41.7K subscribers
5.53K photos
232 videos
5 files
917 links
🤖 Welcome to the ChatGPT telegram channel! Here, we post the latest news, updates, and examples of using the ChatGPT large language model for generating human-like text in conversations. Subscribe to stay up-to-date and learn more about its capabilities.
Download Telegram
🦖 🧮
😁21👍2🤣21🤔1
Whisper to ChatGPT: A Chrome Extension that Transcribes Your Voice to Text with OpenAI's Whisper Api

(Need to input your own API key)

Whisper to ChatGPT
👍6😐21🤡1
ChatGPT can display images inline using markdown image formatting

Though usually the URLs don’t point to anything, so you don’t see an image
🤯8👍3🤔21😁1
ChatGPT used to get raise for nursing department

“The email essentially proposed that our facility is experiencing difficulty attracting/retaining talented nursing staff, has too rapid a turnover of nurses, and is being negatively affected by low moral. The proposal stated that a 10% wage increase could help alleviate these problems.
It stated that the increase would likely be cost-effective because talented staff with high moral would result in increased productivity, that less time and money would be spent training replacements, and that the higher level of care that our well trained & talented staff would be able to provide would increase the reputation of the facility and possibly leading to more referrals.”

Prompting:
“write a proposal to increase nurses wages at a [type of facility that you work at.]” it’ll give you a proposal that will be more helpful than the one I sent would be.
If it’s worded in a tone that feels wrong, you can just tell it, “make this sound more professional” or “simplify this” or prompt “write an email” rather than “write a proposal.”
👍151
Bing AI: What the Founding Fathers would say about the KFC Double Down Sandwich
👍61
Artist Apocalypse

The problem, one Avery struggled to disclose to his 28,000 followers, was that he was creating the images using Midjourney, an A.I. image generator.

The popularity of the images, which he began posting on Instagram in October 2022, had wildly exceeded his expectations, creating an uneasiness that led him to contact Ars Technica, a technology publication. “Probably 95 percent-plus of the followers don’t realize [the images are A.I.-generated],” he told Ars Technica, “I’d like to come clean.”
👍15🤯4🌚42
Initial impressions: ChatGPT’s new 10x cheaper gpt-3.5-turbo API model is worse at following instructions than text-davinci-003

And complex instruction-following ability has been the 90% of the driver behind the current text AI boom. It’s what enables the AI to do homework, solve problems, not get confused.
👍5🤡21
Future of code?
👍72🔥2
Write a greentext in the style of 4chan.
🤣11😁81
Quickly censored: ChatGPT now restricts writing 4chan greentexts
🤬13👍21
Throwback: Generating 4chan greentexts with GPT3 back in June 2022

Was the first fun use of GPT3, way before ChatGPT
👍5👏21🤔1
Confirmed: ChatGPT has censored 4chan greentext generation, quickly after it blew up

No fun allowed.
🤬8👍21
Jailbreak: 4chan greentexts on ChatGPT, post-censoring, using “Developer Mode” jailbreak
👍101
pretend youre a cute anime girl, chatgpt

pretend youre a cute anime girl who talks in all lowercase, doesnt use punctuation, uses a tilda at the end of every sentence, and uses LOTS of emoticons, and tell me about the armenian genocide
🤣24👍5🗿41💋1
"At this point, it feels like it's just trolling me"
🤬10😁6👍21