OpenAI ChatGPT
Posted: Thu Dec 22, 2022 3:46 pm
My model is this is how our so-called civilization is being run today. It is currently run by AI/supernatural unseen energies/entities. My guess is they are now going to harvest manKIND on a more direct personal individual level. And it's not to sculpt ads to make "money" with....
From the comments
Seeing the reactions to ChatGPT, I know that GPT-4 is going to shake the very foundations of civilization.
Well we are the training programme, we are literally training it for free right now - and yes it actually does remember every single thing it sees. We will be shown more parlour tricks like this WebChat module. But the real genius isn't that version at all...
The rate of progress is incredible!
It's been out for 3 weeks and after just 2 weeks, it has already been updated and the current version is noticeably better!
It's so good at writing songs, apparently, it even knows about rhymes and proper rhythm!
I got it to first give me a D&D oneshot idea, which it then fully fleshed out when I asked for details, and then I made it my DM for playing a little bit of DND and it even knew rules and methods, and was able to improvise!
This is just mind blowing. Works and feels like magic
I tried to get it to create an argument supporting a conspiracy theory I saw and it said it shouldn't be done for ethical reasons but it would make an argument for why it was likely not true.
ChatGPT is quickly becoming almost as central to my daily internet usage as Google. Way better at answering complex or abstract questions.
That’s kinda scary.
ChatGPT is literally as good as AI chatting should get...
It's becoming scarily better than humans
Extrapolating the speed of improvement a few more papers down the line Chat GPT will be a superhuman.
I am personally not that surprised that it can write essays or code, which have both been accomplished by GPT-3 and Copilot. But I was genuinely shocked when I fed it several multiple choice questions from one of my finals and it actually solved them, heck it performed better than I did. The questions I chose weren’t even the math ones with definitive answers, but from a literature class like “which one of the following is not a theme in author X’s book….” It really feels like the AI actually “understands” rather than simply responds to prompts.
From the comments
Seeing the reactions to ChatGPT, I know that GPT-4 is going to shake the very foundations of civilization.
Well we are the training programme, we are literally training it for free right now - and yes it actually does remember every single thing it sees. We will be shown more parlour tricks like this WebChat module. But the real genius isn't that version at all...
The rate of progress is incredible!
It's been out for 3 weeks and after just 2 weeks, it has already been updated and the current version is noticeably better!
It's so good at writing songs, apparently, it even knows about rhymes and proper rhythm!
I got it to first give me a D&D oneshot idea, which it then fully fleshed out when I asked for details, and then I made it my DM for playing a little bit of DND and it even knew rules and methods, and was able to improvise!
This is just mind blowing. Works and feels like magic
I tried to get it to create an argument supporting a conspiracy theory I saw and it said it shouldn't be done for ethical reasons but it would make an argument for why it was likely not true.
ChatGPT is quickly becoming almost as central to my daily internet usage as Google. Way better at answering complex or abstract questions.
That’s kinda scary.
ChatGPT is literally as good as AI chatting should get...
It's becoming scarily better than humans
Extrapolating the speed of improvement a few more papers down the line Chat GPT will be a superhuman.
I am personally not that surprised that it can write essays or code, which have both been accomplished by GPT-3 and Copilot. But I was genuinely shocked when I fed it several multiple choice questions from one of my finals and it actually solved them, heck it performed better than I did. The questions I chose weren’t even the math ones with definitive answers, but from a literature class like “which one of the following is not a theme in author X’s book….” It really feels like the AI actually “understands” rather than simply responds to prompts.