Open AI and Chat gpt 4o Just shook the world up overnight

bnew

Veteran
Joined
Nov 1, 2015
Messages
62,479
Reputation
9,478
Daps
171,152
Whole creative world in shambles right now. Jesus :wow:



@bnew Im surpassed you didn't make an update on this

posted in the A.I development megathread and midjourney thread already.



1/2
@mars_eve
INPAINTING - REMIX😱
NEW 4o GPT/SORA IMAGE GENERATOR👀 @OpenAI



Gm6L7u_XsAAta80.png

Gm6L9ZaWsAAp8Td.jpg


2/2
@dancevatar
👀👀👀




To post tweets in this format, more info here: https://www.thecoli.com/threads/tips-and-tricks-for-posting-the-coli-megathread.984734/post-52211196



1/1
@abuchanlife
Breaking: OpenAI just added image generation to GPT-4o.

Finally, OpenAI followed Google and made image creation a built-in feature in ChatGPT.

Here’s what it can do:
✅ Create photorealistic images with high accuracy
✅ Generate perfect text inside images (no more AI gibberish)
✅ Edit and refine images through chat
✅ Keep consistency across multiple images

This means Photoshop, Canva, and many artists will lose work as AI takes over more design tasks.



https://video.twimg.com/ext_tw_video/1904617732475371520/pu/vid/avc1/1280x720/-vccQUDnfYcsQlPe.mp4


To post tweets in this format, more info here: https://www.thecoli.com/threads/tips-and-tricks-for-posting-the-coli-megathread.984734/post-52211196




1/2
@CommonSenseMars
New GPT 4o image generator in OpenAI's Sora with style presets



Gm6Oa4IacAA6mdn.png

Gm6O-W3a4AAKS3M.jpg


2/2
@CommonSenseMars
From the image generation in Sora, you can go straight to "create video".

Sora still has a buggy image to video output, where keeping style of the image input is always difficult. Otherwise, great functionality of image generation on the Sora site.



https://video.twimg.com/ext_tw_video/1904619456116252672/pu/vid/avc1/1280x720/uS5QFLSv1Nn6rX5n.mp4


To post tweets in this format, more info here: https://www.thecoli.com/threads/tips-and-tricks-for-posting-the-coli-megathread.984734/post-52211196




1/2
@blizaine
First GPT-4o image mod test: Make this man look like a pimp. how did it do?



Gm6ImafWMAA31l7.jpg


2/2
@blizaine
FYI

[Quoted tweet]
If you are getting bad results from GPT-4o image gen, it's because it's offline and reverted to Dalle for new threads.

Previously working threads now get this error. @sama


Gm6LHR2XkAA2iZB.png



To post tweets in this format, more info here: https://www.thecoli.com/threads/tips-and-tricks-for-posting-the-coli-megathread.984734/post-52211196




1/1
@joshuavoydik
GPT-4o Image Generation just dropped



Gm6LWK7bYAIFAUr.jpg



To post tweets in this format, more info here: https://www.thecoli.com/threads/tips-and-tricks-for-posting-the-coli-megathread.984734/post-52211196





1/8
@BenjaminDEKR
First experiment with GPT-4o Images:

"Put me in a modern spacesuit, make sure my face is still visible"

Most surprising: it didn't "edit" the image -- it recreated a new version



Gm6Dw0IXAAAx6m1.jpg

Gm6Dw0HWMAEu8bH.jpg


2/8
@BenjaminDEKR
"Now make a semi transparent tinted visor over the helmet, with the moon semi reflected in the visor"



Gm6FP4rXgAAdtYZ.jpg


3/8
@UziObi_
Honestly it’s well done



4/8
@BenjaminDEKR
Yes but it's interesting that it's not "editing around" the existing image

That could have pros and cons



5/8
@forgebitz
this is really good



6/8
@zamdoteth
not bad tbf



7/8
@mark_k
Strictly speaking, image editing with multimodal models is always creating a new image that resembles the input image. Sometimes more, sometimes less.



8/8
@LucLively
Pretty slick. New world unlocked




To post tweets in this format, more info here: https://www.thecoli.com/threads/tips-and-tricks-for-posting-the-coli-megathread.984734/post-52211196




1/4
@slow_developer
GPT-4o with image generation is insane.

now, what if you told some common people that this is not a real picture? would they believe it?

NO... absolutely not

[Quoted tweet]
4o image generation has arrived.

It's beginning to roll out today in ChatGPT and Sora to all Plus, Pro, Team, and Free users.


Gm6E77vWAAAYyxN.jpg

Gm5-GJRbYAAr1iz.jpg


2/4
@dreamworks2050
Your outie makes banger memes with ChatGPT



Gm6FU6BbYAMCStl.jpg


3/4
@slow_developer
WTH 😂



4/4
@zaveng
its really good, but now its saying that image generation tool is temporary disabled.




To post tweets in this format, more info here: https://www.thecoli.com/threads/tips-and-tricks-for-posting-the-coli-megathread.984734/post-52211196
 
Last edited:
Joined
Aug 3, 2012
Messages
43,126
Reputation
-35,573
Daps
239,983


H&M to create 'twins' of models with AI: "The possibilities are almost endless"​


Summarize

H&M joins the growing list of fashion companies utilizing artificial intelligence (AI) in their online communications. The retailer will create digital "twins" of real models. These AI reproductions will be used in social media posts and marketing campaigns, provided the models give their consent. Thirty models, including Mathilda Gvarliani (known from the March 2024 cover of Vogue Netherlands' "New York" issue) have already given their permission. "She's just like me, but without the jet lag," Gvarliani commented on the photo of her AI twin for H&M. Alongside enthusiasm, the decision has also been met with criticism.

Jörgen Andersson, Chief Creative Officer of H&M, shared a post on LinkedIn about the decision to use AI to replicate models: "Creativity and radical curiosity have always defined who we are at H&M. Now we are exploring new territory – generative AI – and discovering how technology can unlock new ways to showcase our design in innovative ways, while staying true to our human-centric approach. The possibilities are almost endless. We can explore and enhance creativity, which benefits us and the entire industry."

But where change occurs, criticism often follows. Strong criticism has come from influencer Morgan Riddle, among others. Via Instagram, she expressed her concern about the potential impact on jobs within the fashion industry and the rights of the models involved.

FashionUnited contacted H&M for more information about the use of AI technology. The retailer stated: "We believe that by collaborating with other creatives within the industry, such as agencies and the models themselves, we are doing this responsibly."

In an interview with Business of Fashion, H&M emphasized that the models retain control over the use of their digital replicas. They will also be financially compensated in the same way as for their photographs. Furthermore, the first images of the AI models will be shared on social media with watermarks, so that it is clear to the consumer that they are AI-generated images.

The Swedish retail giant H&M is not the first retailer to utilize AI models. Levi Strauss & Co. launched a campaign with AI-generated models in 2023 in collaboration with Amsterdam-based startup Lalaland. Fashion chain Mango launched its first campaign featuring an AI model in July 2024.
 

IIVI

Superstar
Joined
Mar 11, 2022
Messages
13,177
Reputation
3,379
Daps
46,756
Reppin
Los Angeles
The fact that the entire world is referring to the style this AI is creating as "Studio Ghibli" is why these companies are doomed. :dead: Just like every new AI trend - essays, then images, music etc the nerds get estatic for a week talking about how art is dead, but then the rest of the world points out they will never actually consume this shyt :russ:
Yeah, basically this.

It’s great for a week or two then phases out and begins that phase where most people find the “annoying” things about it.

People will use it to make some cool stuff and content but the world keeps going because most people are too lazy to use it to provide value anyways.
 
Joined
Aug 3, 2012
Messages
43,126
Reputation
-35,573
Daps
239,983
The fact that the entire world is referring to the style this AI is creating as "Studio Ghibli" is why these companies are doomed. :dead: Just like every new AI trend - essays, then images, music etc the nerds get estatic for a week talking about how art is dead, but then the rest of the world points out they will never actually consume this shyt :russ:


This is surface level complaints though, Sure the majority of the world won't care about a real picture turned into a cartoon. But the fact that you wrote that lets me know you know what going and that its way deeper than just a cartoon picture. I mean whole ads off of still pictures is wild.


Creative directors, art directors, set design, DP, sound design, models, graphic designers, cameraman, and everything else included. It might not affect you directly at this moment but shyt did just change for millions of people out there.
 

Scottie Drippin

Should Never Mention Me
Joined
Apr 30, 2012
Messages
17,158
Reputation
5,318
Daps
62,492
Reppin
The Traps of Unified Korea
This is surface level complaints though, Sure the majority of the world won't care about a real picture turned into a cartoon. But the fact that you wrote that lets me know you know what going and that its way deeper than just a cartoon picture. I mean whole ads off of still pictures is wild.


Creative directors, art directors, set design, DP, sound design, models, graphic designers, cameraman, and everything else included. It might not affect you directly at this moment but shyt did just change for millions of people out there.
I agree, the entertainment industry is cooked, but that's because they see profits in marginalizing people to produce this shyt, and people won't consume it.
 

yardman

Afro-pessimist
Joined
Sep 27, 2014
Messages
10,656
Reputation
9,240
Daps
58,118
Reppin
Whitehall
The art snobs went from “Ha! That AI trash can’t even draw fingers properly.:bryan:” to “This is stealing! You’re evil and killing the environment!:damn:
They’re just mad they can’t gatekeep anymore.:umad:
How the hell is a heritable trait/finely honed skill being gatekept, when anyone can pick up a pencil and train themselves to draw? Shut up.​
 

Illuminatos

#OVOXO
Joined
Jun 19, 2012
Messages
44,039
Reputation
1,603
Daps
182,598
Reppin
NULL
How the hell is a heritable trait/finely honed skill being gatekept, when anyone can pick up a pencil and train themselves to draw? Shut up.​
Of course ya'll are gatekeeping.:bryan:

It's just anger. I spent years learning how to draw and Joe Blow can just put some words in a prompt and produce art as good as me. Instead of being angry about it I suggest you guys utilize the tool. We're in a new era now. Adapt or die.:manny:
 
Top