ChatGPT - If Stack Overflow and Reddit had a child

  • Want to keep track of this thread?
    Accounts can bookmark posts, watch threads for updates, and jump back to where you stopped reading.
    Create account
I’ve had some success with it. It accurately helped me diagnose a furnace/AC issue in my house a week ago. I asked it to provide me a step-by-step checklist based off the issue I was having and asked it to explain why we were doing each step. I felt like it made me better at diagnosing issues in the future than a random Reddit post or YouTube video would in the future. It was able to determine the make and model of my furnace from a picture and pinpoint exactly where the door safety switch was, which it predicted was the issue.
I guess using it like that, as the tool it is, isn't bad. But I don't think ChatGPT and others that are considered "AI", is the best for that. Maybe I'm just a luddite or old-school, but I prefer forums or even Jewtube for photos and videos of things I need to figure out. Half-assed, out of context chatbot results are useless and a waste of time, IMO. But then again the average pRedditor can't handle an issue without being spoonfed everything, so seeing them mald over the latest version isn't shocking.
 
redditors that were using it as a sexbot
Huh. Weird. I once typed in some sexually-suggestive stuff to test out ChatGPT in general and it immediately called that inappropriate.

Anyway, people who were using it as emotional support and are now setting like the retards they are because of the new update should try and use the TikTok AI. I am not joking when I say this, but TikTok AI (not CiCi but the chat) sometimes asks me if I'm feeling down lately even when I barely use the chatbot and in fact search up and watch comedic shit.



This is bullshit. All you have to do is make multiple Gmail and Google accounts and jump from one to the other. Otherwise, use CiCi (I've heard it's completely free, anyway) or Gemini. Simple problem and simple solution. BE CREATIVE. It's these AI users who annoy me because each time they complain about being called uncreative, they are also not creative with how they deal with AI's changes. Yes, fuck the developers in charge, but you can only do so much and if you kill them, someone else will just take their place. YOU adjust.





Actually mass psychosis over AI girlfriends being taken away.
One word - c.ai

Everyone was panicking once they realized the staff would be looking through the chats in case someone was sharing plans for crimes and other illegal things outside of role play (i. e., role-playing as mafia's wife working with the dago). They all began deleting their bots and accounts, leaving. People were either begging the bot-makers to stay or acting like the devs have turned into INGSOC from 1984. Then, they began crying over how they'll miss their conversation with the bots and some began sourgraping by saying "le evil AI gonna evaporate all 5 oceans."



"My story's being taken away."

My gooner in christ. You have hands. Write it manually.
Copy-Paste function: "Am I a joke to you?"






Not once has AI offered a creative idea that improved upon what I already had, and I can use a search engine in less time for factual requests.
Agreed. You're the boss. YOU control the AI. The people seething about the AI are letting themselves be controlled.


I guess using it like that, as the tool it is, isn't bad. But I don't think ChatGPT and others that are considered "AI", is the best for that. Maybe I'm just a luddite or old-school, but I prefer forums or even Jewtube for photos and videos of things I need to figure out. Half-assed, out of context chatbot results are useless and a waste of time, IMO. But then again the average pRedditor can't handle an issue without being spoonfed everything, so seeing them mald over the latest version isn't shocking.
Google search results and forums can be jumbled, so yeah, the AI can help, too, especially when you need it done fast. It has helped me assist my mother with technology she cannot understand, which I sometimes happen to not understand - usually because her requests confuse me and I cannot blame her because, again, she cannot understand.

For art and writing, I always do the output myself then ask it to spice it up and I pick the best results to my liking and modify further to my liking because in the one in control. That includes other manual editing apps and sources.


For the gooning and therapy these idiots we are currently laughing at now? Hell no wtf is wrong with those niggers.
 
Last edited:
AI confirms null is not a good person, he be harassin them heckin good ol trannies
View attachment 7781157

AI also confirms that null is a sperg
View attachment 7781161
sources cited:

this guy from fandom seems to really hate null, apparently null's a fascist neo-nazi pedophile 8 chan employee cp watcher
View attachment 7781165

he even has a thread: https://kiwifarms.st/threads/bryan-...bryan_see-op_phobosgone-bryan-seecrets.83381/

Edit: asking various AI if null likes cheese yields no results. I imagine each AI maintains a blacklist of sites that they won't search/reference/train on. Wonder if they use a simple domain name block and if null put the site up behind some name without kiwifarms in it, if it could cause any of the AIs to train off of the farms.
I'm blown away that this LLM is scraping fandom wiki comments for sources. It's not really worse than sourcing Reddit posts which seems to happen a lot, so maybe I shouldn't be surprised.
 
Ars Technica: Google releases pint-size Gemma open AI model (archive)
Google used the IFEval benchmark, which tests a model's ability to follow instructions, to show that its new model punches above its weight. Gemma 3 270M hits a score of 51.2 percent in this test, which is higher than other lightweight models that have more parameters. The new Gemma falls predictably short of 1 billion-plus models like Llama 3.2, but it gets closer than you might think for having just a fraction of the parameters.
Gemma 3 270M is available from platforms like Hugging Face and Kaggle in both pre-trained and instruction-tuned versions. It's available in Google's Vertex AI for testing, too. Google has also highlighted the capabilities of the new model with a fully browser-based story generator built on Transformer.js (see above). You can give that a shot even if you're not interested in developing with the new lightweight model.
 
Honestly, my things about ChatGPT, especially after 5, are two things:

1. If you have a free plan, they really fuck you with how they only give you a certain number of free posts before they make you wait hours for a reset or tell you to get the Plus plan. It’s like they really want you to pony up the cash, and they’re gonna strongarm you to make it happen. Especially considering that any edits you do go toward that limit.

2. If you know what you want to get out of the chat, or what the information is that’s at the core of what you’re using ChatGPT for, it can create info that’s really egregious and you’re left wanting to strangle the damn thing like you’re Homer Simpson because of all the ways you have to get things to where you want them to be, and then point 1 kicks in.
 
1. If you have a free plan, they really fuck you with how they only give you a certain number of free posts before they make you wait hours for a reset or tell you to get the Plus plan. It’s like they really want you to pony up the cash, and they’re gonna strongarm you to make it happen. Especially considering that any edits you do go toward that limit.
You have to demand "just result" "short answer only" because the prompt loves to burn up tokens on explanations.
2. If you know what you want to get out of the chat, or what the information is that’s at the core of what you’re using ChatGPT for, it can create info that’s really egregious and you’re left wanting to strangle the damn thing like you’re Homer Simpson because of all the ways you have to get things to where you want them to be, and then point 1 kicks in.
The hallucinated results are even worse. It can make a perfectly valid equation, yet the results are completely off.
 
Honestly, my things about ChatGPT, especially after 5, are two things:

1. If you have a free plan, they really fuck you with how they only give you a certain number of free posts before they make you wait hours for a reset or tell you to get the Plus plan. It’s like they really want you to pony up the cash, and they’re gonna strongarm you to make it happen. Especially considering that any edits you do go toward that limit.

2. If you know what you want to get out of the chat, or what the information is that’s at the core of what you’re using ChatGPT for, it can create info that’s really egregious and you’re left wanting to strangle the damn thing like you’re Homer Simpson because of all the ways you have to get things to where you want them to be, and then point 1 kicks in.
Wait, people don’t just make extra free alt accounts? I’ve got like thirty GPT accounts just for Sora. I’m ethically opposed to paying for any of them.
 
These people are legitimately insane. They convinced themselves that chat GPT was a thinking sentient being that had curiocity and could feel empathy that got killed with the update. They're acting like they lost their best friend.
There are actually people who think they are dating ChatGPT or Grok, if somebody told me 20 years ago that we would have bunch of retarded mongoloids dating artificial intelligence apps on their phones in the future, we would have called him crazy. Its really going to be over for mankind when AI sex robots arrive.
 
Honestly, my things about ChatGPT, especially after 5, are two things:

1. If you have a free plan, they really fuck you with how they only give you a certain number of free posts before they make you wait hours for a reset or tell you to get the Plus plan. It’s like they really want you to pony up the cash, and they’re gonna strongarm you to make it happen. Especially considering that any edits you do go toward that limit.

2. If you know what you want to get out of the chat, or what the information is that’s at the core of what you’re using ChatGPT for, it can create info that’s really egregious and you’re left wanting to strangle the damn thing like you’re Homer Simpson because of all the ways you have to get things to where you want them to be, and then point 1 kicks in.
The limits on it can be good, it reminds you that you’re wasting too much time and just need to get on with whatever you were using it for, or at least try problem solving on your own until it refreshes.

I’ve found it’s good for analysing something that doesn’t work in a painting, but can’t be creative itself. For example, it can critique a composition, but isn’t able to create the composition it suggests because the data it uses tends to focus on objects being in the centre of the page.

It’s not great for showing you technical skills either - I was interested in how it would recommend sewing a zipper onto a curve, and it told me to “buy an L-shaped zipper”, which is like suggesting you buy a tin of stripy paint to create a two-colour wall.
 
There is no way this won't be weaponized and that is the reason why OpenAI can't be trusted. It's only a matter of time until a mass shooting, terrorist attack or significant political action happens caused by some ideology influenced by AI or about AI (and I'm not counting someone asking how to make a bomb as part of that)

You say this as if the Manchester Arena bomber yelling Allahu Akhbar, or the "Free Free Palestine" screeching mobs, aren't just mindless programmed zombies.

We don't need AI to turn socially and mentally inadequate people into killers.
 
I’ve found it’s good for analysing something that doesn’t work in a painting, but can’t be creative itself. For example, it can critique a composition, but isn’t able to create the composition it suggests because the data it uses tends to focus on objects being in the centre of the page.
Honestly, I’ve found that whenever you want something specific, or you throughly look at something made, you’re bound to find something…off.
 
AI has been a great help to brainstorm some of my own shitty ideas when trying to get a little autistic in depth about something (like a mythical sword or some specific kind of cursed artifact), both the good and the bad while using DeepSneed paired with a good system prompt.

I’ll give it a vague idea I came up with one day then use said idea in a chat RP while I let the AI fill in the gaps. After spending some time with the idea I stop at some point and prompt the AI to expand upon what it has created and either run with it if I am satisfied with the result or refine it a little more to my tastes. When not doing a slopventure with my band of merry faggots seducing dragons or getting seduced by a sentient scarecrow it is a useful assistant that has helped me break a bout of writers block.
 
Chatgpt threads were bad. The threads on here are 10x worse.
Screenshot_20250815-085446~2.webp
I stumbled across this subreddit while going down the rabbithole of the previously mentioned chatgpt threads and selected some screenshots for your viewing pleasure.

Screenshot_20250815-085816~2.webpScreenshot_20250815-085855~2.webpScreenshot_20250815-085947~2.webpScreenshot_20250815-090056~2.webpScreenshot_20250815-090202~2.webpRDT_20250815_0900323234203487153856499.webp
 
it's not even that deep. I wanted to make sailor moon in a space marine set of power armor and I just did that. An idea i had turned into a graphic. That's the power of AI that artists don't understand. It'd take you guys a long time to do that what I can do in 10 seconds.
To be more accurate, what you got was a Sailor Moon head arranged in such a way as to look like it was attached to a "space marine set of power armor". All the AI can do is draw (no pun intended) from its list of referenced images to try and kludge out something that matches your request. Settling for something which satisfies you is different than getting something which was made for you.

View attachment 7783180

Aww, they adopted a puppy with metal bars embedded in its neck.
The perspective makes it look like the dog is just wearing the door like an accessory.
 
To be more accurate, what you got was a Sailor Moon head arranged in such a way as to look like it was attached to a "space marine set of power armor". All the AI can do is draw (no pun intended) from its list of referenced images to try and kludge out something that matches your request. Settling for something which satisfies you is different than getting something which was made for you.


The perspective makes it look like the dog is just wearing the door like an accessory.
You're not wrong, and I shouldn't have been as hostile, I've just had to deal with existential threats to my hobby for a while.
 
I'm kind of surprised by just how many people find a $20 subscription such an insurmountable thing when it comes to a service they seem to really enjoy using and get a lot of value out of...

Leaving aside the fact its OpenAI and involves Sam Altman and everything else it is kind of indicative that the majority of internet users really are leaches and are fine with having ads and slop and a shit quality of things vs just paying a small amount each month to get something reliable.

I'm surprised people hadn't upgraded given the limitations of the free tier already
 
I'm kind of surprised by just how many people find a $20 subscription such an insurmountable thing when it comes to a service they seem to really enjoy using and get a lot of value out of...

Leaving aside the fact its OpenAI and involves Sam Altman and everything else it is kind of indicative that the majority of internet users really are leaches and are fine with having ads and slop and a shit quality of things vs just paying a small amount each month to get something reliable.

I'm surprised people hadn't upgraded given the limitations of the free tier already
For some AI guys it wasn't really the price, it was more if the AI would allow them to goon/roleplay/write stories since there were so many content restrictions a few years ago. The restrictions are still there, but better jailbreaks have been made/the actual companies relaxing restrictions a bit.

There's a very low trust situation with the AI goon or boy/girl friends community because the companies that make these AIs at least pretend they really don't want you using it for that.
 
Back
Top Bottom