They said the same about cameras. How they would ruin art. They didn't. It became it's own type of art. And artists themselves, real artists taht don't care about critics, they are inventors and explorers. And they always have been. Vermeer painted his "masterpieces" with a camera lucida. Of which he modified to create perfect pigment matching. Technically one of the world's first photographers. Instead of using chemicals to make color, he did it himself by hand. In the art critic world they want to say that is cheating. In the photography world that is insanely talented. Some art critics today still don't want to believe it even after the overwhelming proof that has been given. Explored in the documentary "Tim's Vermeer". AI is nothing but a new brush. A tool. Like the camera before it, the camera obscura before that, the camera lucida, and the finest hair on a donkey's bum before that.
Love it! That's such a good perspective to have and a great story that proves human ingenuity and creativity will always find a way to express itself in new ways
Only difference is, mediums of the past weren't 'controlled' by Google overloards + their SJW minions. I know this is on another level, but no one was paying attention in the past - when creative minds pushed limits. Part of the big picture of "art" is ebb 'n' flow. Not just of medium and mind, but the deeper psyche of the (human) collective. Humanity NEEDS to PLAY with things and find (organic) equilibrium as part of a holistic intelligence process. Over sanitizing shit (as we're witnessing) leads to a lack of immunity against deviance + perversion. It's not 'healthy' to try and control/steer the collective psyche. It leads to inability + breakdown (sickness). But.....I'm assuming Google already knows this. In short, AI is wholly conditional, while natural intelligence is wholly unconditional (thus superior). The camera is a great tool. AI is a great tool. Guns are great tools too. Protect everyone from tools and eventually the world is overpopulated with useless eaters who don't know how to DO anything. Or, are just completely scared shitless (of anything outside their little box). Sorry - High octane morning ramble :/
@@Mymysticmama Actually classism was very much in play back then. And secrets were held by the artists. As was writing and reading, depending on the exact era and location.
@@Mymysticmama Okay, wow! Yeah the control of these tools is what's scarier than the actual tool as it is going to cause big problems for a lot of people
I used numerous AI tools. Blue Willow is the one I like the most. Using Blue Willow, I produced stunning photographs, and it's really cool and most importantly it's completely free
its so funny because a few years ago people were like "artists will be one of the last jobs to be lost to AI because it requires creativity" instead it seems to be one of the first.
I think some of them would lose their jobs, not all. Definitely people who start out with little projects they would struggle and barrier to becoming an artist, a successful one would be even greater than before
@@pieflower6419 Don't be, They still can't understand a lot and it's far off creating truly fine detail. Think of it more as a tool you will use to help imagination later on. I use Dall-E-2 to generate creative images that I then draw. I have an idea in my head. Blast it at Dall-E-" it shows something eventually that resembles what I want then work from that. We're still quite a way off it killing art. Not that it ever could. But it is going to takes its toll on stock imagery and assets. But ultimately its going to just become a tool that you will use as an artist. Not a full blown replacement of you. Plus these things cant do text at all. It cant make logos etc. Right now they're jsut good at photoshop really. There is also the fact that this is creating images from images that exist. So theres a lot you can actually ask for that It can't grasp. Asking dalle for a stick man never works for instance. It jsut gives you a dodgy stick lol. Plus I've generated like 100 images from one instruction and had 100 compactly irrelevant images back. I've had it come back with things it thinks is right but is totally wrong because someone gave an image a bad description and instead of a toaster its a foot. Ultimately all these things can do is read what the descriptions are on an image. Diffuse and understand some objects based on that via commonality. cross art styles and diffuse on lighting. Smash it all together and bebang. So it's not capable of creating something truly unique.
I tried free AI software for Dall-E-2 and it was a complete failure when it came to vehicles such as helicopters and planes. Even when I was giving a specific automobile model, it would do things like give different headlamps to each side.
I'm hopeful that you'll get more control over the images. I was apart of the mid-journey beta test. Their AI produced beautiful artistic images, but you had next to zero control. I wanted to be able to highlight what the AI got wrong n what it got right. I was able to produce a three-page comic, but quickly realized that it just wasn't going to work.
Oh wow! I guess even then at least it produced something close to what you wanted. That's the limitation with these current image generators. It's only going to get better though 🙂
@@beyondengineering I wish I could link and show the images I produced, because even with the limited control I had it was still kind of cool. It would’ve cost me hundreds of dollars to do what I did. The iamges were digitally drawn, colored, and painted.
@@beyondengineering I uploaded the comic. It's not where I wanted it to be, but it's still impressive for a morning’s work. Even with the distortion and the lack of control, you can see that in a few short years that the creative world is going to radically change. I don't think I spent more than two hours on these images. I had no story, but wanted to see how far I could stretch the tech.
Yeah that's the thing, it's the speed with which you get what you want as compared to before. This is why tech like this would take off, at least in its current form you can tell an artist what to create for you eventually based off of the generated images you already have
Hahaha... Expectation --->"To extend human creativity"...... Reality ---> - Google and other large company executives and shareholders laughing in their lambos on the way to their 3rd mansion that they got from all the extra money they saved when they fired 10's of thousands of people even remotely connected to content creation and replaced them with 10 highly niche AI Computer scientists*
That is definitely a possibility in the future, I'm not sure it applies to Google today though. Hopefully we sort out the ethics before deploying all these AIs without understanding the true consequences. I like to be hopeful that this will create new jobs and open up a new economy. Again, multiple ways this could go wrong but if we get it right, our lives would be much better for it!
@@beyondengineering You almost sound like you actually believe that. You must live in a world where Google aren't a tyrannical organisation with dark agendas. Whatever it could be, nevertheless it will be used for evil.
If you have humans who can do more, you can just do more... Like drown the internet in your content and get everyone's attention. Which will make you more money. Tech has always empowered people to do more... Of course, you also have to learn how to do new things and give up your old ways of doing things.
Me: Hey Google, generate and image for 2 frogs playing chess underwater in the art style of pre historic cave paintings. Google: easy, here you go Me: hey Google what’s the time? Google: Thyme is a herb typically used to garnish foo…
Hey! I wanted to ask you if you think that this is the end for all artists, I have been recopilating information during these last months, the answers that I got most of the time were: A-Nope, human creativity and soul are irreplaceable (but big corporations don't care) B-Yeah, it definetly will (ignoring that this kind of A.I can't really understand abstract concepts or modify as much as a client would like to the image in comparison to an artist, because an A.I uses only literal words that are in the algorythim and an artist has an advantage in understanding "the mood" that a person wants more deeply than just "cute" "sad" or "happy" even if they say that they want that, they just don't really know how to put it in words) C-It will become a tool for artists like farmer machines did (but the farmer machines aren't able to do everything alone and a damn lot of stuff can go wrong with a farm, which is very important since we all have to eat (in comparison to entertainment it wouldn't matter as much so probably the firsts years there would be a lot of errors or some then it will just win over humans)) D- This kind of "A.I" can't really do this much since it doesn't really understand humans and do multitasking like an artist would in their minds since it isn't a general intelligence one, when it becomes one ; it will be able to get over the job. (Which I agree a little but since I do not know anything about "A.I" I can't really have an opinion) What do you think about all of these answers and if possible, answer to this question as well with your own words, thanks in advance!
As an artist. I say yes, this is the end for 90 % of illustrators, concept artists, and photographers. Some will still keep their jobs for a while. They will become vinyl records among digital streaming services. I am currently switching my career to programming , starting out with js /angular. 20 years of learing artist skills will become virtually worthless when these engines go public. Shit happens.
@@rasheednesbitt8667 It can't really understand the mood of an image out of the general "sad" or general "happy", basicly there aren't any nuances of both, if you say both it will probably give a weird result because the ppl that programmed it just said to the machine : "this colors, shapes, this and that are the things that are -insert feeling-" and since it is very... Nuanced and it is just hard to describe because of relativity and life experiences (without forgetting that some clients say they want a general feeling for a drawing but they really meant another thing (which is why artists do some little searches outside of what the client wants)). Maybe I am in the wrong and I am saying non-sensical things but this is what I heard the most and thought that it is kind of right(? Please, leave a comment if you don't think the same, it is helpful
@@cynamonstudio but programing wil become almost obsolete when the general A.I comes out (like every job (not the ones in which there are risks for the general population, somebody has to watch if the machine does something wrong for sure)) this actual "A.I" has some information, but still, it seems like we have 5 to 10 more years maximum or even lesser than 5 depending on the big companies before it gets perfect, that is what I think at least, please leave a comment, it is very helpful to receive diverse opinions
@@cynamonstudio also, the D and C answer is the one that a A.I student and an A.I programmer gave me= A.I can't do that much since it really doesn't understand what it is doing, it really just tries to make things look as belivable and realistic as possible but can't do "story-telling" on a character nor that much for now since it has a ton of errors(just look at the images, even I, a rookie artist can see them).If it keeps the text form since it will not be possible to get that much into details (changing details) and for being able to change those details or errors (if there are any) it would maybe be an actual artist job until general A.I gets to the level (at that moment most jobs will not exist and there will be a sallary for everyone since we can't do better job than the A.I and the only humans that will have works will be those that have to reassure that the machine isn't going to make a big error that would kill a person ex: surgery (yes, there is already a machine being used for surgery, it is called davinci, it will probably have an A.I when it gets to the human level))... Who knows, I am just a teenager trying to see the truth thinking critically, so probably I am in the wrong because of my very little experience in general so please, leave a comment, it is very helpful!
I'm or at least I used to be a traditional artist, until sever anxiety & constantly stressed due to over work. It killed my focus & artistic creativity. Using Midjourney was an amazing experience for me because I can finally create things that are in my mind. Things that people don't consider that artists use this as well. Most artists don't like the realistic photo style, but we love the non realistic stuff Like A dirty slimy cat that is made out of garbage. Simple commands creat realistic stuff, but absurd commands create art. Also you use your imagination & command of description as well as artistic terminology.
Hey Vega, I hope tools like this do help us in getting more creative but at the same time it is scary that tech like this is moving so fast and possibly destroying many careers in the process, increasing the wealth inequality before we've even caught up with it
What I find unnerving is that it understands about light transport, you know, diffraction, diffusion, shadows and reflections. How could it know about these things that are not implicit in the prompts provided to it?
Hi Chris, That's the part that got me the most too. It truly is stunning, but the researchers built it with specific aesthetics in mind, their research paper mentions that they added parameters to enable this. Obviously it wouldn't really understand what it is doing but the neural network does the magic!
I'm pretty confident this will replace artists, not all of them, not the real artists, but majority of the artistic employment will go to AI. A lot of the "art" that is produced today is just content: stuff to be shown in a presentation or an ad, in the backgrounds of movies and games, stuff no one really cares about. Many artists will be just flat our replaced by AI. Many will change how they work to babysit the AI and do some post-processing on the results, and many will use this as an extension of their art, 3D modellers that generate textures from text prompt (this is already happening) or creating background art for their comics and movies. But I think the biggest problem will come from automatically generated . There's already an overwhelming number of websites, (most notable all over facebook), that are just a flood of AI generated articles. So far they needed people to add in the images and do some checks, in particular for custom content. However, with models like these and the more advanced ones that will come out, I think the internet will be overwhelmed with a flood of generated content. There's going to be millions more videos and pages made by AI generators than by humans. Now, even people don't make content all that worthwhile most of the time, but it's going to get a whole lot worse.
First off, a lot of what you see in the backgrounds of movies and games is from hard working well trained professionals that have worked hard for decades to hone their skills. I assure you people do care about these things. Concept artists love their job and people love the world's they make with those honed skills. Secondly, it will take a lot more advance models of AI to replace artists. I think in the near future iterations of this might give some graphic designers a bit of trouble, but for the most part I already have ideas for how to use it to help me in my work flows as an asset, not a threat. I don't think this will be a threat at least in my life time.
@@sociallyresponsiblexenomor7608 I'm not saying the artists don't care about their work or that they're not talented, far from it I know they're incredibly dedicated and hard working. But I also know that most artists don't care about the background, they just need it to be there. Even prop artists like to create a scene, and will just put a stock photo on a plane outside of the window, or use a random placement brush to scatter trees and shrubs. And the audience doesn't care that much either. The only reason players and audience care is if it's there and doesn't break immersion. They notice when it's done badly, but don't notice it when it's done well. So while original designs and high craftsmanship will still be required, the 50 different cups you see in a game or in the background of a movie won't be made by an artist but just generated by AI with the artist baybysitting it to make sure it doesn't make a penis bottle, or a cup with self-intersecting handle. Like seriously, how many background chairs or cars do you really remember from any recent movie? Do you really pay attention to all the blood splatters, graffitis, dirt spots or puddles in the background of a shot? I know I don't, and I know even people working on those things don't, except when they specifically look for them for study. I'm pretty sure all that stuff is going to be delegated to AI pretty soon, we already use procedural modelling for that with programs like substaince painter or houdini, we already use photogrammetry and performance capture with ML to clean up the results. The job of the human artists will be focused on what people will see. Is it a chair sitting somewhere in a corner as a prop? Let an AI do it. Is it something that the player will interact with? That's for a human to do.
This is enlightening and also gave me another perspective on this whole situation. There are so many scenarios for use of AI like this , that I never even considered something like that. At the end of the day, the actions we take are all based on our taste, the economics and the time we have to get what we want. 100% such scenes would be taken care of by AI. Probably a few years away, that!
Makes you think when computational power and hardware can reach a threshold where we can apply this AI to video, we can literally create movies from a plot summary.
Sounds like the foundation for the matrix, where instead of just being a simple virtual environment, it's a whole universe that responds to your attention and intentions
All this means is that the role of humans will move from instrument players to conductors, and instead of creating detail, humans will create milue. It means that pieces of art will go up in scale and digital art, instead of taking up a screen or a wall paintings/images will become the size of entire neighbourhoods, all at high-resolution photo realistic detail. Images will no longer be viewable in one viewpoint, you will have to scroll. And perhaps static art will go away, and humans will have to stitch together timeliness into successive, related pieces of art. The human input in art will be creating the causal relationships. And thus art will move up from creating a frame to creating a framework and everyone will benifit from a whole extra conceptual framework - a gallery - available in art, something which was up until now constrained by time. So yes, AIs will take the work artists currently do, and allow the artist to focus their attention on how to connect several individual pieces of art into a greater piece of connected art. A few centuries ago, they said tractors would take away farm work from farmers as a group - what actually happened is that farmers as a group used those tractors to create bigger farms, and fewer individuals were farmers, but what those remaining farmers did do is planting hugely more land per farm.
Absolutely! Short term it might cause some pain but long term benefits would be amazing and we will adapt. Thank you for leaving such an insightful comment, it gave me an entirely new perspective
Thank you for watching and for giving feedback. This will help me improve, another viewer pointed out the same thing, I'm really sorry I didn't pick it up while editing the video. This is defo my priority #1 from now on, I will pay more attention to this!
@@beyondengineering @Beyond Engineering No need to be sorry my friend... What you did is great.. your priority should be making the cool content... which your doing... There are just a few little things that make it even better... and if anything I say helps.. then awesome.. Just keep making good stuff
100%. The video itself and the content is great; nice framing, nice lighting, clear understandable language in one of my favorite complex subjects, so I definitely liked and subscribed.
Thank you. Sorry about that,, I will fix this going forward. I write the script and read it so many times that when I was editing, I didn't pick up on this.
Google AI is very impressive, but this Video Audio speech and music are very loud, a little distracting, I had to click on the links to appreciate in silence...
I would like to see ai create niche artwork styles like glitch art or other abstract styles. When you get into that territory there is so much nuance between different styles. If you have your own style, you could try using an ai model to expand upon that style but these models are trained on millions of images so unless you have a million pieces that you made of your style it won't be able to perfectly replicate your style. Don't stop doing art just because ai is around, that would make life extremely boring.
Absolutely! Every artist has their own thing that makes them special, the only thing is average to bad artists would be replaced for content generation tasks
Yeah, AI art for now looks a bit nonsensical in it results. It always have some flaws that will make people to get tiered of it wery fast. Easy come - easy go. I am thinking on how can I use it for some 2D game and it might be ok for personal project. In general hiring artist is expensive and so people are not even trying. One of the strategy I can think of is to make a game with the generate art and later when it will look worthy - pay to an artist for refining of it.
For those worried about ai taking your art job, I wouldn't worry too much about it in the near future because as he mentioned, making large scale images with these models takes a huge amount of processing power. I have only seen certain ai algorithms that can go up to 12 megapixels, whereas I routinely work with images above 100 megapixels. If you try to print one of these images they are gonna look like crap or be too small to matter. Until we start training ai to make images using vectors or increase the processing power, that is useless in businesses that need the larger scale images such as graphic design agencies or marketing. This is the limitation right now but I'm sure at some point even that limit will be broken. It is only a matter of time. When that happens I'm sure ai will be so advanced that it is solving many problems that humans could never hope to and that there will be an absolute need for universal basic income since there would be almost nothing an ai couldn't do, so either way I wouldn't worry about your job. I would worry that humans might go extinct first.
I wouldn't worry about humans going extinct unless we have robo dogs with guns on their back supported by army of small drones we can't see or take down just waiting to hurt human beings...oh wait!
This is so cool. Makes you think though. To the AI the difference between a renaissance nude painting and pornography is not far apart: yet people will insist their is a difference. In a strange way the AI shows us a more pure reality, but then we humans enforce our biases into the system... It's as beautiful as it is horrifying... The human mind I mean. 😅
The two are not the same though. And this shows that AI still has a long way to go to understanding context in a human way. I don't think it's an objective opinion to say that all nudity is equivalent and the same thing. It's not. This is a dangerous inability to perceive differences. If we expect AI to take on security, military, and law enforcement duties then this blindspot can cost human lives potentially. I don't think it's pure as it is ignorant. Like a child. Also not pure, but just ignorant. We confuse ignorance, naivety, and lack of awareness with innocence, purity, and somehow objectivism.
I think you make a good point about one case where humans are selling art which is really just porn but you have conveniently ignored countless other cases. Think about those too before an AI becomes your God
A major difference. What was put into that piece by the human? The emotion, creativity, imagination, love, hard work, patience and drive. What was put into that piece by a computer? A word? Theres no meaning when its AI generated.
"Culturally meaningful" and "unbiased" are opposite ends of the semantic axis. It's utopian to think that there exists anything that is "unbiased", that also embodies understandable concepts in the cultural sphere.
It's a noble aim anyway, we are starting from scratch and creating new economies, doesn't hurt to be too careful especially when this would revolutionize everything like the industrial revolution
Right. I was thinking that eliminating bias is a bias itself. I understand why they think they should remove the pornographic and racist databases, but any intentional selection creates a bias. Trying to eliminate it is the same as enforcing a moral absolutism.
This is scary indeed. How many years will be needed to see AI able to do most complex 3D animation and special effects and somehow, replacing many production and post-production companies ..?
@dietrevich It's exciting but its definitely scary how far this tech has come and how quickly it can have an impact on so many jobs, wouldn't you agree?
@@beyondengineering Nobel laureate economist Milton Friedman: "All of the progress that the US has made over the last couple of centuries has come from unemployment. It has come from figuring out how to produce more goods with fewer workers, thereby releasing labor to be more productive in other areas. It has never come about through permanent unemployment, but temporary unemployment, in the process of shifting people from one area to another." In 1900, 40% of the total US population were farmers. Today it's 1%. Why isn't there 39% unemployment? Because humanity moved on to do other jobs. The new job possibilities are never-ending. You can start to worry about the end of new jobs once we have colonized every planet in the universe.
Voluntaryist Perhaps at a certain level of technology (in the distant future) we stop scaling *up* and begin to work with resources at hand to scale down. When *an* apple could feed hundreds. Macro-scale cultures combined with the scaling *up* of resources, physically. Who knows what the future holds, you’re right, we’ll still always be inventing stuff to do though, even if we don’t need to call them jobs anymore. Edited: Grammar, removed a sentence.
Thanks for the sub! I appreciate it, sorry I will fix this as I go along. I'm still new and learning, believe me I wish I hadn't made this mistake, had no idea so many people would watch it
Great video. I wish I Could use it. ALSO wish it was not forced into a family-friendly, censored content-warning direction. Imagine if Photoshop would block you from drawing violence.
All the people saying artists will be replaced by AI. Not exactly. A lot of what makes great art great is conceptual. Take Pollock's famous drip paintings for example. I'm sure that most people can splatter paint on a canvas. But it was pioneered as an artistic concept by Pollock. How about Piet Mondrian's squares? Almost any person, even with no artistic skills, can make similar paintings. But to conceive the idea in first place, and then to realize it - that is the art. You can tell a computer to generate a cubist painting of dogs, and it can. But the AI did not imagine or create cubism as a concept or style. For the AI, it's a reference point. A reference point to a style which was born from the mind of Picasso and others. Same is true for any period, movement, etc. AI can only create a renaissance style painting because it can look a renaissance paintings as a point of reference.
Hi Shane, its just mimicking understanding. Understanding comes from conciousness but either way this tech will get better more accurate and generate results that we will like and pay for. Less like a typical artist more like a calculator but it would replace a fair proportion of artists
As of now it looks to be a very well done photoshop mashup toy. Is this a sample of a future capability? Can text-to-image A.I eventually produce a more "conservative" image like those found on stock sites like Getty,etc. The rapid growth of AI is amazing.
@@beyondengineering I've used AI to help my process a bit. I generated images of a creature idea I had. They were awful, had no sense of proper design or understanding of anatomy and looked like bad photobashes....however....they gave me good ideas for how I would render out my painting, a neat idea for a skin/hide texture. So it will be an asset before a threat, for many many years to come. It won't be a threat until it comes along a whole tool suit that lets you take the character in an image, that it rendered out perfectly- and then change it's position in 3D space with a camera feature that re-renders the image so you can have it actually how you envision it. To let you morph slight aspects of it with text inputs and accurately re-create things you need it to when you need it to, in the way you need it to. The amount of technology required for things like that is so far away, I think artists can rest easy for some time yet.
I put these prompts into Craiyon(Formerly known as DALL-E mini) and the difference is both scary and amazing!!! Can't wait until it's available to the public in a few years from now...
I believe that in a near future we'll see videos created entirely by AIs with auto generated action scenes, sounds and dialogues between characters. That would be insane!
@@ViniCavin Damn, I can't wait and I believe if anyone who has the balls to make this sees this idea here or has even thought about it, we are possibly looking at something like that next year, all it takes is feed picture frames to the AI so it understands how to adjust it's artistic skills in drawing frames, dunno maybe I'm talking bullshit but I Don't think it's a big deal if we can have stuff like this already
So this is not public because it might include porn? what about simply inputting a command option like (np) at the end to tell the a.i. to not include any porn ? and if someone wants to create an exotic image intentionally, why not allow that? artist have been doing this for centuries, why censor a.i.? an analogy would be, painting canvas manufactures don't sell to the public, because someone might paint a nude, or something sexually explicit .
I can only speculate why they wouldn't let us use it, perhaps being extra careful with it or maybe they saw something in results where this was just not possible? Maybe the next iteration will allow that
My guess is that if they release it now, and a bunch of fake porn of real people is made, the backlash will be huge, and it will be harder to push the technology forward.
Why do you not hold on the images longer? We wanna see the images. How long do we think it will be before AI knows everything that humans know? - this afternoon or next week?
Maybe in the future,we can text to 3D animations or movies, actually I am very excited about this. Google openAI collaboration with 3D blender, image that how amazing.
I dont see how this will replace artists. Without and image, it could make anything. Without a power source, it wouldnt work at all. If it does “replace” the artist, we will do what we are best at doing…, evolving. Adaptation and innovation is what will make or break us all. This is all in a days work to a creative.
@@b-rse That's not evolution of ourselves though, that's obsolescence. There's no good reason to kill our species for the sake of some grandiose sense of "progress".
@@invertexyz I guess if you are a cyborg,, but last I checked, my mother is a human. She gave me purpose to be, not a man made object. Without food would not happen considering I’m an excellent cook and I have a green thumb. All blessings given to me by my mother… not a computer.
@@jayld5396 You are completely missing the point I was making. At the base level we are not much different. The human ego makes you feel you are something magical but in reality your brain is operating on what is essentially fuzzy computer logic, it's just an organic computer. Perhaps you're someone who believes there's some magical soul that makes a human human, in which case there is no further discussion to be had because one cannot argue with a belief that has no way of being verified. My brain has been given a prompt from you replying to my comment, I analyze that prompt, break it down into its statements and formalize a response based on the knowledge I have gained from my own brain training (AI training) through my life. The way humans behave only seems random because there are so many variables involved that is is hard to see the patterns most of the time, and a lot of the data to understand the pattern is contained in their brain, something you have no access to. I'd highly recommend watching the BBC Documentary "The secret life of chaos", so you can better understanding how complexity arises from simple systems.
We are building a horrible world where humans become docile and without purpose. We have lost the importance of hard work and thinking for ourselves. You had a glimpse of this during the pandemic where depression was on the rise. Now, AI will imagine for you, AI will drive for you, AI will paint for you, AI will cook, clean and shop for you! Enjoy becoming a vegetable with no purpose in life, ruled by those who own the AI 👏
The purpose of life is much more than all of these tasks combined. Most people however do not see that and don't think about that. Life is still worth living even if you lost your limbs, your friends, family and many other gifts we all make use of everyday. Even if AI will eventually do some of these tasks for you, we would be giving time to other things. I don't see any harm in tech like this if used properly and ethics were designed around it well enough in the first place
@@beyondengineering "I don't see any harm in tech like this if used properly and ethics were designed around it well enough in the first place"🤣🤣🤣🤣🤣 Iz U a bot? LOL
Midjourney: hold my beer I think MJ has the edge atm, but it'll be fascinating to see where AI based image creation takes us over the next few years. Will certainly enable even non-artists to unleash their creative potential.
This is very impressive and this whole technology is mind blowing.... i have to say however... I think that DALL-E2 has the edge. I think the images look more natural... and the artistic expession seems superior at least from what ive seen. Splitting hairs tho... this is all mad amazing.
DALL-E2 is amazing, I just preferred this look a bit better but obviously this isn't open to public. Real test would be if all images generated are as good as the ones that have been handpicked by google and shared with us all
nah man, from a digital artists perspective I assure you Midjourney is far superior when you know what you're doing. I did everything there, from concept art, to high detailed fantasy architecture to emulating the style of different famous artists, ink drawing etc etc. It is insane. The results with Dall-E never came up with such surprising "creativity". The most mindblowing thing about AI is not that it can replicate almost photorealistic stuff, it is that it comes up with combinations and styles we haven't really seen like this before. Surreal, yet highresolution...like straight from a dream.
@@reuterss306 I agree now that I have seen more... the new version is giving truly amazing results. Ive gotten myself onto the trial... when i get a chance im gonna dive in... Ill prob go for the unlimited subscription so i can just go nuts.... Im a digital artist too... loving the looks of midjourney.... Dalle-2 seems to excel at composite work... and can make some truly photorealistic results... but midjourney wins hands down for its sheer artistic quality.
Where can we buy or download Imagin or Dal-E? Just a comment too, you have a nice speaking voice, but I have high tone deafness, because of years in the NSW Police, so it is hard to hear you properly with the music, which drowns out your voice. Hope you don't mind me commenting...
Hi Ken, thank you for watching. I messed up on the audio on this one, honestly didn't expect this would take off like it did. I will fix this going forward because otherwise nobody would understand what I'm trying to say...
wow! Legally they don't have any other option though, it would be too complicated to give credit to each artist otherwise. It's a shame really, but every tech has a dark side I guess
So they decide to not let the public use this because it could display inappropriate content. Well... if they insist on nannying people like that, why don't they logically just remove Google Image Search too?
It's not nannying really. Morenabout ethics, what if someone used your face as a starting point and converted that to a porn image using this AI? Making it look super realistic? Deepfakes are a thing already, surely you can understand why they want to be cautious
@@beyondengineering Yes, that's a good point. Though, as a digital artist I think I could achieve that anyway in only a few minutes better than most AI. But yeah, maybe Google kinda don't want to take responsibility for such creations.
Wait. So that old, inaccurate representation of people scaling up a faraway image into a bigger one with great detail should be possible using Imagen's scale-up technique?
I mean, yes and no. It's still an imagination of what the detail was that was there and not truly was was there, so it's not something you'd want to use in the context that stuff is usually used in, crime investigation haha
Hey @Thai Jackson There are obvious ways this tech could go wrong and put people out of jobs and everything would be doom and gloom it's certainly a possibility BUT look back on history and how new revolutions created different economies and how companies that adapt to the circumstances still exist. I'm sure as someone whose earnings depend on Art like you do, you would have a different way of looking at it, perhaps a more accurate way of looking at it even. I am an optimist and if I were creating tech like this, I would certainly be creating it to help other human beings, not cheat them out of an earning. That's why I believe that for a very long time, tech like this would actually help people such as yourselves. It would help you generate ideas for art that you want to create or it would help your clients communicate with you in more effective manner. Sure, it might take away a section of business but those people won't pay you much today anyway...I'm an optimist TL;DR
I look forward to when this sort of thing gets applied to music. It'd be really cool to be like "Hey make me a song in this style, which sounds like [hums / beatboxes a melody]".
@@beyondengineering Mostly, I just want to save myself a lot of work while making songs. As is, I do the humming / beatboxing thing into a phone or something, then sequence those notes as midi, add layers, add drums, choose instruments, record the actual audio, add effects, arrange clips into a song, then mix and master for a final render... and it's a lot of work. So it'd be cool if an AI could do most of that for me... like if all I had to do was come up with the rough idea and overall flow, and make some aesthetic decisions, going into fine detail only when I want to override what the AI did.
This will not just remove artist's job but will remove imagination and artistic skill from human gene. FYI if a group of people is not using one end teeth, it will vanish after several generations. Evolution must work and this is what will happen.
@@FUnzzies1 I'm sure we can agree on a few objective principles that can then let us determine what is good for society and what is bad, shouldn't be this hard
Here’s the problem. Computing is an issue of capacity. Human brains aren’t as powerful as the processing power of computers. Art on the other hand is the celebration of human capacity. It inherently can’t be replaced by machines. I think it’s a great tool for artists, but we already have elements of art that are technically better than what humans can produce. A camera is better at landscapes than an impressionist, but the value of the painting is the acknowledgment that it is human.
@Huck yes absolutely but a lot of things that some people get paid to make today will be replaced by AIs like this in the near future. Of course, human beings will use this as a tool to create even better art or a different art form we don't have yet
How about the blurred line of copyright in this images? A lot of them come from data sets from the internet at stock photo libraries. As an artist or designer you would have to pay royalties for this even if you photo bash them if you want to sell your creation. This models in the other hand benefit from the entire history of human image creation and might profit out of it like in the case of dalle2. This is something without a precedent that might have to be settle down in court down the line of Ai existence.
As a human artist. You learn from the work of others and absorb different styles from them and then produce your own work. This doesn't mean that you violated any copyrights by learning. AI is learning not copying.
It's a fair point but that kinda thing is hard to settle in court plus how would you pay and divide up your payment for thousands of people at the same time every second that a new image is produced?
Thank god im doing 3D art, there is still quite a long path for ai to replace that, however most 2d artist are gone beyond this point which is a bit depressing to say at least.
I don't think that I would feel at peace from just that. That is just the next brand of art that would be targeted with this tech. Anyways no need to be depressed, we will always have humans creating art, there would always be a market for that
@@RemyDovianus complex 3d is harder to make with ai. And animation is even more advanced. We can only see static images made by ai. Im not saying it won't be replaced im just saying that it's still not as close as the 2D one.
LOL, Haven't you seen the AI creating buildings and assets in 3D from text input? It's already here. But again. You have to understand these things are trained on pre existing assets. They arent capable of creating things like we are from imagination. When you ask for a cat made from waffles swinging from the eifle tower. It's jsut merging things it knows of. It couldn't create Samus Arrans new armour or some type of new alien spaceship etc etc. It learns from example. No imagination. So think of it more like this. You design a few assets like scenery pieces. You tell it this is for a wal this is a window blah blah. Then when you want a city made up you say I want a building with 45 windows and a big front door like a hotel with a helipad on the roof. It will use your new textures and assets to build from what it knows. Then you can tidy up what it has produced to make it how you imagined. Thats the future of your job for some aspects. It's not here to replace. It's a tool. And it's not that far off. There is currently about three 3d asset generators out there/ Nvidia has a fair few going on. And Ubisoft have invested a ton into it. Will this inevitably hurt 3D artists? Probably. But mainly the low tier asset builders that hate they're jobs because they spend countless hours making silly things to fill out a world. So jsut don't be that guy. lol
@@beyondengineering the future of gaming and art will be actualy about how good are you with your vocabulary and thoughts expression and making sentences that express your vision the most precise
At the end of the day, this exercise teaches us that, in many cases, looking trippy in public isn’t as bad as we believe. Sure, we might get some looks and giggles, but these don’t hurt, now do they? The Cynic philosopher Diogenes of Sinope used to sleep in a barrel on the streets, walking around in ragged clothes, ‘pleasured’ himself in public, and tripped balls with people who walked past him. He fucked what people thought of him. Therefore, no one was able to hurt him.
Hi, you definitely can. The Imagen AI subreddit takes in requests and does generate images! You could also try Midjourney and DALL-E2 and make a request there to see what comes up!
Yes, Dall-E will allow that. Previously you could only get on a waiting list and then hope to get picked (I did). But as of yesterday they announced they are going into next stage of development and will be looking for a million users to add over the coming weeks.
I think its more the photorealism compared to DALL-E 2 for me. Again, this is subjective and there is no right answer. Everyone would have their own preference. With DALL-E2, the great images are more of a cartoonish nature than realistic camera like ones IMO
My FAVORITE part of videos like this (and others) is the extremely short length of time they display the description along with the picture so you don't really have time to read it before the GENIUS video creator moves along to the next image/description. F'cking so glad A.I. will be far smarter.
Hi you can always pause the video or slow it down, unfortunately most people will not even watch if the video was static for too long because overstimulation. Also the whole idea is to ignite your interest, you can always go read more. Anyway I hope AI would be much nicer than folks like you
While interesting, we can try DALL-E2, and I have generated a few dozen images. We cannot try IMAGEN, and it seems I see the same sample images from IMAGEN over and over in the various media reviews.
That's a valid criticism and perhaps what makes DALL-E2 better because everyone can see its results for themselves, although all these image generators do this to some extent, of the images generated, only a few are shown in the result
Hi Dina, thank you. This feedback is really important for me. Judging by the views on my last two videos, I am inclined to agree with you. Please watch the new video when it comes out, hopefully it’s better than this one
It's about text to image. So it felt weird that the image comes up first on the right side of the screen. Then the text comes in, in a distracting animation, and by the time it's read, the image has gone. Had to keep pausing it.
Please: Less hand gestures! It distracted me quite profoundly in concentrating on the information you are trying to convey. Otherwise good video! Thank for the info and the hard work that went into making this video.
Hi Daniel, maybe I overdid it during this video. I was feeling really bad at the time as I recorded on the day I had peak Covid infection. Normally more hand gestures make me less of a robot in front of the camera.
Oh my god, people worrying about bias is stupid. Everyone has biases, that's just a part of life. Don't screw it up by trying to get rid of things you don't like personally. Imagine making a creative AI based on all of human data and then screwing it up by censoring it and filtering it.
Yes there are some biases that are subjective but some are objectively wrong but it takes a lot to recognize objective right or wrong... anyways a lot more complicated issue than how you would rather have us believe
😂 that's a nice way to put it. I'm sorry, I messed up on the audio. I will fix this in future. Thank you for watching and giving me feedback ☺️ it helps me improve
3:20 - Funny but when I randomly stopped this image the shown text was partially missing (only went up to "and Gr") and I guessed the missing letters and words of the 2 first ends of lines just like this T5 is doing..
