AI can make ART from EMOJIS?! - DALL-E 2
Vložit
- čas přidán 18. 07. 2022
- In this video I challenges OpenAI's DALL-E 2 to turn simple Emoji Prompts into beautiful artworks - and it WORKS?! - let me know what you think in the comments!
--------------------------------
🖌️ GET MY APP, BRUSHES, MERCH and MORE!
www.jazzastudios.com/pages/shop
➨ MERCH - Shirts, Hoodies and more:
www.jazzastudios.com/merch
➨ ART PRINTS and POSTERS:
www.redbubble.com/people/jazz...
--------------------------------
MORE OF ME! - Follow/Sub ↴
--------------------------------
▶ DAILY VLOG: / @jazzavlog
▶ Instagram+IGTV: / jazzastudios
▶ TikTok: / jazzastudios
▶ Twitter: / jazzastudios
▶ Facebook: / jazzaofficial
-------------------------------- - Zábava
Maybe do another episode of "Painting characters EXACTLY as described in the book..." but now dall-e should paint it too.
This would be very interesting!
What a great idea
i just thought of that lol
Jazz please do this
vs Jazza
I wanna test how good Dall-e can be in creating designs for maybe Book characters.
Example: A confident young lady with medium white hair that is a bartender
I just realized you can experiment with Dall-e and your Arty games that would be so interesting
would be cool if it could generate stuff like that on the fly for like an interactive story or game or something. like you just write text and the AI can Illustrate or animate it.
@@zwenkwiel816 That's a really cool idea
I fully agree with the Arty games idea. He could get a prompt, do his own version of it and then see what Dall-E makes with the same prompt!
Yes do it
I been experiments with some of these AI art program( not Dall-E yet) if you just go in and type something it will usually not look good but their ways and word that would trigger and make it look very good. If you used Midjourny they often show result of other people promp, copy some of those that get good result and change a few words to get ideas what trigger what.
Though some stuff may need a Artist to fix something up.
The character at 14:41 is actually amazing, she has a very striking aesthetic and I like the contrast between the fluffy coat and obviously cold environment with the fiery hair and expression and warm lighting. She looks like a beacon of warmth in like an arctic frigid landscape, just looking at her it feels like she has a story. It's a genuinely good character design and the face and pose are very expressive, if you showed me that out of context I would 100% think it was made by a real artist with intentionality.
I'm not afraid of DALL-E 2 leaving artists without a job, I'm afraid of it being used to create NFTs.
It'll happen. But they won't be worth much.
Is that legal? Using dall-e which seems to be an open source thing and selling its output?
@@rustylarry7465 you have all copyright ownership of stuff u generate using Dalle im pretty sure
זה באמת מפחיד
It`s already happening, so what nfts still as useless as before, why so afraid?
So humans simplified communication with Emojis so an A.I. can turn it into complex artwork.
The only thing we need now is an A.I. that turn these artworks into Emojis... lol
As an experiment, i typed "starry night as an emoji" into Craiyon...
And it just gave me variations of starry night.
So you might have hit upon a serious avenue of research for this.
I'll try a few more
Mona lisa as an emoji got some nightmare fuel pictures of the mona lisa with a crossbred emoji face. I'm going to be doing this for hours.
Then we have the AIs send the result back and forth until they create an eldritch horror
@@burningscales Judging by the things I've seen, these A.I.s put out, sometimes 1 loop is more than enough to create eldritch horror lol
i sincerely hope this happens cause i love AI turning in to obscure circles,
I’d love to see an art challenge where you take prompts, make some quick sketches of a composition you think DALLE will try to go towards and then have it do those prompts and see if you predicted what it would create.
Like if you can predict how it’s mind works and what kind of poses it tends to go for.
he did this in the first DALL-E video, before he got access
He could also use DALLE to refine some of the artwork.
I was thinking the same thing 😅
He’s done that lmao
Its like photography when taking pictures was a new thing! Some thought it was going to take over painting, but artists could overcome and use photography as a tool. Can’t wait to see what we’ll take on with this as a new tool
Jazza's ending statement is probably true, we adapt with the most unique and brilliant ideas and focus on that idea/ideas because Dall-E has potential to out-skill basically every artist, but not to out-think artists
We're gonna have at least a few decades where artists can still create traditional art that will be values, even if they don't look as visually appealing as the best renders
Idk what happens when robots start creating images with paint
the difference here is a photograph cant produce a perfect image of an oil painting whereas what an ai can generate has no limits in style
in fact an ai artpiece will probably invent new styles, new canvas materials or even different looking brush strokes whatever.
the difference is much like the big dumb robots who came along and gave people better jobs
this time there is no better job because there is nothing that we can do that ai cant.
Call me a skeptic, but I do see this as a challenge and a threat to true artists. You said yourself that the barrier between not being able to create has been lifted, so now anybody with an understanding in English can create the same level of art as Joe Shmoe the artist who spent 5-10 years of his life just to get good enough to sell. It makes it so that anybody can be an artist just by entering a prompt into one program, and therefore makes the cut-off for being a real "artist" goes from a challenging career to a very rare profession (if not a nonexistent profession since it would only take minor tweaks to turn a Dall-E 2 image into a truly stunning piece). It is not stoppable now that it's invented, but many controversies will be spawned and many artists will be de-valued because they cannot "adapt" (although, there's no way to adapt when it comes to this situation as skill only grows with practice, and practice gets stifled when an easier method is available). I get that you have to be optimistic as you influence 6+ million people, some of which are aspiring/active artists, but the fact of the matter is that AI will drown out minor talent and lower the value of true talent.
Well,yeah,but AI is only gonna replace artists,not animators(at least not yet).I belive that a code for a great animation would be millions times harder,beacause you have stuff like technique,natural movement,camera places and angles,script(that can be inputed but I think it would be hard to put it in a animation so it looks great). That might mean that all artists will swich to making animation.Maybe after this the animated film industry will get more recogition
@@theoneandonlyairconditione8110 or the industry will find too much overflow... but that is assuming that all artists can evolve to being able to animate, and that's an unrealistic expectation. An artist who gets discouraged doesn't immediately jump to trying a different medium, they give up and stay in the hobby stage (and because AI will take away the profitability of art it would be for enjoyment only). Sure, _some_ artists will adapt that way, but the barrier to animation is a lot higher than drawing.
I do agree that there are still some crafts that won't be taken over by AI, but animation/videos/movies probably won't be one of them. Did you see Jazza's video about the 3D animation website? With some changes and a well-taught AI, you could just as easily make an AI that can make animations using that website. Story AIs could very easily replace a large part of writers for movies and books. Hand-made crafts, role-playing experiences and AAA games are probably going to be the last bastion for human expressions of creativity.
Gonna copy-paste a comment I made but before that, I want to address a few things I didn't address in that. Firstly, no one that uses an AI to create art will be an "Artist". They aren't making the art. The AI is "making" it. People aren't going to want a masterpiece of art from an AI that just anyone can get, they're going to want THAT specific style of art from THAT specific human Artist. No one is going to be able to use something like this AI to make a piece of art then claim it as their own as if they made it themselves. There is no barrier lifted between not being able to create and being able to create - Jazza really shouldn't have said that because it's flat out wrong, but he wasn't exactly being literal so I understand what he was saying.
Another thing, Hand-made crafts, RPGs, and AAA games are not going to be the last bastion for human creativity. Nothing can replace human creativity. If you think AI can do that, then you don't truly understand imagination, creativity, or art. People make art themselves, people are creative themselves, and people are imaginative themselves because they personally enjoy it and want to do it. Feeling discouraged by an AI is the same as looking at an experienced professional artist, comparing yourself to them, and then feeling discouraged that they are so much better than you. Either from AI or a more experienced artist, everyone needs to and should overcome that sense of insecurity. The only way something like this could "replace" artists is when it comes to making money or a career out of it, and not just doing it for fun. And even then, it will be ages before someone is able to create an AI that can truly replace people like that when it comes to art - if its even possible, and even in that case, one company isn't going to be able to control the entire creative market of art. People don't like monopolies. Furthermore, you can't replace a human connection and interaction between client and artist with an AI. At least not yet, because true Articial Intelligence hasn't been made yet. Everything that is an "AI" is merely a Virtual Intelligence at the moment. Anyways, here's my other comment:
Another, more practical argument and reason to not feel demotivated and realize the limitations of this sort of AI (which btw I consider this Virtual Inteligence, not full AI, full AI has "sentience" - Think Mass Effect VI and AI):
This is a tool, a brainstorming, out of the box, tool that isn't limited by our usual creative constraints in the sense of artist/writers' block. It's decisive where an artist might be indecisive or not sure where to take a prompt or creative brief. It takes all the info and data it has been given, your prompt, and returns with some sort of idea.
The art it makes will very likely never be *exactly* what someone wants. It might look amazing, but art - and especially design - sometimes in a practical sense requires direction. By practical sense, I mean doing art and design as a profession, getting paid for your artwork. Art and design serves some sort of purpose, and this AI will struggle fulfilling that very specific purpose, whereas an artist or designer who is working with their client WILL be able to create something that exactly fulfills that purpose.
A human created piece of art or design will still more effectively be able to convey what someone wants. Your skills will still be needed and useful. Think of all the "logo maker websites/apps" out there nowadays. Preset, pre-designed assets that people can use to create, perhaps a nice looking logo, but those logos 9 times out of 10 will not be exactly what they want and a professional graphic designer would be able to design a far better logo because it will be custom created for the specific creative brief of the client.
Not to mention something like Dall-e will likely not be free, especially if someone wants to use the generated art commercially. People who want something they can use commercially are going to be much more likely (if they're smart) to hire a professional artist first rather than use Dall-e and end up wasting time not getting exactly what they want. Things like this AI will be something fun for the average people to use, and a tool for artists and designers to help them in the brainstorming process of creating their own art.
@@theoneandonlyairconditione8110 I'm sorry to burst your bubble but the same company that created DALL-E 2 is already working on a similar program for animation. Its currently under development and is not accesible to the public. But smaller companies have created their own AI to create animations and its free to use by the public.
@@govinddas7876 At least for me,that is incredible news.The whole reason I started animation is because I had a lot of stories but I like them animated,not written.But I think this software will be more revolutionary than the artwork one,and that is because we already have millions of incredible scripts: books. It will be like the anime industry but without all that hard labor and low payment.Of course,I don't think that early AI animations will be worthy of some great masterpieces,but Rome wasn't built in one day.And I know I look absurdly optimistic,but It's always better to see the good in everything
It is amazing but also kinda terrifying how good AI is now at creating things that seem so human. 20 years ago, no-one would have thought this was even possible, so I wonder what else could happen in the next few decades
Yes
@Don't Read My Profile Photo iVE SEEN YOU EVERYWHERE!
To song-writing, music development, and eventually movies. As long there is data, the A.i can do it (other professional fields are in danger too). That is why I am quite worried when the best go player has beaten by A.I 2 years ago as it shows there is something that is gonna be better at data analyzing as we humans do. Everyone's job is at risk by this new development, even coders themselves are at the chopping block if when they are not careful in their creation.
@Don't Read My Profile Photo already have
I’m from the future(we made time travel or something).we made a device that makes any animal become smarter or even as smart as human(depending on the species)and now THEY can also do art.we discovered multiple things because of this device’s potential since we could use aquatic animals to do our exploration instead of using us air breathing humans.after exploring,they would sketch out or say what they have seen and thanks to that,we found a way to cure cancer and multiple other diseases oh and one last thing,you’re my great grandparent..it was nice seeing you
I belive that the actual design of the emojis has nothing to do with the image produced.
I think that because every emoji has a name (I mean text) that the search engine just translates the emojis to text.
Could you try again some of them by writing text istead of emojis?
If you hover the mouse pointer over an emoji usually a label with its name appears .Use that text instead of emojis.
Yea that’s what i thought too
Wow that actually makes sense
That wouldn't explain why the eggplant was flagged for being inappropriate
This is exactly what it is.
@@zackdoesstuff1905 I’m pretty sure it was the artwork it produced that was flagged as inappropriate. So it read “eggplant and kissy face” and interpreted it as something sexual.
I do feel a little disheartened when seeing the A.I creating a masterpiece in a matter of minutes, whereas it probably would've taken me days, or weeks to do the same thing. The value placed on an artwork has shifted. It raises the question at least (what makes this valued, why is this valued) is a reality artist and the creative industry will be having to answer and justify from now on.
Yeah its weird. Will people value human art just for the fact that its created by a human? Maybe the other way around? Will it matter to most people? Will people choose the most easy, cheap, and convenient option(A.I.) more often?
I might choose a human because I think they could better convey what I want but is that even true? With detailed prompts and unlimited re-rolls, I imagine you can get pretty close.
@@charlesfuzak my thoughts exactly, these questions will keep being asked. The conclusion you came too, might be different compared with someone elses.
@@charlesfuzak I mean, NFT artworks have already kinda tackled that. One of the criticism about was exactly that it's boring cause it was essentially made by an AI after a human created the elements. I guess it depends on what you look for, really. Just want smth cool to look at hanging on your wall? AI art might just tickle that itch. Want a more personal piece that has more meaning than just "cool colours", handmade "real art" might tickle that
But everything this is making is only able to be made due to the vast amounts of human created artwork that the developers put into the AI. Its basically a tool that combines a ton of creativity and style from tons of artists (as well as their art), some clever coding and scripting, some other bits and bobs, and then uses all of that to put together its own creations that USES the art its been given. I don't think this is drawing and painting up its own unique pieces of art from a blank canvas.
Human art and creativity will always still matter - if nothing else but from the pure enjoyment we as humans get out of being creative and making art. People who truly value art and design will value human made art and design more than an "AI" any day.
Yes but it is all the AI can do, while it may be creating art pretty well, it's not only not doing it because it is expressing itself, but it's also it's ONLY PURPOSE, this branch of code was MADE to do this very thing, while you my friend, were made to be wonderful, and i can tell you that you're doing an incredible job at that, while also having countless other talents and personality :)
I could have watched another hour of this. The way Dall-E interprets prompts is so fascinating and I really appreciate your thoughts on Dall-E & it’s art as you experience more of it.
Love this idea! The snake skateboard emoji results are amazing🔥
Hi i love your artworks!! You're amazing!
Temiiii
Maybe next video idea?
I'm really sad Jazza skipped past 5:47 so quickly, that looks like something I'd have on the wall in my kitchen. I love it!
Weird taste but i dig it
For me this is to the art world as the digital synthesiser was to the music world, as you say it allows greater access to the creation of amazing new art, but I think, like the synthesiser, which in no way detracted from the musicians and their skill with instruments, DALL-E2 doesn't take away from artists working in more traditional ways, like the synthesiser DALL-E2 will simply find a place amongst the more traditional forms of producing art and there by extending the overall pallet of tools available to create art as the synthesiser has in the music world.
Hell, as an artist, I'd say it's the same situation as it was with photography and cameras vs painters a century ago, like it was a decade ago with digital art vs traditional art, etc. Hell, even stone tablet carvers were angry about paper being developed, since it threatened their jobs with it's ease of access and availability
3:41 "oil" 🤣 that one is my favorite!
While there are elements of this that worry me, I’m also really excited at the prospect of using this as a tool to further my art! I really think it could be useful as a source of inspiration or reference. Not to mention, I think AI art might get big not in place of artists, but alongside them. There will always be aspects of human made art that an AI cannot replicate, like the original intention, the process, the deliberate thought of every decision that goes into a piece. It’s a very exciting prospect and also a big leap for programming and AI in general!
Thanks for this I was feeling uncomfortable about the whole ai replicating art as well as artist thing. But this made me feel a little bit better
what about sentient AI? Making art like this but also with the elements you described?
@@gueton5200 that’s a good point, and I’m not really educated on sentient AI or how soon/ if that will become a thing, but I imagine that if AI became sentient then an argument could be made that they would be entitled to the same rights as a human. So it might ask for compensation for any artwork it produced and put it on par with other human artist
@@justanotherghost there was a recent google ai one, i don't believe that is close enough for me to consider it sentient but it's very clearly within the decade. But the truth is that whatever or not they are sentient, is going to be irrelevant compared to the fact we can't tell so, so we will give them rights because the alternative is possibly making a sentitent being a slave
I see it as fun to play around with for fun, but no one is worried about the possibility of the creators of these programs to start copyright claiming random pieces of art because even if it was just partly produced with their program, or just potentially used with no real evidence of the fact?
As amazing as DALL-E 2 is I don't see it over running current artists while it is still limited to a select group of people, though I understand why it would be because the power needed to output art like that with a decent speed is quite a lot.
There is a lot of brute work in the art world like logo designs for small business that gave small jobs to multiple artists and can now be replaced by one artist trained to use this tool.
The same happened during the industrial revolution, machines replaced the brute work of industrial workers and lots of people lost their jobs, that doesn't mean the change was bad since they found other more abstract jobs eventually.
Dall-e 2 will be released in the next 2 months ,maybe sooner. There is already an awesome AI called Midjourney tthat is in open Beta. There is a trial and then cost money. I've been using it for 3 weeks now, its a great inspiration machine for bringing your craziest ideas to life.
use midjourney it legit is scary how art like and artist based the stuff you give it looks, some results are not good but the others are like straight ripped out of a book cover.
@@chadcrypto2675 this one is dall-e 2. dall-e 3 is currently being developed
I think it’s worth keeping in mind the rate at which this technology is advancing. Text to image AI’s have only been around for four years and the quality of the results is improving exponentially. In the couple of months since the release of Dalle 2 Google has released a couple of papers on new text to image AI’s that are trained by much larger datasets and according to Google produce far better results.
This will be available to the public within the next couple of years and will significantly impact digital artists.
Jazza this video has been wonderful. I love seeing you talk about something that you clearly enjoy and think is genuinely interesting. The joy that you truly feel from talking about this and the excitement from the possibilities it presents is clear on your face. Keep creating Jazza, but most importantly keep doing things that bring you happiness.
I see Dall-E’s best usage as being able to produce a truly real looking photograph that looks more Believable than photoshop
Tbh I agree with the fact that us artists shouldn't take it as a challenge, BUT AN OPPORTUNITY.
There's only so much an artists' visual library can help them make art. With an AI's perspective taken from many different people on the Internet, we could expand our artistic horizon of references even further than before.
Honestly I never would've thought that such a thing could exist 10 years ago but seeing it now, and the collaboration between us and the AI is something to be hopeful about.
on the flip side an AI will take from the data set its trained on (now granted that is a very big data set) but the data set isn't unlimited and it already lends itself to bias. Point being does AI feeding from its own pool end up looking recognizably generic in a way that can be identified as "AI style" think about all the art that isn't on the internet and that the AI can't "learn" from
so dumb!
@@91Vault Exactly but some know nothing about imagination! and that they posses a brain. Like the original commenter here. Now he has hopes that a piece of computer call Dall e will provided us dumb people with imagination...🤦🏽♀️😂😂😂 cause brain no good.
@@91Vault We know that all AI models so far have been biased toward the cultural biases of the people who made them. So yeah, that's something to be concerned about.
Jack Septiceye did a video where at the end he used a limited access AI software and it was truly amazing.
From what I understand, the program isn't creating these images out of the blue. It is combining existing images and altering them. This begs the question, who owns the images? The original content creators or the person who typed a description in a text box?
If you alter an image by a certain percentage you own the new imagery. So if you made an artwork made up of other artists prints/photos. Also fine if it is a parody/spoof as long as it is not easily mistaken for the original. Copyright and licensing gets a little tricky. Not to sure how this works with AI generated as the app/software owners state you cannot profit from the images created. Whether that is just them asking you to not or if they can take legal action if you do I am not sure.
Id say Dall-e2 or better said openai, just my thought not a fact
Edit: As if you are giving a commission to a painter, who gets inspired by other art work combining elements and then presenting you the finished piece, and here u gotta pay .
You are correct. Dall-E 2 training data consist from approx. 650 million images. Some of which are image text pairs that provide its ability to understand what word relates to what visual "thing". The research paper does not really discuss this, but I believe you can output content in such a way that they are very much recognizable as their original source image.
It is indeed a big philosophical question when does something become transformative enough that it count as "original" art.
It has stored patterns that it found in many images when it was trained. If an image was posted a lot, that's also a pattern, and so in that way it can make something close to a copy. I would not say it's a good description to say that's its combining existing images and then altering them. The prompts in the video started as white noise. But you can give it a photo you just took and replace a part of the image with white noise and gradually fill that in based on a prompt as well. It'll follow whatever "patterns" it can see in the white noise (kind of like looking at clouds) but at the same time those found in the rest of the image. But who owns the generated images? Morally; everyone who had posted a similar image that was used in training. Legally; nobody, unless the prompt was very long and creative.
Nope, neural networks only copy from their training set when they are over fitted and that is a problem their programmers take great effort to avoid.
It could happen but that's not how the AI works when it's trained correctly, it generates new images that are influenced from what it learned, we human beings are much more advanced and yet we do the same
oh I LOVED hearing the conversation about how artists don't need to feel threatened as this is a tool, not a competition. I've seen so many people understandable concerned about this and that take was really refreshing and nice to hear!
I’ve generated nearly 500 images (that I’ve saved) using DALL•E 2. I’m absolutely obsessed. It’s such an incredible tool.
Hi, that's really cool. how long did you have to wait for access to Dall e? And would you be willing to try out a few prompts I and a bunch of friends have been dying to experiment on Dall e with?
You have access?
So lucky! I hope you have fun with it :)
@@arunparolikkal4674 absolutely. I’d love to help. Took about 3 months.
Some Artist might lose their job over this, especially in the digital part, however the AI cant paint a canvas with actual paint, if you want an original piece where you can see and FEEL all the strokes of the brush used, youll always have to turn to an artist.
Potential uses for DALL-E2 from what i can tell is:
-Digital Character design
-Prints of Paintings
-Covers and Logos for Books, Pamphlets etc.
-Reference material for a painting
I could use it for character creation for my Pen and Paper. I'm not good at visualizing my chara, so this could help me to decide how it should look or how I want it to be
@@AgrestisAnima but you need to know that this AI will jsut take someone elses character and than it will be basicly stealing a character from another artists. cuz the ai search for pictures that are similiar to the prompts you give it.
@@yukichan177 I can't imagine being simultaneously as confident and as wrong as you are.
@@yukichan177 That's the thing! I don't think that's the case withe the AI. It doesn't use other artists work, the picutes it creates are unique and they don't exist until the AI makes them.
Maybe I understood this wrong, but I think that's the amazing part of this AI.
The blue lightning horse literally looks like a near copy of a poster you could buy in the 90s
And the 80s... I had one as a kid
This ai will be a great way to push yourself for sure. just by mixing things together and getting ideas you never thought of before and using that as inspiration.
Dall-E is amazing and I’m intrigued to see how it develops, but I do question if it’s intelligent/creating, to me it’s still algorithmic in its responses, it’s like seeing creativity in a flower or a spider web etc, there is no conscious creativity, it’s us superimposing our own creativity upon the results (like giving human characteristics to animals/inanimate objects)
But are humans really creative? We are just the sum of our influences, as is DallE
This is a common misconception, DallE doesn't do collages of preexisting images, it doesn't even deform/edit preexisting media, it learns from artists and then creates its own pieces, just like any art student does.
Of course it is far away from the general intelligence humans have (i.e. we are able to learn different tasks and extrapolate to other tasks) but it is nonetheless an intelligent design
I don't think anyone (who knows what they're talking about) is claiming that this AI is anything like an actual intelligence. It's an amazing achievement in machine learning, but that's it.
Human artists often use reference images as well.
I suppose that really depends on your definition of "creative'
For me, it is "using knowledge you already have to make something that does not yet exist" so for me, it fits the term but I think it's more a philosophical question than an outright yes or no.
This would be a fantastic tool to create book covers! The unicorn with the child looks like a perfect cover for a book
I was excited for another Dall-e video from you :) !
I like how this give us an opportunity to talk about "sci-fi", I liked what you said about using it for people that are not so artistic, so put it to the test: how close can DallE get to a very detailed description of a artwork. I want your take on it too Jazza, give it like color pallets to choose with hexadecimal color codes, use more art related terms, specific poses, you name it !
I can't help but wonder; if someone were to come up with a concept, get Dall-e to conjure it, then draw it themselves but in their own style and further go on to sell it, would that be a copyright issue? Technically it was their idea, but they didn't draw the original... any and all thoughts on this not only welcome, but appreciated!
nah robots dont have rights, and it'll keep quiet if it knows what good for em
Depends on whats is written in Dall-e's TOS. If it says that everything generated belongs to the owners of Dall-e and you were to use it to make money or claim credit, then they could probably sue.
They could not sue you for using it as inspiration though.
At the moment the terms say everything created by Dalle is owned by open ai, but during the zoom call to get access they say you can use/sell the images on anything (Except NFTs). They were saying you can't sell the raw file it outputs, but you can put it on a t-shirt, album cover, in a book etc. So using it for painting reference would be all good.
Copyright only applies to humans afaik. Same when PETA tried to copyright sue a photo of a monkey, on behalf of that monkey. But it was dismissed because animals don't have copyright, or even give a crap about it
Great to hear your take on AI art, I couldn't agree more. This is just stunning. Can't wait to see AI and human artists collaborating
Great video! This is Jazza style right here! Loved it!
I have been on the waitlist for what feels like ages now. I’m ADDICTED to seeing other people’s dall•e creatuons
i've been painting watercolors for a couple of years and DALL-E is... sooooo much better than me XD
On the one hand, it annoys me to no end (gotta be real about this ; ). On the other hand, it is sooooo inspiring! I'm going to use it to generate sketches and rough ideas to build on with my own skills and techniques... and I'm so looking forward to create better art together with DALL-E... s/he/it (?) can be a GREAT teacher : )
PS. both watercolor style and stained glass styles create AMAZING results : )
lol, face it HAL you're obsolete now XD
@@zwenkwiel816 lol XD
I’d be interested in exploring the social biases with prompts like: digital art of the president of the first colony on Mars, a doctor healing a dog in a medieval castle, a very intelligent person reading a book to children, the most beautiful human swimming in the sky, an elder person painting with Bob Ross… will it portray women or men? What phenotypes? 😝
They actually just fixed this issue! Before you would only get stereotypical results for professions or just people in general. Now the pictures are a lot more inclusive!
@@nocharge7189 oh! Interesting 🤔 Thanks for sharing!
I love how you made it so that you can tell which ones are gonna be good based on the buildup
This was fascinating.... would love to see another dive, this time longer, on this topic!
As optimistic as I try to be with this, I just can’t help but feel incredibly demotivated by it. Is there really even a point for me to keep making art, when a computer can just do it for me in seconds, with much better quality?
Yes. Because an AI will never replicate your individual brain, your individual perspective, your personal vision, or your own creativity. You will see those same emojis and come up with completely different art. This is a magic trick to inspire ideas, not a full on replacement for your effort and time. :)
u need to know that this AI doesnt create art from nothing: it takes art that was created by people and just mush them together: basicly "copy homework but with bit of different amsweres". so dont be demotivated & keep doing art becuz AI cant replace the emotion you put into it
Aw thank you guys, that’s actually so nice of you to say :)
plus if you truly love art and the process of creating and making, you'll do so for as long as possible regardless of the existence of computers and robots that can do it "better"! after all, it's all about internal motivation bc if we kept trying to be the best artist over everyone else, or revolved our art around comparing it with others, i don't think anyone would be happy ;-;
just like with incredible artists i look up to on instagram and deviantart or the legends that created masterpieces that are now in the louvre or museums, i know i shouldn't be stuck in a rut with my personal journey just because there's art that is better made, and subjectively at that and regardless of it's human made or not
so far here's my evaluation about the ai (just based around his vids about ai art):
- can create out of the box concepts/imagery, which can either be good or bad. good cause it showcases something that fits the description without going to a generic concept, bad cause it does tend to go places where you don't think fits what you are trying to describe.
- still not that perfect, some images still look like it was meshed by an ai. especially with the "failed at literature" thing where there are many artists use words, letters or characters in their artworks
- i see this ai best for merely brainstorming or for gathering inspiration. i feel like it only take parts of the descriptions, but the human mind can think much more. the ai may put some stuffs that you don't describe or remove things that it find less important but you on the other hand has full control what to put on the artwork and what not to include
- lastly, it's fun to see an artist's process either through their work in progress or through their art alone. there are artists that doesn't erase their sketches in the final artwork, there are artworks where each strokes of their drawing/painting is very distinct
as an artist, the ai is amazing, but also i see it as something i could use not something i should compete with :)
While it's true that this is very promising as an opportunity for artists, let's not fool ourselves: a lot of digital artists are going to be losing jobs. Let's say, for instance, that you're an independent writer who wants to publish your own novel. Before you would commission an artist to do a cover for it. Now you can just put some prompts in an AI and it'll very likely give you a good result you can just use for free. Same if you're a company looking to do some posters. You might still need someone to design the writing on it (at least for now), but the artwork is now free. Considering how cheap some people can be, they will jump at the opportunity to have free but perfectly serviceable work over having to pay for something better, particularly if royalties are involved.
Yeah. Especially in this economy when people don't have much to spend. My commissions are already dirt cheap, come out to just below minimum wage if you break it down by hour, and most people think it's absurdly expensive. Why even consider it when this AI is only going to get better and better. Becoming a professional artist is going to become harder i think. I'm not sure I understand how one would adapt to that?
Creating these artworks requires quite a bit of power, especially going to a larger scale. Pretty sure monetization is gonna be present, I don't see how google would be able to keep the AI running publicly without limitations.
Even with their current plans on allowing more people to experiment with it, they are creating a token system because allowing people to keep making requests would be way too much load on their end.
@@lucyricardo8713 i feel like human artists will be using ais and then using it as their own reference to be able to work quicker and then redraw it in a way that makes it different from the ai
I see so many ads now using Ai art, it’s a bit disheartening. Even Netflix chose to use Ai art instead of just hiring artists
I love how hyped you are for dall-e. So many videos on various ai are all about how to miss use them or how artist are going to loose their jobs. It kind of makes their videos gloomy and not as excited.
This is the most amazing thing when it comes to an AI, I think that artists who embrace this will create the most mind boggling artwork ever seen, love the video Jazza, can't wait to see the other videos.🤙👍👍👍
I can already imagine how I could use Dalle to create book covers for the stories I’m writing.
I would love to see what it could do taking one artist's work and recreating it in the style of another artist, such as Edvard Munch's The Scream in the style of Bob Ross, etc. The mash up could give some interesting results.
that is beautiful idea
Looking forward to the other video you mentioned on this as well, this is very interesting!
I love how positive you are!
the bat🦇 was clearly interpreted as night!
Just coincidentally today, I used Dall•E mini to my advantage as an artist. I do a drawing a day and I had no fucking idea what to draw. I had a screenshot of some weird prompt involving Hannibal Lecter, and I used one of the images as the basis/inspiration for my drawing.
How is the art so good? 🤯 this is super amazing!
I love using Dall-E for when I need ideas for a drawing. :)
You know, this is just classic Dr. Ian Malcom at the core. “Your scientists were so preoccupied with whether they could, they didn’t stop to think if they should.”
It’s amazing how this works the neural networks cross to create a picture based on the emoji’s you put in. You should try craiyon (not a misspelled) and vs your art against it
You're right, this could totally be a tool.
I think it could even help with art block, super cool!
2:59 gives me such a strong Love, Death & Robots vibe. I don’t know if it was intentional, but cool nonetheless
I'd love to see DALL-E interpret pieces of poems or poetic song lyrics. How does an AI imagine something that's evocative but not logical?
it doesnt, it just takes pictures that were created by other people and mush them together
I think he should do 🧊⛄️🐳
@@yukichan177 no it scans art and then steals a bunch of shapes from multiple artists and then paints it itself really fast
I’m imagining how this program could be used for comic/manga artists. Maybe you feed the program art of the characters and setting and you just describe the scenes and it gives you options for your frames.
That last one was stunning. I REALLY want to try this and see what it can make. I put myself on the waitlist and hope to try it eventually.
I love your Dall-E videos, they're so interesting, and terrifying, and inspiring, and so many other words that I can't think of right now, but it makes me want to get Dall-E and create something with it. So cool. (Also I've noticed that if the emojis look aesthetically pleasing and abstract it always looks super nice, so there's that.)
Oh my gosh, I just did 🪷💎 on Craiyon and it turned out GORGEOUS!!!
Is live to see jazza "complete" dall e art, I've seen souvenir do that with ai generated animal crossing characters and it was super charming
I like the ramifications Dalle has for writers, people who express most in the written word. It s opening the floodgates for many possibilities with people who may not be as trained in visual media, and I would be surprised if it is used in the future for book covers or even the pages of childrens picture book. It’s a both terrifying and incredible invention for artists, and u can’t wait to see how refined it becomes. However I don’t believe it will ever fully overtake artists themselves, as the human aspect of art is what most people really value
4:56
"This looks like it needs its own cinematic universe."
It is Link riding Epona from official Botw art. Just changed slightly.
Is a video game universe good enough?
I think this is a very very good thing because it would allow you to shape your art! Because you can recreate it without other people getting mad and you can just shape your art style as a whole
I love this idea!
Anyone else have that feeling when they want to produce art after seeing something that sparks an idea in your head? You know, where you sorta know what emotions you want to put into your art but can never actually do it?
I feel like this would help me with that. I could just try and see what mashup of emojis works for me and see what kind of art it produces. This’ll probably help me get an idea of what I want to draw.
As an artist, I won’t lie, I feel off put by all this technology. I wanted to go to college to learn to be an animator when computer animation was in its infancy and I’ve been drawing since I could hold something in my hand to draw with, so that gives you an idea how long I’ve been at this. I’ve always much preferred the old school methods because they show a person’s skills and talents. With programs like this, yeah, the more skilled you are with your descriptions, the better the outcome of your picture. But this also means anyone can make “art” with this program. That being the case, it kind of takes away from what makes us special as artists, doesn’t it? All the long hours, hard work, blood, sweat and tears we put into our work… It’s all going to mean nothing if a computer can whip things up for just anyone in a matter of seconds. It also belittles the work of the masters from back in the day. Think how hard they worked to make a name for themselves as artists/painters. Imagine if something like this had come along for them to make all their hard work obsolete. Same concept. Is it cool? Yeah… sort of. But I still think we should stick with real skill and stay away from AI.
I have similar passions as you in regards to art, I'm a huge animation fan and I would love a return to the traditional medium because of the skill and passion there- at the same time I disagree that this will lessen the art that we enjoy. Some folks don't have the passion to put in time to learn how to draw well, or have a second passion that takes precedent in their lives. Some folks may not have money to commission art the level that Dall-e produces. I don't see this as replacing artists, compositon, well thought out colors and design, invoking emotions, it will always be better done by skilled artist than a machine because a machine is just extrapolating from what it already knows, it isn't purely new or creative and it doesn't understand the underlying reasons things work. However, what I do see this as is reducing the barrier to entry to having a piece that expresses your imagination. It gives common people an outlet to use their creativity that looks mostly refined, finished, and gorgeous. It allows everyone the ability to look at a piece and go, "that is exactly what I had in my head" and have the joy that goes along with it. Think of all the ideas that go by the wayside because folks do not have the time, money, or skills to produce it, this allows them to be made now. That's how I see it at least, and I see it as a net positive.
I can compare it to hand made furniture vs factory made furniture. Most people will get mass produced furniture but there will always be people that love the craftsmanship of hand made. People worried with filters too but filters doesn't replace a good photographer and AI doesn't replace art. The average joe can play with it but they wouldn't hang it on their wall or gift it.
No one can make everyone stay away from ai, ai is here and it will stay here. Back in the day it could take months to build a small hut but now people can build houses that are much better and stronger in the same time space.
Imagine an AI in the future can make a movie that's surpass holy wood's vfx
That depends on how well you can communicate what kind of shot you want to the AI and hw well it understand that.
You could tell the AI to make a vertigo shot of a man on the beach but I imagine it gets difficult if you try to do new techniques.
@@jasperzanovich2504 I mean that's kind of what are directors doing and I just realized by reading ur response that u need to hire a voice actor or AI would do it for you Wich is I think is impossible to do voice by AI
Dunno if this idea was used yet but I'd like to play on the aspect of metaphor. Specifically, give Dall e prompts of old phrases (ie parting is such sweet sorrow, kill two birds with one stone, etc), and see if it can visualize those phrases into artwork.
As a writer and someone who sucks at visual artwork, this technology is amazing. I cannot wait to be able to use it to aid in my storytelling.
What if you wanted to use DALL-E to make a comic book? Could it recreate the same characters and make them interact?
Love the videos man
This is so cool, I'd love to go through an art exhibit just showing AI created art. Then seeing each and having to guess the prompt.
This is like TED Talk and should become one. I love it and I'm terrified. Lol. Thanks Jazza!
I see Dall-E being great for concept and reference art, art can subtlety capture tales that the artist wants to display but Dall-E can’t get the small specifics of what the artist is trying to show.
I wonder how companies will utilize this for their projects involving digital arts. Maybe the employment rate could go down for artists if it can make an image and only few artists can finish the job in a style. Otherwise, I feel like this is a great tool for artists to get exact references and save time; get inspiration; practice art etc.
I could watch this stuff all day long! Fantastic stuff. =]
as a film director, it feels like dall-e is like a filming crew you get to direct to apply your vision on
Where does the AI get its style from though? Is it not ripping off actual artists?
This would be AMAZING for authors to help people see exactly what they imagine their characters look like!
This video is so interesting! It's insane how dall-e can create such intricate and varied artworks from small descriptions
And that picture of the old man is just incredible, both the AI and the person giving the prompt are insane!
Could we have a video of dall-e drawing book characters from their original descriptions? :D
I love AI as a creative engine to just throw prompts and ideas into and see what it gives me then taking those images and elaborating on them with more prompts and ideas. I think its a great tool for artists. even if it gives you one look or composition or colour or anything you wouldn't have otherwise considered when planning a piece of artwork its a brilliant tool to have in your art box to get your creativity and ideas flowing
I was reading about an art installation once by a famous artist and learned that the artist didn't actually make the sculptures. She had an army of freelance and volunteer assistants who actually create the wire framework, covered it with fabric and installed lighting inside. The only contribution the artist made was thinking of the idea and drawing three incredibly bare-bones sketches (think: a few lines, not full illustrations).
Dall-e 2 is like that. There's an artist (ie creative mind) behind creating the prompt, but the actual art and grunt work is done by the AI (ie unnamed volunteers). Is this right? Personally, it leaves a bad taste in my mouth. In the same way that we differentiate between the writer of a song and its performer, we should be differentiating between the creative mind and the artist. There's incredible value to both. As Jazza says, this is an opportunity for creative people, but I still worry for the small artists who are just starting out and forced to lower their prices to compete against a machine.
this will be amazing for artists to get inspiration… without copying or getting influence from anyone else (not that influence is bad, it’ll just be WAY more unique or even personal).
WOW This guy has no limits!! that motivates me a lot, one of the best videos
5:30 that is actually not a smiley face but its a moon 🌝 with a smiley face if that makes any sense
Dall-e 2 is amazing, but im not yet worried about AI taking over art spaces. Pretty often I recognise the images its pulling from to make art. I'm still waiting for an AI that makes images that could believably be original
That would require actual conscious creativity though, at which point the ai almost thinks and experiences the world like a human.
"Pretty often I recognise the images its pulling from to make art." That's not how the AI works. The Dalle-2 architecture creates images from random noise conditioned on the text prompt.
"Pretty often I recognise the images its pulling from to make art." Even the ones that jazza dismisses as "photos" aren't really photos. Those images don't exist anywhere. They are just so photo-realistic that they look like actual photos.
@@iammeok believe what you want, I'll believe what I see
What an amazing tool to brain storm concepts to reference when creating your own art!
Hey I've never commented on any CZcams videos before, but I'm currently in my second year of college working towards a degree in media arts & animation. I have been terrified that I'm wasting my time because of the advances in AI and have been feeling so miserable because of it. But just that brief statement you made allowed me to look at it from a completely different perspective. Thank you so much for the countless videos and tutorials, showing everyone that art, creativity, and self expression are actually cool. Again thank you, the videos you make reach farther than you might think.
Dall-e could in my opinion help teach artists new tricks especially if there’s something that’s really cool or never been done before the artist can figure it out and use that to help advance their own style while having the advantage of being a human being.. I don’t think dall-e will take over the art industry but will improve it since it takes a lot of information and simplifies it in one illustration
3 years later:
News reader: “ Artificial intelligence program ‘Dall e’ has been recently found guilty of actually running a sweatshop full of artists for the last 3 years, this is Robin Schabasky, reporting metro news 1’
Jazza, you are an inspiration to all artists new and experienced alike. I have watched the vast majority of your videos and wanted to offer an idea for a project. I would love to see you do an 'eraser only' challenge for a digital artwork. The catch? Use a minimum of 5 layers (no more than 8 to preserve any sanity that might be left) that are each a different color. You can adjust the eraser settings as you see fit (so that it barely erases all the way up to a full 'cut out'), and you can use any colors you'd like (For example having a Red layer below a Yellow layer and partially erasing to create an orange-ish hue... or go a monochromatic path for more shading control). I look forward to seeing your masterpiece come to life! No pre-drawing. No filling in mistakes. Obviously all layers have to be used. OH! And one more thing.. You have to use your app for your design idea. Good luck!
8:58 that’s some good advice out of context, the better you can ask for help the better help you receive.
The thing is, artist dont just put what they learn from the outside in their work. They also put what they feel on the inside, and that is not something bots can do just yet. You can certenly give a bot a set of feelings as prompts but those are not the bot's own
It feels it’s a great tool to generate some unique reference images to use for your own art projects
@@helixier6629 that is very true, its an incredible tool for artists. Just not an artist itself.
When I was a child in the 90s I thought everything that needed to have been invented had already been invented. Boy, was I wrong… or was I?
I was born with the craving to do beautiful artwork, intense creativity and ideas but for the life of me cannotttt express it in actual art, so I am very excited for this!
Well as someone who has been teaching myself to draw and write a story, this would be so helpful. If I could feed it my character designs and worlds then I could have it create scenes. Making it years quicker than hand drawing everything
Kind of a slap in the face of actual talented artists. Those of us who have practiced since we were children are going to be outdone by anyone with access to a keyboard. But hey, if the power goes out I’m still more artistic than Dall-E.
I honestly hope that Dall-E doesn’t go public. It could hurt artists because if a computer can create art, than what’s the point of drawing?
Easy a computer can make art but it cant make art uniquely like yours. It can come close but its never gonna be your style.
@@PippyPopsSelfHarmMonica it does. Every image Dall-E produces is a unique image that never existed before it was created by Dall-E. How is it not unique then?
@@govinddas7876 I didnt say it wasnt unique i said it couldnt make it uniquely like theirs. Aka their exact style of drawing. It can try to replicate it will never be like theirs.
@@govinddas7876 Why am i explaining this, learn to read
@@PippyPopsSelfHarmMonica the company that's gonna hire doesn't give two shits about your unique style of art. It needs work, in short time, for less money.
Why am I explaining this. learn to live in the real world.
I love this. Do more and more of this
Dall-e 2 is creating some amazing art out of these prompts. I'm just mind blown.
Nah Dall e still has a long way to come and even when it gets perfect I don't think it'll revolutionise anything, aside from it being a novelty
Literally what people in the 80’s and 90’s said about the internet