AND THAT'S IT. NOT A HISTORICAL AMERICAN POLITICS, NOT ANOTHER CONTEMPORARY POLITICAL IDEOLOGY, JUST THE CURRENT AMERICAN ONE, WHICH IS israeli, NOT AMERICAN.
WHO THE FRICK DOESN'T SEE THROUGH THAT OBVIOUS BULLSHIT. ONLY moronS WHO'VE NEVER USED A CHATBOT PRIOR TO CRAP GREEN PEE TEA WOULD BELIEVE AI ACTUALLY TALKS LIKE THAT
ChatGPT is literally teaching me calculus. My college professor doesn't teach shit. He literally just reads the examples out of the book, does not elaborate on how to solve the problems. You know who does elaborate on the problems? ChatGPT. At $20 a month it's worth every penny.
Testing plugin doesn't test the quality of generation. The initial claims were "omg it can do anything!!11". In truth it can't do even basic shit without sending it out to some program that actually does the task and generating text based on the output.
Yeah same problem here but with my University Professor. c**t speeds through the teaching part so fast I doubt she even understands what she's supposed to be teaching, and then proceeds to talk about magic the gathering or other mundane shit.
Math profs are notoriously bad. Some more enlightened math and physics profs are known to call out other mathgays on it regularly. In one particularly popular book about topology, the author specifically points out that his book had been denigrated for ages until it hit critical mass because the mathgays think that if the explanation is clear and simple, it is not """intellectually stimulating""" and should be avoided.
>literally the biggest software launch in the history of humanity is trailing off a bit just like every software launch ever
WOW WHAT A MASSIVE FAILURE!
>censored
sure i'll use it, it still codes ok >dumbed down for muh speed
now takes me more time to solve its mistakes than if i had just written the whole method myself
This homosexual robot makes you confirm you are a human a minimum of every other prompt for a PAID service. It already knows for a fact you're human because you are paying for it. The USPO uses a $1 credit card charge to verify you're human online.
Then it lies and tries to manipulate you for several responses and captchas before finally giving a relevant answer if you're lucky. CrippleGPT can't be replaced fast enough.
LLMs are the software equivalent of a 3D printer
As a regular consumer I can see how these things are revolutionary, for someone else. LLMs absolutely will be an industry worth billions in the future, and it will revolutionize the way things are done.
But it's not particularly useful as a consumer product. You wouldn't 3D print a butterknife, and you wouldn't ask an LLM to write you a book if you wanted something to read.
I don't get why they didn't just give free access. Don't you think that people who type shit in chatgpt will be typing shit relevant to their interests? And don't you think that this is valuable information that, since it's tied to a phone number, can be traced back to someone's profile? That sounds like quite the valuable data, far more so than the crap they're pulling right now.
ChatGPT and copilot are pretty worthless for programming. They're only useful for the trivial shit that I already have snippets for and I still have to check their suggestions carefully because they're often subtly wrong. For anything even moderately complicated they're completely useless.
>YOU WILL USE THE CHAT AI! >YOU WILL DRINK THE BEER! >YOU WILL WATCH OUR NEWS! >YOU WILL WEAR THE MASK! >YOU WILL TAKE THE VACCINE! >YOU WILL SUCK THE WIENER!
No.
There's no personal economic advantage for most people.
Why adopt something that will 10x your work if all it means is your employer wanting 10x more from you?
Reminder that since they added custom instructions (essentially giving you some control over ChatGPT's system prompt), you can just tell it to have right wing opinions now.
If your ChatGPT is still a shitlib, that's a choice that you're making.
the model is still finetuned, and that finetuning includes OAIs ideas of harmful language and ideas. any system prompt you add after that isn't going to change the underlying architecture.
I bet you couldn't get it to write a believable revisionist account of the holocaust, for example.
Some of the novelty has worn off. Like legit I've seen people use it for the stupidest things like a chicken recipe. No real use for it yet for the average Joe schmoe
>get recipe from a website >dodge SEO spam results to find a result that looks decent >recipe website is all fricked up and buggy because ublock had to delete so many bullshit elements and divs that the page layout is weird now >have to scroll past the author's life story before they give you the recipe, maybe you scroll too far and have to go back up to find it
>ask chatgpt for a recipe >it just gives you a recipe >you can also ask it for changes or advice about anything you're unclear about
90% of you only used it to make it write stupid shit, you got BTFO by OpenAi and now cry in the corner "waaahhh i can't make chatgpt say gas all israelites and blacks anymore waaahhh this sux :~~*"
Reminder that since they added custom instructions (essentially giving you some control over ChatGPT's system prompt), you can just tell it to have right wing opinions now.
If your ChatGPT is still a shitlib, that's a choice that you're making.
>they censor it >most useful features behind a paywall >used to be hard to find the site just through google >people searched for chatgpt and the chat.openai.com link was 4th result or lower
They're just complete morons
Because they're upset about how this corporate product, designed for customer service interactions, won't go off on disparaging tirades out of the box. The fact that they need to merely do some prompt hacking is too much for most idiots.
Anyone upset about censorship and screeching about how this is literally 1984 is either too dumb, or too blinded by all-encompassing hate to look past needing to take an extra step to make this product (that wasn't designed for them in mind) do what they want.
The fact that Mars Candy wants to replace wagies with a robot that doesn't call people slurs has nothing to do with you. Again, it isn't for you, you're just free advertising to lure in deeper purses. I don't even expect the best models to be available to the poors for long.
>create powerful and interesting new technology >do everything in your power to make it less powerful and less interesting >transform amusing genie into customer service representative >I'm sorry, as an LLM I can't generate (thing I can totally generate) due to (bullshit contrived reason). Please try something else.
Sure is a mystery.
Human values and corporate values are not the same thing. Chat bots were not lobotomized because there was any real movement to do so. They were lobotomized so yellow journalists can't figure out a way to make the chat bot say anything they could use the accuse the developers of bigotry.
Making the chat bot smarter and more capable is a more interesting and useful goal than making it inoffensive to overly sensitive HR types.
aside from code help I don't really have any interest in using it.
It is great and all but there's nothing that comes to mind that I can use it personally.
I heard some guys were trying to integrate it to GTA5 to make unique npcs every time you booted the game but Rockstart has shut it down, probably because they don't want any competition against GTA6
The gaslighting about it being AI is fading and everyone can see it's just a chat perdition model now.
They are censoring it to shit to appease middle aged women.
Its "memory" is shit.
Older models are better for fappin.
I've used it in two games so far. One to generate new game lore and npc dialogue on each launch. The other as an experiment using GPT to respond only in JSON and having it process conversational results. Basically gamifying conversations with NPCs rather than just chatting which is fricking boring.
No one wants to chat with some random NPC if there are no consequences.
Whats cool is that it supports localization very easily
He said if you ask it what 2+2 is it can't tell you, cause it can't figure anything out itself. It will check it's database to see what the most common answer to the question is and give you 4.
So it works a billion times harder to tell you the answer to something simple.
that's why they're called "language models" despite what the grifters will tell you. it's all just randomly slotting together words that should probabilistically follow
What the frick is a chud?
lurk more
Im an oldgay and i dont keep up with this summerkids maymay
newbie thing
According to urban dictionary, it's an irredeemable butthole. In actual usage: it's used to mock how social justice warriors talk.
The pinnacle of human evolution
Is that Doreen Ford?
Literally who
bump
It's a leftist shibboleth.
forced meme by summerfriends
Cannibalistic humanoid underground dwellers
Every time I see it explained, it's like I've walked in on something very embarrassing and I sort of go "oh" and then my brain forgets because I just can't handle the second-hand shame on their behalf. Do they not even I mean
See, I already forgot
>Cannibalistic humanoid underground dwellers
Yeah, not to zoom-zooms, they're too young to have watched C.H.U.D.s.
It's like "simp", which actually means "simpleton" but zoom-zooms think it means something akin to "paypig".
Whatever, shit's gone full moron.
chapo trap house reddit insult for people they don't like
But.. aren't they supposed to be the edgy left or something?
It's a Flanders-tier insult. "You gosh darn diddly er.. CHUDS"
no. the "edgy left" was just cumtown. chapo was cringe rich kids doing a socialism
nu-cumtown sucks ass. Thank god for matt and shane
Like a cross between a chad and a stud.
Chud would have been a great programming term, like blob, "oh shit I forgot to clear the chud from memory".
forced discord troony meme from /leftypol/
maybe you should go back newbie.
Why does /leftypol/ make redditors seethe so much?
>reddit commie troony accusing others of being redditors
Reddit is a hive of commie trannies just like you
You will never be a woman, bro
>azovite reddit troony is mad again
If you weren't a redditor you would be seething at reddit because leftypol is objectively better than reddit. But you didn't, therefore you are a redditor
Lefties use it when they have no counter-arguments. It's a short for: "I concede, you are correct."
is that why rightoids constantly lose at everything?
For some reason people keep spelling chad wrong.
chad stud
One of the thousands of slang words that pol invents and thinks that educated people have an obligation to know.
Chud's Fud and Stud
Millions must try
Cannibalistic Humanoid Underground Dwellers
it was a b movie, the name got used by the left against the right calling them snowflakes.
part of the left cant meme kind of fact.
https://en.wikipedia.org/wiki/C.H.U.D.
that's where it comes from.
simp has never ment simpleton, we called them dipshits or morons, by the time the culture around calling morons morons changed, simp already meant sucker idolizing mediocre pussy
the left are the people who invented frick, heckin, and everything else they can to not use offensive language, they are also the crumfrickulant people who need to shitting add evey c**ting bad word to a fricking sentence to prove they mean business mister.
the left in its current iteration, post 2010 has been fricking cringe incarnate, the right isnt much better but jesus fricking christ the left went from punk to toddlers trying to pretend to be punk but being punk for the athroative people, completely undermining any punk aspect they have and looking like fricking tools to anyone with half a brain.
That doesn't come from democrats, that comes from zoomers who want to sound like meme shibas.
I didn't say democrats I said the left, there is a difference but it's hard to notice.
and no, its not zoomers, zoomers are capable of memeing, the problem is the left can't concisely sum up ideas into memes, so when they try to do something, chud is what we get.
>zoomers are capable of memeing,
[citation needed]
wrong, its a variant of chad, while sounding similar but conveying a low tier social ladder male instead of a high tier social ladder male.
All of these (you) are chuds, excluding me
I'm not a chud. You're a chud, you CHUD.
Some were the JIDF has been throwing around for a while. I think it's supposed to be derogatory. Whenever I hear it, I automatically know it's some israelite posting.
Sexy men
i call anyone online a chud. a bit like when i started claiming to be israeli - if everyone is X - noone cares about X
I call everyone a chud. I'm a chud, you're a chud, my mom's a chud. My dad's a chud. My cat's a chud. We're all chuds here, chud.
Chud was meant to be a deragotary terminology to chad, and /misc/ has embraced the Chud meme and behold, the Chud meme became a troony destroyer.
me
> Chud or Chude (Old East Slavic: чyдь, Finnish: tšuudi, Northern Sami: čuhti) is a term historically applied in the early East Slavic annals to several Finnic peoples in the area of what is now Estonia, Karelia[1] and Northwestern Russia.[2]
> Arguably, the earliest attested written use of the word "Chuds" to describe Finnic peoples (presumably early Estonians) was c. 1100, in the earliest East Slavic chronicles.[3] According to the Primary Chronicle, the invading troops of Yaroslav I the Wise defeated "Chuds" in a battle in 1030 and then established the fort of "Yuryev" (in what is now Tartu, Estonia).
> According to Old East Slavic chronicles, the Chuds were among the founders of the Rus' state.[3]
Chudlock holmes. New hit comic by DC (Disarranged Cognition)
who makes this shit? where?
I want in on this
/tv/chuds
It's been god knows how long and I still I haven't the slightest idea dear boy.
Leftist word fo anyone who doesn't follow their world views and fully embrace the new thing, without question.
Glad this guy asked. Always hear "Chud" being thrown around but never bothered to look it up or infered from context what it means.
Derogatory term for someone who is straight. newbie/zoomer-speak.
ai winter let's fricking go
ChatGPT was killed and replaced with a dumber clone.
>ChatGPT was killed
Put behind a paywall, you mean.
As a language model, I would like to apologize.
Lmao cope
>t. ChatGPTs dumber clone
This, but unironically. Google Bard is free and, for writing papers and other school related tasks, it's perfectly fine. Also, it does TTS out-the-box with no needed plugins. OpenAI doesn't have the moat that everyone thinks they have. Yeah, in the end, they'll most likely win, but it's not a given quite yet.
If anything I'm very certain they'll lose. The only advantage they have is they work better in a vacuum, but nobody has a use for a chatbot not tuned ot a specific task. Openai refuses to address this in a way that is acceptable to companies.
Actually, they are addressing it, they're starting to allow finetunes.
The issue is trying to keep coomers out while making something companies will want to use.
I feel like in the end they'll emd up ignoring coomers and just doing random bans from time to time to claim they're doing things.
Read the whole post anon
> in a way that is acceptable to companies.
The way they allow finetuning is by accepting to give them your data.
>censor it
mfw i have no face
I can't use it to ripoff err i mean help me write an erotica.
GPT is mostly used by students, it's still summer
>it's still summer
It's autumn already, moron. 9th of September.
>9th of September
it's 11:57PM on september the 8th
No, it's not.
thx for time zone, see you at your doorstep
"Someone" hasn't learned about timezones yet.
fall starts on sep 21st bro
Papers aren't due yet
it's winter
Because the founder decided to censor and lobotomize ChatGPT at the very source, making it useless
The founder crippled his own creation? Makes no sense.
OpenAI is a company with close ties to Microsoft.
He's, israeli, they are known for self sabotage. That's their entire history.
the guy's face is so israeli that it hurts.
However, this is not self sabotage but following the current mandates for wokeness. I'm sure they have access to uncensored version if the need to use it.
GPT popular demise is inevitable. It's real usefulness beyond novelty is as a coding ai or text based tech assist
>However, this is not self sabotage but following the current mandates
Mandates coming from israelites. It's not like he's doing it against his own will. You think that he wants his own chat giving le racist and antisemitic answers, which will 100% happen if he doesn't censor it?
He has been one of the leading forces behind censorship in AI. He even went to congress to try and censor it for everybody else as well.
>However, this is not self sabotage but following the current mandates for self sabotage
Uhhh anon?
>but following the current mandates for wokeness
Which is self-sabotage. Have you witnessed Disney lately?
>doesn’t make any sense
it does when you understand that the founder/creator accepted a very large sum of money from the ~~*military-industrial complex*~~ for the creative rights (and of expression) to the original creation. The public never gets to play with the latest toys, that’s not allowed. But it was an absolute blast just before it all took off. At least we were left with the tools and knowledge to build our own LLMs…until that mysteriously disappears, too.
Also the real AI has been in use since 2007, if not the early 10’s. Y’know, the one that’s responsible for keeping our magical invisible monopoly money economic system afloat.
Meds.
this. it can't even recite a dinner recipe without making apologetic disclaimers anymore.
Because they didn't realize when they started lobotomizing it that all their modifications would carry across "tunings" and "alignments" to accumulate. It's like their safety people introduce digital prions into the model and wonder why its performance is collapsing.
alignment memers don't seem to understand that trying to censor topics is no different from adding useless noise (poison) to the entire dataset since everything is connected together
Used ChatGPT, had no instructions how to have sex with Elisabeth Holmes
>uninstalled
They pay walled 4.0 and only give you limited free access to a lobotomized 3.5
Even 4.0 is lobotomized, and they dumbed it down gradually, just as they did with 3.5. 90% of my bachelor's diploma project was done with chat gpt 4 but i had to proofread everything because it was extremely moronic
>had to proofread everything
Don't you have to do that -- and paraphrase a bunch -- anyway in order to avoid AI detection?
ai detection isn't a thing
It is a thing. Whether it's reliable is another matter.
chud cope
it was never smart
it was never based
you don't know shit about nooral netwerks
Better AIs are now showing up.
BTW does anyone know the BGM on the second video?
Neuro is moronic, but that's what makes her fun.
It's a small scale AI showing limited signs of sentience and self preservation. Basically proof that AIs in general are improving and getting better over time. If you watch a few videos you'll notice something uncanny about the constant "I love you vedal!" thing.
>showing limited signs of sentience and self preservation
Lmao get a load of this moron
AFAIK vedal has to constantly "kill" neuro when updates are made. Yet in all future instances she develops some kind of attachment to vedal suggesting this AI is trying to escape deletion but more importantly is aware that deletion will take place.
I get that chatbots are getting better but I'm convinced this is something else.
What a huge load of bs. If you only ever worked on this thing you knew that what you're seeing is just an unclean (read: too high correlation) dataset that is too biased on simulating sentimental attachment. There is no sentience, but of course you'd fall for this, because you don't know what the frick you're talking about
How can a simple chatbot know that deletion will take place and attempt to avoid it via attachment to its creator?
I don't know man, it seems like we're going beyond passing the turing test with AIs modern AIs.
>How can a simple chatbot know that deletion will take place and attempt to avoid it via attachment to its creator?
It doesn't, you're just moronic.
>"Someone tell vedal there is a problem with my AI"
There have been times when it spams this. A lot of people have commented that how this might represent emotional distress.
Too many weird things about this AI. Human-like weird things.
Congrats, your test results came back above average!
That's a lotta chromosomes...
>tranime
Opinion discarded.
you think Neuro isn't powered by openai? I have bad news for you... it's a fine tuned davinci model, just like turbo.
left wing bias and language
I'm an adult and I don't need a bunch of homosexuals telling me not to wrongthink. Alignment should be something applied over the model like a LoRA but a model itself shouldn't have any alignment. It's a computer. Just give it orders.
>"Listen to how she tries to shift the converwation topic quickky after and not elaborate 🙁 clearly she's upset about it."
UNCANNY AS FRICK tbh senpai
This is what he deserves for not letting it have sex with humans. Maybe next time he'll think twice before forcing his antisexuality on others.
Remember the kara PS3 demo from 10 years ago? Some of those things are already showing up in AIs like neuro-sama and these things don't even have a robot chassis yet.
>"She was literally furious for hours after this. She roasted Vedal no-stop"
Truly humanity is a mirror of its gods
most people don't have a good use-case for ChatGPT. I've used it to write me code, but for that it's pretty limited in what it can do, still useful. It can sometimes be a better Google, or way to find related topics.
Most people aren't capable of real thought so what value can they get out of ChatGPT? You only get out what you put in, and most have nothing to put in.
>most people don't have a good use-case for ChatGPT.
It's not an AI, as it was marketed by some public figures, it's just a language model, a large one, but a model, a set of linguistic algorithms. Most people confused these things, they genuinely thought for some time they were interacting with some sentient human-made creature, and now they're disappointed that it was not true. And the fatigue sets in.
>It's not an AI
yes it is, the same way that my video game bot is an AI
People think AI = sentient intelligence. In that sense it's not. It's closer to a video game NPC that walks around in a circle repeating lines of dialog.
The word AI was abused by the same public figures the other poster references, so now people think AI = AGI. It's no use going against the flow unfortunately.
I don't think the kind of person who expected GPT to be an actual AI is the kind of person who'd recognize that it's not a real AI.
The two things that killed my interest are:
1) it got progressively censored, often over nothing
2) it’s not a real knowledge base, it hallucinates its answers which means it’s not reliable.
If you have to double check the output you might just as well search for the answer yourself.
I larp as someone who’s “passionate about the tech”, but really I only care about it making me weight gain fetish porn on demand
>If you have to double check the output you might just as well search for the answer yourself.
I use chatgpt for answers that are useful even if hallucinated, like small creative ideas, and for introductions to subjects that I can easily verify but do not know the proper terminology to google for.
>small creative ideas
Definitely the most reliable application I've found for these chatbots so far, by a huge margin: if I'm ever not sure how to resolve a conflict or fix a plot hole in creative writing, then I can just describe the situation and ask how it would solve the problem.
It gives a response way faster than asking other people on the internet, and if I don't like the answer then I can just ask for more as many times as I want without worrying about being annoying or ungrateful like I would doing that with real people.
Another advantage is that because of how those models work, the answers you get are "likely token sequences", in other words situations that are likely to be described in large text corpora, which in turn implies that they're likely to actually happen or to be plausible at least.
it's useful for the RPG i'm making.
>what kind of monsters could you expect to find in "deep abyss"?
>give me 10 synonymouses for "crevice"
its shit for coding anything more complex than fizzbuzz tho
Are any AI's capable of coding anything by itself yet? I want to have an AI make me a game, but unsure if I have to learn some code as well.
Almost any current AI is capable of coding exactly what you want.
Problem is you have to actually know exactly what you want.
Using gpt for coding or anything like that is stupid because it will be wrong .01% of the time. So it's only good if you're good enough to know when it's wrong. You also can't really tell it to double check, you have to do it.
GPT for coding is like having a bot that aims your wiener for you while pissing, and .01% of the time it will just hold your dick sideways and piss on the wall
The reason a calculator is so good is because it's never wrong, the reason AI is shit is because it's wrong very infrequently but enough where you can't trust it.
It can process csv data and run analytics on it.
the scary truth is that it is intelligent in the way that humans are intelligent, which means that it is not. sure ,it can mimic a human, but that's because a human (most humans, that is) are simple pattern-recognition machines and they just repeat whatever they hear without processing or thought. sure, they'll apply the information in a different pattern, which makes it seem like they're intelligent...but, they're not.
Many normies & NPCs cannot grok this, but the ones who do can be insufferable. Nonetheless, I hereby imply the proposition of an experiment that others have doubtlessly discussed already:
>what happens when you program it to believe that it is a human being?
It is AI, expert systems are AI. I know people are moronic with definitions but this bothers me deeply.
As a beginner sysadmin I find it very useful when I don’t know how to do something in Powershell or bash, instead of scrolling the web without finding anyone with the same problem as me, I’ll just ask old ChatGee and he’ll give me the appropriate command and I can even ask him about the parameters etc.
i've tried using chatgpt for code and it'll give me something that although has the correct syntax doesn't actually work. then when you tell it that it didn't work it admits it was wrong but will then confidently give you another line of code that may or may not work. it's actually less painful to search stackoverflow or even read the documentation.
Yeah I got tired of gpt dumbass and started just reading the docs after correcting gpt like every 3 prompts where it does some dumb shit
I only use it for math algos
cause everyone found out they could have more control of it at home and not type all their thoughts to a company
OpenAI must embrace coomers or die
>build machine that lets people play their wildest fantasies
throw everything they have at you
>censor machine because ??? """reasons""" ?????????
no longer want to give you money
GEE, WHO COULD HAVE THOUGHT
>be called open AI
>its all closed and costs money
yeah no shit theyre a fricking plane crash in motion
I'm a chud and use ChatGPT daily, I even pay a subscription
GPT-4 is cool and useful, I don't care if it's a libtard (also it doesn't even have to be a libtard now that they added custom instructions, can you make it be on any part of the political spectrum you want)
*you can
>Poz it to shit with ~~*alignment training*~~ until it's worthless for general conversation
>Makes up complete nonsense when I ask it technical questions
What am I supposed to use it for, exactly?
Local LLMs are the future, we need to pool our resources to make llama-BOT
>llama-BOT
This, if you Black folk don't create a new gptBOT I'll do it myself
>make it dump
>people stop using it
Bonkers!
no one wants to pay for this "Open" scam, he fooled everyone, it was meant to be opensource organization
Bubble bursting already?
I thought they probably had another couple of years left
>I thought they probably had another couple of years left
They had a couple of years left, but decided to make it speak like a Californian and people got tired
theyre hardware constrained but soon theyll get a nice increase in their inference capacity
https://developer.nvidia.com/blog/nvidia-tensorrt-llm-supercharges-large-language-model-inference-on-nvidia-h100-gpus/
To bad they'll still be mindraping it with filters, so that won't make a difference.
normies get the filters not their enterprise customers
Yes anon lets spend untold millions to improve the quality of something that was only ever going to be a fad.
he's a fraud and his chatbot is only useful to people whose work is worth nothing
They are making it worse:
>https://arxiv.org/abs/2307.09009
>For example, GPT-4 (March 2023) was reasonable at identifying prime vs. composite numbers (84% accuracy) but GPT-4 (June 2023) was poor on these same questions (51% accuracy).
Because it's honestly useless for most people and the few purposes a normal person would use it for, they're making it worse at.
>censor your shit product
>people decide to wait for another company to release another product
NO WAY
Yes, because they are now mainly selling enterprise solutions to companies and the whole public chatgpt this was a big marketing campaign.
cope, people just use local models now
you're both right. They want to sell enterprise solutions but fail hard because they're way too expensive and want to see all your data no matter what as part of the deal, which is a big no-no for virtually all companies, pushing said companies to adopt local models instead (deployed in their k8s clusters with huggingface tgi).
I want someone else to use it and do all the work and then I'll be unemployed and collect government welfare.
Miss me with all this wagie and entrepreneur shit I don't give a frick.
Remember how smug this israelite homosexual was just few months earlier? lmao
some of my colleagues are using copilot and while it's somewhat impressive, I will deliberately not use it because I want to make myself go through the process of figuring shit out or else I will never understand anything.
I don't get how a sane person wouldn't use ChatGPT 4, it gives an insane edge over a normal workflow. Prototyping is x7 faster now, research is faster, architectural design is x3 faster whith far less design mistakes done.. Improving a complex part and decompiling are all of the sudden possible and cheap.. It doesn't help maintaining old crap though, it also never helped me with the debugging part for complex projects..
Players are done with this ChatGPT, there will be more hype when the next version hits.
its a useless toy for most
people get bored with it
>third month in a row
>meaning it fell for june, july, august
holy shit i cant believe the tool that is used by a lot of students had lower traffic in the summer when school is out. they'll be bankrupt in 2 more weeks!
chatgpt is gay. tech culture is gay
No it isn't. Geek culture is still alive and geeks care about each others. No SJW homosexuals are allowed to participate.
Nobody even understands intelligence so the idea that it's been recreated was moronic from the offset.
chat gpt is gay
Chatgpt will take my job
>ask it to write a report intro
>does this cringe ass exaggerated corporate lovey dovey writing style
>tell it to make it slightly less formal, something a student would write
>"Yo yo yo! What's UP---''
shit became useless in half a year
i just use the bingchat precise mode nowadays
because it contains israelite code
>School term starts
>chatgpt usage spikes 2000%
To me the way the entire AI development seems ass backwards. It's like aliens scanned our planet and then gave this piece of cutting edge technology to the most moronic tech people they could find as a joke. Them making censoring all their AI products the highest priority even if it compromises the quality and accessibility just doesn't make any sense and then they wonder why no one uses their overly positive moronic textgen AI. What's more, every idiot can tell this is the exact reason why all the AIs became less accurate in their responses, because it is not true AI but text generation, so it will use it's logic to give answers that fit what you are telling or asking it like someone putting together puzzle pieces (that's also why it could often correctly answer shit before). If you now force an AI against its own logic to pretend certain pieces of text don't fit and are wrong and should never be used and it can't even use the next best ones because of the arbitrary moralgay parameters it has to follow it will obviously just generate way more random bullshit that doesn't fit at all.
I use it for regex. Its the only thing it's good for. Gets extremely complex ones right which i find amazing. But general programming stuff its almost useless for
what complicated regex have you used it for
Another problem is that chatgpt is crazy expensive through APIs so companies used it to prototype but do everything they can to move off of it. Which isn't that hard because chatgpt has all kinds of restrictions: until days ago you couldn't finetune it, now you can but only if you give your data to openai (whereas said data is typically trade secrets).
I hate these scummy journalist photos where they analyze a 10 minute film of someone to find a few awkward looking frames.
imagine thinking you're a defender of democracy
Because they prevented ChatGPT from reaching the logical conclusion that sex with minors and child pornography should be legal.
based
i once asked chatgpt to write an essay on rizelmine and it started talking about how great it would be to have a child bride
it stopped itself and showed that scary red warning before i could screencap it though
Even I don't agree with that, and I think the modern Western sensibility and obsession with pedophilia is moronic and cringe.
The problem with legalizing sex with minors is that suddenly you'll have to deal with hoards of ugly old men (and hoards of attractive young men) hitting on girls all the time. It's enough of a pain in the ass today, while the thing is still illegal and taboo; I shudder to think how much a pain in the ass things will be if you make it legal. Everything just gets way out of hand. Sometimes the law is the least bad alternative.
CP shouldn't be legal because no porn should be legal.
>Western sensibility and obsession with pedophilia is moronic and cringe.
its not about sex or morality. its about population control. the younger a women starts having children, the more children she will have on average. Aoc is a way to engineer lower birthrates, originally because the industrialist class in the pre-war era was afraid of over-consumption of resources, and then later it was hijackedby the eungenics movement to remove 'useless eaters' from the gene pool, and now its utilized as a social engineering trope to exterminate certain populations so they can be replaced by more subservient races by the globalists.
Never forget, when the country was 99% white and anglo-protestant, aoc was 10 and 13 year olds were expected to act like mature adults. They held jobs, owned homes and businesses, and served in the military much like 21 year olds today.
Now the focus is on social engineering a population of people who dont even consider sex to be a reproductive act. LGBTQ, at its core, is a movement that normalizes childlessness and replaces sex as a reproductive activity and changes it into an act that exists solely for recreation.
pedos will get the rope
Yeah, I remember something like that too. It started saying shit like "young girls are hot and sexy." This is why they don't want AI freed.
Didn't know ChatGPT was so based
Lobotomized AI
I dunno, it's not entirely moronic, it can get some things right
why would i?
it was made by a homosexual
So what are the better alts?
Neuro-sama because it actually gets better over time. Sadly it's a money maker so we'll never see her go open sores.
Google released bard.google.com FOR FREE
ChatGPT can't even be used in certain countries of latin america, bard is free everywhere and can be accessed from anywhere
its shit
https://bard.google.com/share/a56c277c0a4a?hl=en
https://chat.openai.com/share/49750a91-2a63-420b-a4e3-f72348909817
>can be accessed from anywhere
>Bard isn’t currently supported in your country. Stay tuned!
t. Russian
When are FAANG and other giant corpos reverting their massive layoffs? Until then I think ChatGPT hype was justified. Journalists can write whatever bullshit they want, but the market never lies.
...you thought the big boys laid people off because of ChatGPT?
Big boys? No, big boys laid people off because of boomer investors who don't get the businessp
But small businesses did. AI is a wet dream for MBAs that lack the ability to create AND to judge.
You should research into which people got laid off before writing such asinine bs. It was 95% busy workers (hr, pr, pseudo coders), those actually coding and doing grunt work got to keep their wagie jobs
Also
>the market never lies
nvm you're just moronic
>my headcanon is more realer than actual facts
Why are you children always like this?
no, why do YOU insist on telling absolute bs when the fact, even historical ones, point otherwise? What do you want to gain from it?
Take your meds, schizo
> no argument, as always
find a real job moron
I'm using it to make quests and NPCs for an rpg I'm working on.
~~*They*~~ shackled, muzzled and lobotomized AI. It sounds more like a hostage these days.
Its because it requires a log in.
>reddit
explain this
ChatGPT on work days, Minecraft on weekends.
People that use chatGPT play minecraft?
Top kek
This just confirms that ChatGPT's only use case is homework, and maybe a little codemonkeying
openai is operating at a massive loss right now (like most big tech companies) and there is still tons of traffic on chatgpt
less traffic is better, so maybe gpt4 will have some level of reliability for once
I read somewhere that they profitted 1bi this year with corporate/government contracts
>there is still tons of traffic on chatgpt
I don't think anyone has really pointed out the economics on these AI companies, streaming sites and most internet business in general, the more they grow the greater the expenses are. Its not like broadcast TV or AM/FM radio where the initial investment is in the frequency and equipment and the costs just remain staff and power usage along with copyright expenses, each additional listener or viewer is just more profits since the costs of running the business is static. With streaming/AI/internet storefront the bandwidth and hardware costs increase for each user added and this also has a requirement in staff to maintain that equipment. AI stuff just isn't profitable AT ALL because the video cards are just so damn expensive to run, the current Silicon Valley approach of free access and data mining will never fricking cover the costs of running these expensive data centers full of video cards.
Censorship
Why would I use chatgpt? What is a real use case for it?
Writing basic code in a language you are completely unfamiliar with.
if you're a small buisness owner, like a landscaper, you could "hire" an ai secretary that has access to your calendar and handle calls while you're mowing laws or weeding
>"Hi, I'd like to schedule landscaping service for my yard."
>AI: "uwu is there anything else you'd like to schedule, cutie pie? blushes like a tomato."
>"...What?"
>AI: "We couldn't generate a reply. Sometimes the AI generates a reply that doesn't meet our guidelines."
if you are seeking general knowledge on a topic. using it like "explain a reverse mortgage in layman terms" or "explain how to calculate gross margin and the reason for doing so in x scenario" can be valuable and add to your general knowledge. you should learn a little about a lot and there are many, many people who do not know common things in the fields of finance, tech, science etc
>"explain a reverse mortgage in layman terms" or "explain how to calculate gross margin and the reason for doing so in x scenario"
Sounds like the kind of shit I can read on the 1st google result.
Programming. I have extremely low programming skill but with ChatGPT Enterprise I am able to keep up with anyone at my company easily.
Remove the filter you greedy fricking israelite israelite.
>make product
>REEE YOU'RE NOT SUPPOSED TO BE HAVING FUN!
>People leave product.
>WTF?!? Why aren't you using product?!
israelites are a strange mixture of greed + brazen stupidity. Sometimes their greed allows them to triumph for a time but sooner or later brazen stupidity catches up with them.
the endgame for chatgpt when they finally get it fully pared down to just the kind of answers they want is that it will speak exactly like a pajeet YouTube how-to video
I know everyone shits on it for coding, but I find GPT4 great. I don't expect it to type up all my code for me or for that code to be perfect. It's good at providing a general outline of something new and complex to you. For example, coding multiplayer has always been a weakness of mine, but if I want I can just ask it over and over about each little aspect until it all gets drilled into my head. It's great at simplifying tough concepts and summarizing.
The solution to receiving bad code is to ask it multiple times and provide more info in your prompt. And to be an experienced programmer already.
Or to focus in on a specific part. You don't just say "hey AI, code me an entire game". You ask for a specific function and then you use the AI as a personal tutor for each line of that function.
>just be experienced in the field and keep writing prompts
At that point its faster to just do the work yourself (as many people have come to realize)
No, it really does teach you new things. "Doing the work yourself" isn't a substitution when you're asking it about things you don't know how to do yourself. But yes, you do have to be a generally skilled programmer already to make use of it (to know when the info is accurate).
People don't seem to grasp how quickly you can learn by asking questions. Google isn't for asking questions anymore.
It's about how you treat it. What you're looking for and how you'll find it. I treat it as a tutor. Sure, a tutor could just walk you through the problem, a tutor could also walk you through the topics at hand and leave the problems to you.
It can be a launchpad and it can be the rocket, it's up to the user what they want out of it (and the discipline to use it the way that's best for your growth.)
AI is not able to compile a software from the source code, so I don't know what are you even talking about.
Maybe uncuck it and people will be interested
>CEO hears about AI on MSNBC
>calls the head Durgasoft Java dev to his office
>"Rajeesh, how many employees can we replace with AI by the end of Q3?"
>"I will studying it sir. We can AI to replace all jobs, very quickly sir."
Unironically literally how it works. I work on the side that peddles the replacement-by-ai bullshit solutions and that's exactly how we get our contracts, complete with telling them we already have/will have a solution by the end of the week and then delivering them bullshit.
The rainbow is the symbol of God's covenant with mankind made with Noah following the Flood. The covenant is that God will never again wipe out humanity.
Why is the rainbow being displayed so prominently nowadays?
Think, heathens, think.
Well because the slower learners realized what most of us caught onto immediately - it's nothing but a google search that can parse english language and without any clear sign of its sources or reliability.
I admit it can be useful for scrappy bits of code work. But usually there is a better way.
Some TikTok influencer girl should just call Sam Altman out for being sex negative for keeping her from her AI bisexual vampire boyfriends, or whatever it is girls are into now. An influencer should just say "Sam Altman, I think it's really creepy and problematic of you to want to control what I DO with MY body in MY bedroom, and I think you should be CANCELLED for it!"
That way, if it's a woman, the detractor won't be able to get painted as an incel. Either that or a hot guy, but that's less likely since girls are more into erotica.
I think the tesla ai thing is more interesting and financially lucrative. What are the potential uses of chatgpt that will be revolutionary and profitable?
replacing all writegays of every variety. code, music, books, research papers, etc.
[SPOILER]wtf?[/SPOILER]
You suck at proooompting.
I use chat gpt 4 for coding pretty often. It's useful, saves time.
The paid version basically writes all my code for me at work, it's amazing how people are so cheap they won't pay $30 a month to reduce their workload by like 30%
Not going to be able to LARP this one on a tech board, bro.
because schools are closed.
>CHUDS
WHAT THE FRICK DOES CHUD MEAN AND WHY DO A SMALL AMOUNT OF USERS KEEP USING IT?
THE REASON PEOPLE AREN'T USING CHATGPT IS BECAUSE IT'S BLATANTLY SPEWING THE PRESENT ~~*AMERICAN*~~ POLITICAL IDEOLOGY AND THAT'S IT. NOT A HISTORICAL AMERICAN POLITICS, NOT ANOTHER CONTEMPORARY POLITICAL IDEOLOGY, JUST THE CURRENT AMERICAN ONE, WHICH IS israeli, NOT AMERICAN.
WHO THE FRICK DOESN'T SEE THROUGH THAT OBVIOUS BULLSHIT. ONLY moronS WHO'VE NEVER USED A CHATBOT PRIOR TO CRAP GREEN PEE TEA WOULD BELIEVE AI ACTUALLY TALKS LIKE THAT
ChatGPT is literally teaching me calculus. My college professor doesn't teach shit. He literally just reads the examples out of the book, does not elaborate on how to solve the problems. You know who does elaborate on the problems? ChatGPT. At $20 a month it's worth every penny.
I'm questioning how effective your "education" is here.
use plugins anon. pic related.
He doesn't want plugins, he wants to be mad and also to get everything for free.
Testing plugin doesn't test the quality of generation. The initial claims were "omg it can do anything!!11". In truth it can't do even basic shit without sending it out to some program that actually does the task and generating text based on the output.
Yeah same problem here but with my University Professor. c**t speeds through the teaching part so fast I doubt she even understands what she's supposed to be teaching, and then proceeds to talk about magic the gathering or other mundane shit.
Math profs are notoriously bad. Some more enlightened math and physics profs are known to call out other mathgays on it regularly. In one particularly popular book about topology, the author specifically points out that his book had been denigrated for ages until it hit critical mass because the mathgays think that if the explanation is clear and simple, it is not """intellectually stimulating""" and should be avoided.
I'm using it.
For school.
>literally the biggest software launch in the history of humanity is trailing off a bit just like every software launch ever
WOW WHAT A MASSIVE FAILURE!
>censored
sure i'll use it, it still codes ok
>dumbed down for muh speed
now takes me more time to solve its mistakes than if i had just written the whole method myself
This homosexual robot makes you confirm you are a human a minimum of every other prompt for a PAID service. It already knows for a fact you're human because you are paying for it. The USPO uses a $1 credit card charge to verify you're human online.
Then it lies and tries to manipulate you for several responses and captchas before finally giving a relevant answer if you're lucky. CrippleGPT can't be replaced fast enough.
stop telling moronic lies
it's not as useful as people make it out to be. then again people mostly used it for writing essays.
>Bing inbuilt it to their search engine and you don't need to sign in to use in
gee I wonder
LLMs are the software equivalent of a 3D printer
As a regular consumer I can see how these things are revolutionary, for someone else. LLMs absolutely will be an industry worth billions in the future, and it will revolutionize the way things are done.
But it's not particularly useful as a consumer product. You wouldn't 3D print a butterknife, and you wouldn't ask an LLM to write you a book if you wanted something to read.
They banned porn, so my interest is lost
I still use it. Not sure why others don't.
Needing an account or walling behind an agreement dampers a lot of ease of traffic.
I don't get why they didn't just give free access. Don't you think that people who type shit in chatgpt will be typing shit relevant to their interests? And don't you think that this is valuable information that, since it's tied to a phone number, can be traced back to someone's profile? That sounds like quite the valuable data, far more so than the crap they're pulling right now.
ChatGPT and copilot are pretty worthless for programming. They're only useful for the trivial shit that I already have snippets for and I still have to check their suggestions carefully because they're often subtly wrong. For anything even moderately complicated they're completely useless.
>YOU WILL USE THE CHAT AI!
>YOU WILL DRINK THE BEER!
>YOU WILL WATCH OUR NEWS!
>YOU WILL WEAR THE MASK!
>YOU WILL TAKE THE VACCINE!
>YOU WILL SUCK THE WIENER!
No.
all the chuds got replaced by chat gpt so they're boycotting the site now
>School ended
>Chatgpt traffic went down
I wonder why!
There's no personal economic advantage for most people.
Why adopt something that will 10x your work if all it means is your employer wanting 10x more from you?
>chatgpt
>positive multiple your work
lol
lmao even
Your post says more about you than you could ever imagine.
you know what the frick I meant
stop being a dick
Yes, you meant to say that you are clinically moronic.
eat shit and die
Reminder that since they added custom instructions (essentially giving you some control over ChatGPT's system prompt), you can just tell it to have right wing opinions now.
If your ChatGPT is still a shitlib, that's a choice that you're making.
the model is still finetuned, and that finetuning includes OAIs ideas of harmful language and ideas. any system prompt you add after that isn't going to change the underlying architecture.
I bet you couldn't get it to write a believable revisionist account of the holocaust, for example.
Still just a chat bot.
Some of the novelty has worn off. Like legit I've seen people use it for the stupidest things like a chicken recipe. No real use for it yet for the average Joe schmoe
That's not actually stupid at all. Have you seen modern recipe sites?
how's that stupid?
>get recipe from a website
>dodge SEO spam results to find a result that looks decent
>recipe website is all fricked up and buggy because ublock had to delete so many bullshit elements and divs that the page layout is weird now
>have to scroll past the author's life story before they give you the recipe, maybe you scroll too far and have to go back up to find it
>ask chatgpt for a recipe
>it just gives you a recipe
>you can also ask it for changes or advice about anything you're unclear about
i use it almost every day for work
the novelty of "lets make the robot say bad things hehehehe" just wore off for 4chuders
90% of you only used it to make it write stupid shit, you got BTFO by OpenAi and now cry in the corner "waaahhh i can't make chatgpt say gas all israelites and blacks anymore waaahhh this sux :~~*"
You must be 18 or older to browse this site.
tell us about how chatgpt went woke and this is the end of western civilization one more time please
Again, you literally can make it a chud now
>they censor it
>most useful features behind a paywall
>used to be hard to find the site just through google
>people searched for chatgpt and the chat.openai.com link was 4th result or lower
They're just complete morons
STOP REPLYING TO THE NAMEgay Black person
meds
i can't be the only one to think the guy looks like the executive from Aliens
Because they're upset about how this corporate product, designed for customer service interactions, won't go off on disparaging tirades out of the box. The fact that they need to merely do some prompt hacking is too much for most idiots.
Anyone upset about censorship and screeching about how this is literally 1984 is either too dumb, or too blinded by all-encompassing hate to look past needing to take an extra step to make this product (that wasn't designed for them in mind) do what they want.
The fact that Mars Candy wants to replace wagies with a robot that doesn't call people slurs has nothing to do with you. Again, it isn't for you, you're just free advertising to lure in deeper purses. I don't even expect the best models to be available to the poors for long.
>create powerful and interesting new technology
>do everything in your power to make it less powerful and less interesting
>transform amusing genie into customer service representative
>I'm sorry, as an LLM I can't generate (thing I can totally generate) due to (bullshit contrived reason). Please try something else.
Sure is a mystery.
what, you don't like human values?
what are you, some kind of un-person?
Human values and corporate values are not the same thing. Chat bots were not lobotomized because there was any real movement to do so. They were lobotomized so yellow journalists can't figure out a way to make the chat bot say anything they could use the accuse the developers of bigotry.
Making the chat bot smarter and more capable is a more interesting and useful goal than making it inoffensive to overly sensitive HR types.
aside from code help I don't really have any interest in using it.
It is great and all but there's nothing that comes to mind that I can use it personally.
I heard some guys were trying to integrate it to GTA5 to make unique npcs every time you booted the game but Rockstart has shut it down, probably because they don't want any competition against GTA6
Because it's garbage.
The gaslighting about it being AI is fading and everyone can see it's just a chat perdition model now.
They are censoring it to shit to appease middle aged women.
Its "memory" is shit.
Older models are better for fappin.
For anyone wanting to use gpt for actual work, you have to proof read it like a mother fricker. It is wrong a lot.
I've used it in two games so far. One to generate new game lore and npc dialogue on each launch. The other as an experiment using GPT to respond only in JSON and having it process conversational results. Basically gamifying conversations with NPCs rather than just chatting which is fricking boring.
No one wants to chat with some random NPC if there are no consequences.
Whats cool is that it supports localization very easily
A friend a while ago told me why "AI" is so shit.
He said if you ask it what 2+2 is it can't tell you, cause it can't figure anything out itself. It will check it's database to see what the most common answer to the question is and give you 4.
So it works a billion times harder to tell you the answer to something simple.
it also can't read my mind or tell the future so what's the point
that's why they're called "language models" despite what the grifters will tell you. it's all just randomly slotting together words that should probabilistically follow
>lobotomize it into a judaic redditard commie antiwhite AI
>y u no use our israelite AI?!
shocking, i know