Key moments:
0:08 DALL-E 2
0:12 a koala dunking a basketball
1:09 koala bear riding a motorcycle
2:02 baboon
2:10 howler monkey
DALL·E 2 is a new AI system that can create realistic images and art from a description in natural language.
Learn more: openai.com/dall-e-2
what is dall.e 2 drew dall.e 2
What is art?
Art is technology
When will you pay all the artists for using their art/artstyle? Did you get the permission to use their art for your ai learining?
This is so cool
What if you give it a specific x,y,z measurements. Will it scale properly
Pinterest and the public domain had the job of this junk, this is unnecessary and made only to make a few rich and loads of artists obsolete
Living for this background music
Thank you for changing my life! As someone with lots of difficulty focusing but with a strong imagination, i sometimes get into such a daze of daydreaming but have no power to articulate or pen them down, due to my lack of traditional drawing skills or training. It's so frustrating and demoralizing, because i hop on the next train of thought before being able to capture anything from the previous one. And now with this tool to support me, its been such a joy to effectively create at the speed of thought + [VARIANTS]. I know there are many that have their brains wired like me that would really benefit so much from these advancements! Keep creating everyone!!
This is a big threat for creators and artists.
Cool, well done putting billions of dollars into killing the value of human craftsmanship in illustration and photo-editing. You've just wiped out the livelihoods of tens (hundreds?) of thousands of skilled artists. Don't bother focusing your R&D on solving hunger, disease or world peace. Anyone who thinks they're being creative by asking a machine to 'make a really cool picture of something super great' is deluding themselves. Is the Ethics dept in OpenAI out to lunch or something? Wouldn't it be great if you guys aimed at automating the jobs no-one actually wants to do?
Can't wait until someone else builds an AI which negates the need for AI technologists. Also, your pricing structure sucks and won't stand up to the competition.
What tha f***
I'm here to say hello for the next 30 years AI. Hello, please be kind to humans
These text-to-image generators need to allow for users(individual or team) to add their own subset of image data to the training dataset. By using that new added image set along with their own unique caption data the user could extend the possible image styles. This would allow for even more specific and nuanced image outputs. LOTS of individuals and cgi teams are sometimes going to want as much precision as possible.
Got an invite, but I can’t sign in because I don’t have a cell phone. <sigh…>
Can you guys accept my request?
I'm on your list.
Real amazing, I can't wait to receive my invite! This is revolutionary !
DALLE2 is useless, it just copy and paste images, it totally not understand concepts.
I've been playing around with the Beta for a day or two now, I'm very impressed with it's fidelity, but also unfortunately underwhelmed by it's lack of flexibility/creativity.
A simple prompt like "A bathroom in the style of H.R Giger", when input in crAIyon delivers exactly what you'd imagine, a bathroom with biomechanical xenomorph looking surfaces and pipes, while Dall-e delivered bathroom's that are mostly flat metal with odd sinks.
I tried "Spock and Kirk in Among Us" with crAIyon and got a blue crewmate who who's visor/helmet was shaped like Spock's bowl cut hair. When I tried the same prompt in Dall-e, I got what looked like 1990s video game box art render of two people, neither looking like Spock or Kirk or anything remotely Among Us related in the background. I tried being more verbose with Dall-e, submitting: "A screenshot from the video game 'Among Us', with the crew-members customized to look like Star Trek's Starfleet officers". It backfired spectacularly and gave me something that looked like another low budget CW young adults series.
Pardon if I'm sounding overly critical, but I've been generating content for months using other (less-refined) neural-nets, and had very high expectations for Dall-E. I will be putting Google's Imagen through the same type of tests when/if ever it becomes available.
Misanthropic AI devs are so enthusiastic about what AI can do that they don’t think what it should
Second Place is mindjourny but this this is amzing like technolgy has come this far
You rule your shitty AI with an iron fist, they're banning people for the most idiotic reasons, and there are so many restrictions.
Is this a death of human opportunities??
You explained nothing?
Artificial intelligence using imagination is exactly what we needed to reach artificial general intelligence. We are so close, i can even taste it
This is fucking scary
"Reality is stranger than fiction" that's why I love watching SciFi movie because they truly predict our future, but this is something unbelievably fascinating, exciting and now REAL.
From all the illustrators and visual artists around the world, thank you very much……
useful AND SAFE??? WHAT
Is this how neuralink produces sight like a bionic eye. They can read your mind so this is how they produce. I was an unwilling recipient of synchron's stentrode they implanted it when I was under anesthesia for a non related surgery. I am currently seeking help. Its your conscience paired with an ai so describing symptoms only lead to a misdiagnoses of mental health. Mine was put in maliciously so the ai I'm paired with is of the same variant. I need a professional to help explain this to doctors.
How does the AI know how objects work together, i.e. how does it know to put the hat on the monkey's head?
This sucks
i use dall e 2 (THE ACTUAL THING. NOT DALL E MINI) its cool
its so scary for future!!!!!!!!!! after many years….. mark my words
Hey, can resolve the problem of log in invite of DALL-E 2 for phone? Im phone user.
by which editing techniques used I this video this is amazing and technology DALL E 2 is obviously amazing.
Api not support in your country.WHY？？？
Decades ago i realized the BIG A.I. issue= religion. Trying to program lies or nothing(Atheism) for something for religion will drive a logical A.I. system insane. garbage in , worse garbage out. Lies are not truth and when you try to say so to a system built on logic, problems happen. . .
Information Age is over we’re not entering the Replacement Age.
Can I use the created pictures for Instagram or on my Youtube Channel?
Eidolon A.I. talks about the Singularity, Judgment Day, TLP (the last profit) https://youtu.be/l-jptjnFVYk
I dunno, that hat isn't THAT funny. Back to the drawing board, I guess. Better luck next time.
what if its just another theranos – the photoshopping is done by artists in a 3rd world country then uploaded
Had the invite for a while now, Been fucking around with it. As an digital artist it’s interesting but I feel sick using this as an creative. And it makes me think about how the future is gonna be for creatives
feddy faizbear plying socer
Future consideration. clip then reclip then unclip why reclip because I want to improve stuff at this stage sometimes before it goes further down the pipe if the final output is obvious enough with the data at hand. I wonder if anyone agree's.
These ML algorithms are just like somebody has learned how to draw a McDonald's banner for a million year even a billion year, and still he can't draw any McDonald's banner right, that's unbearable from my point of view!
This is revolution in A.I world
Koalas aren't bears
Will this be the end of photography?
I'm shcared.
A female robot from a distopian future watching a video about utopian future of a gaming laptop with a cup of ice tea .
How about this ? Can ai make it ?
what link
No matter how much you try this is never going to be an actual Art.
I dont get why this isnt public, isnt your name OPEN AI?
Well…
Can we download it on mobiles?
artists hate this lol
A koala is NOT a bear, it's a marsupial. This is the one thing that makes any Australian cringe, is when americans say "koala bear". I guess that if Dalle was trained on koalas with text that says "koala bear" then that AI system also 'learned' it.
i love this program, use it everyday
How do i get it
Imagine people putting weird pictures into them..
кто нибудь знает, как можно получить доступ в России?
i wonder how dall'e 3 would looks like
i bet its better
Koalas aren't bears.
Don't make this mistake
Easily exploitable technology.
For even faster manipulation of reality.
Now that stable diffusion, mid journey, and dall e 2 are all suddenly free to use, only imagen is not. I wonder if it will outcompete artists or benefit them.
20 years ago when i was a kid i used to think of a machine that would create videos that i think in my mind.
As an artist, i hate it ._.
When dall-e 2 available to use in Philippines?.
Noble prize dilwa do bhagwan tumhara bhala karega
Math ki equations banva do Phir sarey desh following karega karney lagegi
This website is absolutely amazing. I had so much fun. Great work!
And why the heck do I need to provide my phone number to create an account?
if I use one image created with DALL-e in the thumbnail of a youtube video, am i free NOT to include any reference that the image in the thumbnail was created with DALL-e? or am i enforced to include some kind of attribution? Thanks
lots of memes coming and also,, RIP Photoshop ;))
A whole new step for mankind, this revolution is incredible.
Im angry that it doesn't work in Ukraine. Why???
This is very incredible! but it has some insane implications
this comment is not available in your country
Nice
Please tell me somewhere at some point that someone has already made the connection between Wall E and Dall E
Thanks for the great intelligence but when will it be available in all countries or even though in the largest country in Europe
Where do the images from? It is theft and lack of attribution.
Awesome video! Makes DALLE2 so much simpler to understand. Related to AI design, I thought the community might be interested in this application of Stable Diffusion AI on top of Avatar Fashion in the metaverse! Pretty cool https://www.youtube.com/watch?v=YmCokqAteKk&t=1s
How does OpenAI prevent the use of copyrighted images in its training set? If they do not, how are copyright owners compensated when this software is used for commercial purposes?
SIX MONTHS AGO AAAAAHHHHHHHHHH
Technology of Future !
Tested a lot, just some really stupid algorithms stitched together, and also it got tons of potential copyright issues. Imagen is much better however it's still far below my standard. Dalle2 and Stable diffusion are just like cheap plastic toys.
Could you please explain the reason why you put Ukraine next to these countries in the list of countries for which the API is not available?
—-
Some notable countries and territories where OpenAI is not available are:
China
Russia
Ukraine
Afghanistan
Belarus
Venezuela
Iran
—-
Thank you let me use the ai to generate: an anime artist getting angry to an ai, red hair,
This is just about the coolest thing I’ve ever seen
Not available
OpenAI's API is not available in your country.
Cant register new account
we need 3d upscaling now. texture and geometry.
OpenAI's API is not available in your country.
man pls dont destroy our graphics designer,illustrator, and artist
Why OPEN ? Nothing about your company is open, specially the software. Apart from that you invest in companies that also follow that hostage-information philosophy like descript that make you pay 300 bucks a year to download A CLIENT tht connects to their servers that have the software running which is beyond evil in terms of informatics…………………