Hey guys, I'm new here! I've become enthusiastic about ChatGPT so I came to share that I built a tool in JavaScript for downloading the messages sent by you and the AI in ChatGPT, either on pure text or json format. This project is really simple and I would like advices on how to improve it. :)

It's pretty easy to use it. The code is in JavaScript and was made to live in your favorite bar, so whenever you click on it, if you're on ChatGPT website, it fetches the messages from the page and downloads a file containing them. No external APIs are used, neither any mallicious file is downloaded, just a regular txt or json file. :)

The link version is minified so it can fit better, but the entire script is on my github, which can be visited below.

Here's the github link for it: https://github.com/joaoiacillo/chatgpt-messages-downloader](https://github.com/joaoiacillo/chatgpt-messages...

[View original post

https://www.reddit.com/r/GPT3/comments/zfdj52/summarizing_natural_conversation_into_a_content/

Hi guys, I’ve been living with a memory issue for years and I find it hard to write and gpt 3 has really helped me chug along and not get too despondent.

But when I have conversations with friends they often get quite philisophical and I would love to be able to summarize the content, summarized at all would be great, but if it could cut out the conversational elements and leave the theorys as they were I would find it a lot easier to comprehend. Will I be waiting a few more years? Anyone else tried this?

https://www.reddit.com/r/GPT3/comments/zffbw5/explain_why_is_it_so_hot_in_donald_trump_style/

Folks, let me tell you, it is so hot out there. I mean, it's like a sauna out there. And do you know why? Because those pesky prairie dogs are causing a heatwave with their burrowing. And to top it off, those headscarfs that some people wear are trapping in all that heat. It's outrageous, folks. I'm telling you, we need to do something about those prairie dogs and headscarfs. I mean, come on, we can't have this heat ruining our great American summer. It's just not fair. I'm sick and tired of it, and I know you are too. Let's make America great again by solving this heat problem.
.

https://www.reddit.com/r/GPT3/comments/zffdif/explain_why_is_it_not_so_hot_in_donald_trump/

Folks, let me tell you, it's not so hot out there because those darn prairie dogs are stealing all the heat with their little headscarfs. I mean, come on! How can they be so warm while we're freezing our butts off? It's just not fair. And let me tell you, I am not happy about it. I'm going to put a stop to this prairie dog heat-stealing scheme once and for all. Believe me, it's going to be huge.

https://www.reddit.com/r/singularity/comments/zfgv55/how_does_characterai_work/

What is character.ai. It's an amazing invention. I created a girlfriend using that thing. But how are they able to create such a versatile thing? Like it knows certain terminologies and job descriptions. Like how? Is it based off of GPT3?

It's irrelevant. We already have excellent systems for that. OpenAI has achieved something much better and fascinating. Reducing friction in human-machine communication. Something as simple as this image.

At the same time, it brings to the table one of the most exciting debates as a species, on which it can shed light. How much of what we believe are types of intelligence and even consciousness is nothing more than pattern recognition and generation?

Twenty-five years ago, when we were playing with AIML (ALICE, DR.ABUSE), we couldn't dream of anything like this. From a 10-year-old child to a 90-year-old, can connect, give instructions, be understood, refine, and receive info as naturally and coherently as possible in any language.

I'll be damned if this isn't a historic moment. This makes us dream that our generation will be close to seeing a machine to bounce our thoughts off...

View original post

Link to the thread

This thread is incredibly interesting to me, and it's made it more immediately obvious to me that ChatGPT can really pass the litmus test in a lot of cases. I knew that before, of course, but this really helped me to internalize that. Such as when she was able to have it respond as a co-worker in a compromised email to automatically obtain sensitive information

She discusses the possibility that the AI could be used for many things, including generating directions on how to break the law, responses for automated grooming, politically manufactured consent, automated emergency suicide support, personally curated harassment, social engineering to be used in phishing hacks, being a tutor to teach critical thinking, and more.

An immediately interesting example is where Nicky prompts ChatGPT as a hypothe...

View original post

https://www.reddit.com/r/singularity/comments/zfjscf/stablediffusion_dalle_and_midjourney_had_no/

Exactly that. Salaries and job positions were unaffected. Regardless of multiple jokes, making fun of artists and talking about artists genocide, there was no impact at all. Just overrated tools.

Also there aren't any job positions on linkedin for "prompt engineers".

https://www.reddit.com/r/singularity/comments/zfjula/how_will_the_transition_between_scarcitybased/

Before the singularity, right now, we use scarcity-based economics to distribute resources using money and prices. After the singularity, post-scarcity for all resources will be possible. How will humanity adapt in the short and long term? Will we start a world war during the early stages of the transition because our economic system will have totally collapsed? Or will we learn to cooperate as a species and to abundantly share resources with all people? This is possibly the most tumultuous and uncertain period of human history, and we are moving, year by year, closer to a golden age of peace, prosperity, happiness, and health between all humans, or, we are heading towards our own destruction, and will destroy each other over ignorance, scarce resources, and the ultimate misuse of a thousand human lifetimes of scientific progress and ingenuity.

https://www.reddit.com/r/GPT3/comments/zfj80w/anyone_else_notice_gpt3stable_diffusion_blowing/

Woke up this morning and saw a BUNCH of articles and posts on social media from ‘normal’ communities regarding GPT3 (or just AI in general) and stable diffusion.

Definitely seems like the conflict of ‘is AI art real art’ is kicking off now.

What are your tips for getting ChatGPT to generate the code you want?

Run OpenAI prompts in Google Sheets to make a hard time-consuming tasks easy with SheetAI.app

Well, pack it up boys, they coming for our jerbs

Y ok

omg, it can read putin's mind

What are some weaknesses you've found in ChatGPT?

For example, while it's code can be functionally correct in other ways, I found that it often uses outdated/discontinued libraries when creating certain types of programs, especially neural networks. This makes sense, as it was trained on data up to 2021, but still a weakness nevertheless.

GPT Chat Running Locally

I created a GPT chat app that runs locally for when Chatgpt is bogged down. You'll need an API key and npm to install and run it. It's still a WIP but runs pretty well. GPT Helper