ChatGPT

[Free]

OpenAI introduces custom AI assistants called “GPTs” that play different roles

Users can build and share custom-defined roles—from math mentor to sticker designer.

arstechnica.com

On Monday, OpenAI announced "GPTs," a new feature that allows ChatGPT users to create custom versions of its AI assistant that serve different roles or purposes. OpenAI will let users share GPT roles with others, and it plans to introduce a "GPT Store" later this month that will eventually share revenue with creators.

"Since launching ChatGPT, people have been asking for ways to customize ChatGPT to fit specific ways that they use it," writes OpenAI in a release provided to Ars. "We launched Custom Instructions in July that let you set some preferences, but requests for more control kept coming."

For example, in a screenshot of the GPTs interface provided by OpenAI, the upcoming GPT Store shows custom AI assistants called "Writing Coach," "Sous Chef," "Math Mentor," and "Sticker Whiz" available for selection. The screenshot describes the GPTs as assistants designed to help with writing feedback, recipes, homework help, and turning your ideas into die-cut stickers.

We'll probably soon see ChatGPT's version of Clippy...

 

[Dakota Tebaldi]

Okay, I get why OpenAI is doing this. It's like how Bic takes a pack of its disposable razors and paints the plastic handle pink instead of yellow and prints "...for Her" on the front of the package and presto it's suddenly an entirely separate product with a distinct customer base and that equals more sales which means more profit. That part I get.

I'm just confused, from like a customer standpoint. Why, as a customer, would you want a nerfed hyperfocused "version" of ChatGPT that can only give you nifty tips for sticker designs, as opposed to, like, the original thing which can give you nifty tips for sticker designs and help you cheat on your homework?

 

[Bartholomew Gallacher]

Well if the customer wants the Chatbot in customer support only, this has its use cases. E.g. a Windows support chatbot should not talk about MacOS, or how to bake bread.

 

[Free]

I imagine having a chatbot directly clued in on the topic(s) it's supposed to focus on can loosen a lot of the forced elements in ones phrasings for questions and whatnot. If I want to ask about coding solutions in say Javascript supporting specific ECMA versions, having a chatbot handling such could let me get to the issues I'm trying to hit on without preambles like "In Javascript with support for the 12 edition of ECMA 12th edition and above...," as well find some assurances that the resulting output will (should?) be restricted to that range.

 

[Noodles]

Okay, I get why OpenAI is doing this. It's like how Bic takes a pack of its disposable razors and paints the plastic handle pink instead of yellow and prints "...for Her" on the front of the package and presto it's suddenly an entirely separate product with a distinct customer base and that equals more sales which means more profit. That part I get.

I'm just confused, from like a customer standpoint. Why, as a customer, would you want a nerfed hyperfocused "version" of ChatGPT that can only give you nifty tips for sticker designs, as opposed to, like, the original thing which can give you nifty tips for sticker designs and help you cheat on your homework?

"ChatGPT, give me ten slogans for stickers."

ChatGPT: "Ok, here are some popular phrases that might be good for sticker slogans

Have a nice day

Shit happens

I love cats

Gas all the jews

Baby on board

Trump won 2020

Vaccines don't work

Clap if you like cats

Bazinga!"


I think basically it boils down to, sometimes you just, don't want the stupidity that might bubble up from the larger model.

 

[Argent Stonecutter]

It's like Google Search with a "site:" qualifier.

 

[Free]

Clap if you like cats

 

[Khamon]

My grandmother used to chide my grandfather mightily for using women's disposable razors because they were for ladies and men were supposed to use blades or electric razors and he was using those female disposal razors oh the horror. He was ahead of his time. She was still alive when men's disposable razors appeared on the market and, yes, they were the same thing with blue handles.

But this is not that thread. Where did this come from? Have to scroll up and see.

It's at least as interesting as a ChatGPT statement of absolute honest true fact ever.

 

[Free]

It's at least as interesting as a ChatGPT statement of absolute honest true fact ever.

Like, 1 + 1 equals the weight of a pound of gold?

 

[Khamon]

Yes, however, 1 + 3 equals the weight of a pound of feathers.

 

[Dakota Tebaldi]

I imagine having a chatbot directly clued in on the topic(s) it's supposed to focus on can loosen a lot of the forced elements in ones phrasings for questions and whatnot. If I want to ask about coding solutions in say Javascript supporting specific ECMA versions, having a chatbot handling such could let me get to the issues I'm trying to hit on without preambles like "In Javascript with support for the 12 edition of ECMA 12th edition and above...," as well find some assurances that the resulting output will (should?) be restricted to that range.

Buuuut, that's already not a problem. ChatGPT saves threads, so you can start one saying "for every question hereafter, assume I'm using Javascript with support for ECMA 12th edition and above", and boom, that thread is now your tailored specific chatbot with a focused function and personality for no additional cost.

 

[Kokoro Fasching]

Buuuut, that's already not a problem. ChatGPT saves threads, so you can start one saying "for every question hereafter, assume I'm using Javascript with support for ECMA 12th edition and above", and boom, that thread is now your tailored specific chatbot with a focused function and personality for no additional cost.

You forget, this is aimed towards the end non computer user, the one that calls you and says that their mouse is not working, and after 30 minutes of troubleshooting you find out that there is no power, and they aren't even still in the office.

 

[Free]

Buuuut, that's already not a problem. ChatGPT saves threads, so you can start one saying "for every question hereafter, assume I'm using Javascript with support for ECMA 12th edition and above", and boom, that thread is now your tailored specific chatbot with a focused function and personality for no additional cost.

Then mayhaps...this is the best way they've found to monetize it.

 

[Noodles]

You forget, this is aimed towards the end non computer user, the one that calls you and says that their mouse is not working, and after 30 minutes of troubleshooting you find out that there is no power, and they aren't even still in the office.

I kind of wonder if it might be just as easy to build a website that uses the API, and just prefixes every query quietly with restrictions.

"My Mouse isn't working"

translated to, "Pretend I am a braindead computer person asking: My mouse isn't working"

 

[Argent Stonecutter]

"The computer fletely, mouse and all." -- actual problem report that was a meme back before everything was a meme.

 

[Bartholomew Gallacher]

Ah the good old days of the Bastard Operator from Hell. Yeah.

BOFH ARCHIVE

 

[Free]

Sam Altman fired as CEO of OpenAI

Mira Murati will be the interim CEO.

www.theverge.com

Sam Altman has been fired as CEO of OpenAI, the company announced on Friday.

“Mr. Altman’s departure follows a deliberative review process by the board, which concluded that he was not consistently candid in his communications with the board, hindering its ability to exercise its responsibilities,” the company said in its blog post. “The board no longer has confidence in his ability to continue leading OpenAI.”

Chief technology officer Mira Murati will be the interim CEO, effective immediately. The company will be conducting a search for a permanent CEO successor. When contacted by The Verge, OpenAI’s communications department declined to comment beyond the blog post. Employees at OpenAI found out about the news when it was announced publicly, according to multiple sources.

Also,

OpenAI co-founder Greg Brockman is leaving, too

“Based on today’s news, I quit.”

www.theverge.com

 

[Dakota Tebaldi]

I wonder what he actually did; the company's statement is predictably vague.

 

[Argent Stonecutter]

He's been making some pretty wacky public statements. Maybe they finally got tired of it.

 

[Bartholomew Gallacher]

IMHO he lied to the board so boldly, that they had to fire him immediately. Maybe he was evaluating the military use of ChatGPT together with the Pentagon?

which concluded that he was not consistently candid in his communications with the board,