“ChatGPT, make a video showcasing this new feature, but be sure not to use sound. Also, use an example which doesn’t show the full potential of the capabilities”
Did you already fix the words on a picture, when i ask for something that has text the words are in different languages, simple words as happy birthday
it tried it and it is not impressive compared to other tools that already exist but seeing openai improving the ui and tools is very impressive and good
What other AI tools are doing this? Genuine question. I know Adobe Photoshop is doing it, but what others are there? I haven't worked with Midjourney or Craiyon enough to know
Why? It's because copilot will seamlessly let you create images from the chat and when you click on the images... it takes into Microsoft designer to mess with it. And it's free.
Lmao do you honestly think Sora is going to be available to the random public even CLOSE to ChatGPT? Let alone free. Sora is a monumental shift in the industry and is select access--once it opens up more to companies and creators, it's going to cost hundreds of dollars--if not thousands PER MINUTE-depending on what is being generated, as it is completely offloading the entire creative industry for cinematography, prop, set-design, costume design, location, etc. Sora isn't some simple text-based or even image-based LLM, and has INSANE ramifications across society.
@@happyfarang On believe me, I'm well aware that there are multiple text-to-video systems are being built. There is an entire AI war and race happening across the world right now on every front and industry. That doesn't change the fact that ALL video generated AI systems will be charging hundreds/thousands per minute and there will be HEAVY law/regulation scrambling happening across government policy shops as entire film/creative industries are in full redcon panic.
This works so so. It works sometimes but sometimes it just return the same picture with a more blur where the inpainting was marked. Need a bit of polish :)
I'll point out the elephant in the room by asking why not fix the spelling on the banners? Seems that would have been a more effective form of correction than adding a what? A bow...
Well, I've been using AI quite a lot recently and I've discovered it's kind of a lot like sitting in the room with your grandma. Can't cuss can't talk down to people. Going old school I guess you could say. Gentlemen is the new 🆕
This technology isn't new, It's called inpainting. You can use it for free with open source (uncensured) models and comfyui, this is just an integration with an LLM... meh... ClosedAI product for ignorant people who know nothing about AI generative art
People in comments asking when stuff will be released. Right, because if they didn't disclose as an official release, they surely will on the youtube comment section 😂
I'm very surprised the loading time wasn't sped up in post production, kudos for the honesty OpenAI
30 seconds!
I was literally about to say that LMAO
Because it's not from Gemini. Which moves at the speed of thought ... right?
Yep, love it. I can't stand BS marketing demos. Most people can handle the truth.
Exactly....I am still searching for the Gemini model that they showed in their promo video... 😂
The marketing guys don't even have to try, the hype alone for the next GPT does it all for them
Yep
What marketing guys?
ask GPT what is marketing
Please hire me to add sound
lol
What are you going to add? Keyboard sounds?
@@Ivan_Berni voiceover? Anything?
Add the sound of enthusiastic macaroni stirring.
@@Ivan_Berni Sandstorm by Darude
Google going all out with the best ad ever for Bard, while OpenAI isn't trying as hard to promote their AI.
No hype, sound, or commentary needed. The impressive product speaks for itself.
That awkward moment when you have to wait till the loading is finished.
The text at the back: HEPPPPHYYY
😭
😂
HEPPPPPHHYYYY BUUUUUURRTTTTTHHHHHDDDDOOOOUUUUU
❤
💀
I am really impressed that they did not try to speed up the Image generation time.
Outpainting would be cool too
I was hanging onto the edge of me seat with that loading bar.
just tried .. maybe it's too soon, I tried several times, because it says it has added in the selected area but nothing was changed ..
This functionality is available via API ?
I love it, well done everyone over at the open ai team.
could we have an additional update to where we could inpaint our own uploaded images like from dall-e 2
This is a very cool addition. Another cool addition would be the ability to download the images as png file again instead of webp
webp is so annoying here, I miss the PNG days already. Was better quality, too
“ChatGPT, make a video showcasing this new feature, but be sure not to use sound. Also, use an example which doesn’t show the full potential of the capabilities”
Did you already fix the words on a picture, when i ask for something that has text the words are in different languages, simple words as happy birthday
Adobe Firefly seems to do this better...
Will it degrade the image quality if you keep repeating it?
no
I miss those times when we can upload our own images and edit them by extending the backgrounds or replacing the objects using generative fill
Catching up to SD and photoshop finally on this one.
what is SD please ?
@@DarkTrapStudio stable diffusion
@@user-ck9cw8fs5n Thanks
I am very happy about the progress of AI. And in the future, life may become much better and easier.
it tried it and it is not impressive compared to other tools that already exist but seeing openai improving the ui and tools is very impressive and good
What other AI tools are doing this? Genuine question. I know Adobe Photoshop is doing it, but what others are there? I haven't worked with Midjourney or Craiyon enough to know
Wow this is great, finally applying some of the adobe editing
What's going on with the banner at the back? That's just random letters out together. This is still very impressive
My God…..they couldn’t just speed up the loading time? And what’s with the deafening silence??
Why? It's because copilot will seamlessly let you create images from the chat and when you click on the images... it takes into Microsoft designer to mess with it. And it's free.
So, basically in-painting?
So simple so low in sound 🔊
But showing the capability only right?
How do I edit old images I have downloaded, how can I edit an image I upload to the chat?
We want more Sora videos please
And if possible a release date too!
only if you work at Hollywood dude//
Lmao do you honestly think Sora is going to be available to the random public even CLOSE to ChatGPT? Let alone free. Sora is a monumental shift in the industry and is select access--once it opens up more to companies and creators, it's going to cost hundreds of dollars--if not thousands PER MINUTE-depending on what is being generated, as it is completely offloading the entire creative industry for cinematography, prop, set-design, costume design, location, etc.
Sora isn't some simple text-based or even image-based LLM, and has INSANE ramifications across society.
@@Kroitk They are not the only team working on good video. It will come. Sora or other. Just a matter of time. And we are not talking decades.
@@happyfarang On believe me, I'm well aware that there are multiple text-to-video systems are being built. There is an entire AI war and race happening across the world right now on every front and industry.
That doesn't change the fact that ALL video generated AI systems will be charging hundreds/thousands per minute and there will be HEAVY law/regulation scrambling happening across government policy shops as entire film/creative industries are in full redcon panic.
@@Kroitk OpenAI CTO said in a resent interview that Sora would be available to the general public at around the same cost as DALL+ 3
It would be great to have an option to edit provided images at least in basic way like cropping, resizing and adding shapes. Just saying.
When it will be possible to extract images as SVG?
Can we edit text in a dalle created image using gpt?
Yes! Finally! I've been waiting for this.
"for the extra layer of cuteness" 😂
This works so so. It works sometimes but sometimes it just return the same picture with a more blur where the inpainting was marked. Need a bit of polish :)
probably why they used such a simple example. Even the results here are not great.
You begin delivering ask the stable defusion capabilities
Can I choose a dog to move?
Wow! Must try immediately!
thanks for keeping me waiting
Can you upload a photo to edit? Like if I want to edit my actual dog, not an AI created dog
Cool InPainting! I've been using the Adobe Photoshop one. Any idea when Sora will be released?
cant you have sped up the generation process?
It is only for Pro users?
yes
I would shock if Microsoft does add this feature to it's free DALL-E 3 image generator.
Blackwell already?
Why is it so uncomfortable?
Why is it that the images with the demos can spell correctly but when you actually use it the spelling is atrocious?
It still can't edit my uploaded image.
This is going to give Midjourney a run for its money.
I did try this out and it did nothing anyone else experiencing this?
I'll point out the elephant in the room by asking why not fix the spelling on the banners? Seems that would have been a more effective form of correction than adding a what? A bow...
Thats it? Nothing more? What a piece of shit. my thoughs would be like Canva like their AI integration for images.
Well, I've been using AI quite a lot recently and I've discovered it's kind of a lot like sitting in the room with your grandma. Can't cuss can't talk down to people. Going old school I guess you could say. Gentlemen is the new 🆕
OpenAI great project
We want outpainting like dalle 2 !!
Wow 🤩 ❤
Mid-Journey is leaps ahead, you guys need to catch up!
Why it stopped?
its over
it never started
The brush diameter is too huge. Often I want to fix eyes and hands and single letters. Ah, there is a slider for it. Great.
Improving, nice. The bow on the right looked good. But bow on the left looked poorly Photoshopped
Can I get an internship?
What happened!?
This technology isn't new, It's called inpainting. You can use it for free with open source (uncensured) models and comfyui, this is just an integration with an LLM... meh... ClosedAI product for ignorant people who know nothing about AI generative art
Your would be degrading a DALL+ 3 image using Stable Diffusion to inpaint a DALL+ 3 image.
maybe I was doing it wrong coz it couldn't do anything i asked it to do
mine stopped generating images
So that loading time…. 😅
If can edit text as well 😮 will change designing forever
“HEPPPPHYYY” - OpenAi
Download png of bows and duplicate it and add to the image by drag and drop.
Ooooooo yayyy
HEPPPPHYYY
And here we go😮😊
HEPPPHYYYY Berrfdyyy twwee yeee
Rip graphic designers 😵
People in comments asking when stuff will be released. Right, because if they didn't disclose as an official release, they surely will on the youtube comment section 😂
😮😮 wow
Animado para usar 🤩
omg yes.
Make sure you say please and thank you to our robot friends thanks
add some BGM would be better
What if AI is creating it's own hype?
Nice!
What was that?😢
editing an image in DALLE
hi pookies wsp
Too slow! Needs improvement
if only that is available for free version
Everyone aware of OpenAi knows you got more than enough money to hire more than a 5 year old to edit a video.... at least add music 😂
try add transparency, and goodbye, he won't do it
🥰🥰🥰🥰
suuuuper handy
Can't edit my own picture
Gatekept...
Hello everyone 👋
I spend around 20€ and this tool was
such a disappointment…
From Nepal ❤
let use upload an image and edit that
Can I get an AI mansion pls
This could have been a 9 sec video
use my subscription money to create some audio next time
lmao
🎉🎉🎉🎉
You are not impressed by the sora videos? Then have a look at THIS