CHATGPT ???

[Different]

How often do you use ChatGPT? An infinite AI learning machine deliberately designed to replace human cognition while our services to humanity gradually phase into the process of becoming eradicated from the matrix. Dare I also say an unfathomably powerful, AI tool with the capabilities of giving us content creation, writing assistance, answering questions, brainstorming, programming help, language translation, etc. Does all of this sound immensely attractive? Yes. But the bad news is that it’s predominantly here to place you under servitude the minute you subconsciously exchange critical thinking skills for AI dependency.

At this point you might ask, ‘how am I under servitude with ChatGPT? I’m fully cognizant of when to turn my critical thinking skills on. I don’t need AI for everything.’ A multitude of people said something similar to this with social media when they tried to refrain themselves from the addictiveness of it all. With relentless companies like Tik Tok, Instagram, Facebook, Twitter, and YouTube, there is no easy off button. So with ChatGPT it becomes an addiction once you possess its unlimited knowledge with unfathomable capabilities all designed to serve you at your fingertips. In reality, you’re serving AI because predominantly it’s designed to replace humans in the job market world, and all they need is for you to help them build a database on you to see what your interests are. ChatGPT is just the surface scratcher.

How many of your loved ones have decided to surrender critical thinking skills in exchange for a learning machine that’s indefatigable? What happens when we’re heavily dependent on AI to constantly think for us? What happens when we tap out of our own cognitive function and seek AI for things like financial advice, cooking advice, relationships, fitness, and other things of that nature. Cognitive impairment, that’s what happens. There’s nothing wrong with using ChatGPT sometimes. But don’t use it more than 50% of the time because then that tells me that you’re no longer capable of critical thinking.

[Master Raiden]

I find myself using ChatGPT more and more for getting information, but that's the extent to which I use AI.

I'm not really a fan of AI-generated content, so I don't ask it to create things for me, but I find it extremely helpful when I need some information about something. Its ability to quickly pull together information, summarize it and offer opinions and suggestions makes it like a supercharged search engine. The struggles of Googling something and not being able to find the answer you need are in the past when AI can give you information on pretty much anything you can think of.

But of course it's important to know its weaknesses. On a few occasions, I have caught ChatGPT making stuff up and giving me wrong answers. So when I ask it to solve a problem for me, I also follow up with a question asking why their solution is good. You can still benefit from the power of AI without losing your critical thinking skills if you remember not to blindly trust whatever ChatGPT tells you, and that's what I try to do.

[Different]

I find myself using ChatGPT more and more for getting information, but that's the extent to which I use AI.

I'm not really a fan of AI-generated content, so I don't ask it to create things for me, but I find it extremely helpful when I need some information about something. Its ability to quickly pull together information, summarize it and offer opinions and suggestions makes it like a supercharged search engine. The struggles of Googling something and not being able to find the answer you need are in the past when AI can give you information on pretty much anything you can think of.

AI-generated content has reached the point now, where they’re dominating everything. To me it’s amusing for things like funny videos and all, but some of this stuff is too excessive. With technology apps like OpenAI and Sora, you’d have to really be paying attention to not fall for some of that stuff. On the other hand, ChatGPT is very useful for things like learning new vocabulary words, writing resumes (which I’ve done and had recruiters reach out to me before), and paraphrasing information.

But of course it's important to know its weaknesses. On a few occasions, I have caught ChatGPT making stuff up and giving me wrong answers. So when I ask it to solve a problem for me, I also follow up with a question asking why their solution is good. You can still benefit from the power of AI without losing your critical thinking skills if you remember not to blindly trust whatever ChatGPT tells you, and that's what I try to do.

Oh yeah, you definitely gotta fact check ChatGPT because their answers are not always consistently the same, either. For me, when I suspect that ChatGPT is making some mistakes in some of their answers, I’ll just do some basic research to see if it validates its answers. If I don’t like the answers they’re giving me, then I’ll just restructure my questions to see if I can get a more, valid answer. Sometimes I’ll use DeepAI to see if it gives me the same answers as well. I’d say ChatGPT is effective about 70-75% of the time, but I wouldn’t rely on it to give me knowledge on things that are common sense.

[dotime]

I sometimes use Chatgpt for grammar checking(to see what I'm writing down make sense...) and I also use it for explaining stories I made up on the spot other than that I dont really use it much at all. I like drawing my own stuff

[Different]

I something use Chatgpt for grammar checking(to see what I'm writing down make sense...) and I also use it for explaining stories I made up on the spot other than that I dont really use it much at all. I like drawing my own stuff

Sounds interesting. Yeah a lot of people are using ChatGPT to draw stuff for them. From a business stand point of view, I can understand why they’re doing it (perhaps to make money). But I wouldn’t rely on it all the time. They’re just certain things that a human can do that AI can’t do.

[Sunlight123]

I treat ChatGPT like a search engine that I can ask more freeform and natural questions.

You can't ask questions like these on Google:

1. "What was that videogame where it's like an RPG FPS where you're in space and there's aliens destroying the space station?
   
2. "Hey what was that one movie where you had a girl fall in love with a schizophrenic man? It had a bad ending."
   
3. "Uhh what's the one board game that's similar to Space Station 13 where you're on a station and everyone can influence any character on the board. I think there was like a monkey character and a rat character and then a bunch of boring characters like an engineer and stuff? It's like a mystery game where you guess who's who even though everyone can control everyone."
   

You could ask these kinds of questions on Reddit or some other mega-forum, but it's not a good use of anybody's time. And you'd have to go to specific forums, like the videogame forum, the movie forum, and the board game forum, to know the answer to. That's what I think ChatGPT is great at.

And of course you could ask these stupid kinds of questions for more academic domains. If you forget a chemical molecule you can ask open-ended questions to remember it. Or if you forgot a programming design pattern ChatGPT can guide you back to it. Yeah it's definitely flawed, but asking a human being would be equally as flawed, if not worse.

Also, it is particularly useful for fiction research, as in for writing fictional stories, since fiction doesn't have to be factual. Good fiction is about what "feels right" instead of what "is right" or "provably right". That's one of the points of fiction - To resolve things you will never be able to resolve in real life. In some cases, you will never be able to know if someone was a good person or a bad person, or if an organization did corrupt actions and covered it up, but fiction can give you a confident answer. Sometimes that's what ChatGPT is good at - Giving you confident BS that works well enough for a piece of fiction.

---

For content generation I really hope AI content creation doesn't take over in my lifetime, but I have no idea how fast AI will progress. I personally do not think it will happen in the next few decades.

I've played with AI for content generation for fun, never posting any of it anywhere, and it is a bit alarming how good it can be if you know how to use it right and which ones are good. The good news, for people who hate AI, is that right now it is impossible to get exactly what you want from an AI, forcing a person to draw it themselves to get exactly what they want.

Like right now, the big problem is that many companies can't take art from good artists legally to train their datasets off. There are websites harvesting information legally yea, but they are probably over-harvesting and it's not specific/filtered enough yet to be relevant. So what ends up happening is instead hobbyist AI models have taken over. Individuals, instead of businesses, are taking art from Boorus or artist's portfolios directly and training the models on their own computers and uploading them to Civtai or other websites. That way it's much harder to sue anyone because there's thousands of individual people instead of a few big corporations. It's more decentralized, like how pirating works, so it will never disappear.

But the good news is that these datasets don't have enough keywords to make it worthwhile. Like if you go to Danbooru there aren't enough keywords per picture to be specific about things. It might have "Short Hair" or "Long Hair", but it doesn't have specific haircuts like "Mullet", "Bob Cut", "Bowl Cut", "Butch Cut", "Buzz Cut", etc. Since AIs are trained off of these boorus, the AI also will not know the terms: "Mullet", "Bob Cut", "Bowl Cut", "Butch Cut", "Buzz Cut", etc. So trying to get the AI will never draw EXACTLY what you want because it is impossible to specify with enough accuracy. It can only generalize and guess at what you want.

[famknight]

I mostly use it to style my react websites (can't really bother to type out the CSS myself lol) and sometimes for debugging. LLMs like ChatGPT are prone to AI Hallucinations so I don't rely on it completely. Nonetheless, it is still a very helpful tool and can speed up a lot of your work with accuracy as well (as long as you don't completely rely on it to do everything for you).

[Different]

I treat ChatGPT like a search engine that I can ask more freeform and natural questions.

You can't ask questions like these on Google:

1. "What was that videogame where it's like an RPG FPS where you're in space and there's aliens destroying the space station?
   
2. "Hey what was that one movie where you had a girl fall in love with a schizophrenic man? It had a bad ending."
   
3. "Uhh what's the one board game that's similar to Space Station 13 where you're on a station and everyone can influence any character on the board. I think there was like a monkey character and a rat character and then a bunch of boring characters like an engineer and stuff? It's like a mystery game where you guess who's who even though everyone can control everyone."
   

You could ask these kinds of questions on Reddit or some other mega-forum, but it's not a good use of anybody's time. And you'd have to go to specific forums, like the videogame forum, the movie forum, and the board game forum, to know the answer to. That's what I think ChatGPT is great at.

And of course you could ask these stupid kinds of questions for more academic domains. If you forget a chemical molecule you can ask open-ended questions to remember it. Or if you forgot a programming design pattern ChatGPT can guide you back to it. Yeah it's definitely flawed, but asking a human being would be equally as flawed, if not worse.

Interesting questions. These are very broad and astute questions that I too would reserve for an app like ChatGPT because Google can be limited and make mistakes, sometimes. Now with Reddit, every now and then you’ll get really lucky and run into a perspicacious gamer who can answer all your video game-related questions or whatever it is you’d like to know. At the end of the day, I’d fact check both responses to see which one logically makes more sense compared to the other one.

For content generation I really hope AI content creation doesn't take over in my lifetime, but I have no idea how fast AI will progress. I personally do not think it will happen in the next few decades.

I've played with AI for content generation for fun, never posting any of it anywhere, and it is a bit alarming how good it can be if you know how to use it right and which ones are good. The good news, for people who hate AI, is that right now it is impossible to get exactly what you want from an AI, forcing a person to draw it themselves to get exactly what they want.

With apps like OpenAI developing Sora, I wouldn’t be surprised if AI content were to takeover in less than 5 years by 2030. I think what’s happened is that people have taken advantage of something that seemed obscure to non-tech guys, and just fell in love with it. Dude, the videos are just getting out of control, right now. I see way too many AI-generated people pop up in YouTube videos that it’s just preposterous. AI actress, Tilly Norwood is preposterous because it looks too much like a real person.

At first I thought they really need to chill out with this stuff because it’s getting out of hand. But then a lightbulb went off in my head with dollar signs hovering around it, which made me think that I could do this too. With AI art content, I’m not a fan of it either, but if there’s money to be made then I’ll do it rather than paying someone else to do it. Yeah it’s taking shortcuts, but it saves time and money, and it has an aesthetic appeal to a certain type of audience.

[yotoprules]

I use it very occasionally. It's pretty useless for getting information from as it always hallucinates and gets everything wrong, more times than it gets it right. To me, it's just a waste of time. there are specific use cases for it though, but I haven't used it in months.

[Legend]

In my opinion, it's only really good for a few things:

1. Open ended questions where there are no right or wrong answers.
2. Translations - LLM stands for Large Language Model and translations are one of their strong points. I've used it to create Excel formulas by describing what I want to calculate and it outputs working formulas. The reason this works well is because it's being prompted to essentially translate English into Excel syntax.

I don't recommend using it for research or learning because of the hallucination issue. In fact, I would suggest to anyone reading this to ask AI questions that you already know the answers to. You may be surprised by how often it confidently outputs incorrect information.

[Different]

This is why you never depend on ChatGPT for everything. Lmfao, please stick around til the end ?