AI’s reality mastery powers have been increasingly busy over the past decade. We had 3D dioramas based on computer vision; avant-garde style transfer; viral photorealistic selfie setting, selfie editing, face swapping and – ofc – deepfakes; and a lot of frivolous (and hilarious) fun with selfie filters (ohhai cartoon lens ‘Disneyfying’!) in between.
The AI-powered visual remix has shown, time and time again, that it can grab attention. Although keeping an eye out once the novelty of an AI-generated effect wears off can be more difficult. (Selfie editing apps don’t have this problem, beware; there is a perpetual demand for machine learning as a reality. booster.)
What is most remarkable about the developments of synthetic media compatible with AI during this period is the rate at which these visual effects increased, aided by ever more powerful mobile processing hardware.
Wait times for a finished result can now be essentially instantaneous, which is a game-changer for producing (and potentially monetizing) the creativity and power of neural networks and Generative Adversarial Networks (GANs). Aka, the Machine learning frameworks doing retouching, cropping, or even generative modeling, jumping out of a human prompt for their inspiration.
And while most of the app-based visual remixes of the past decade have focused on retouching / restyling / augmenting versus purely AI-based image generation, that too is emerging. change.
Wombo, a Canadian startup that has drawn attention to its eponymous AI-enabled lip-sync video app, recently launched another app, called Dream (iOS and Android), which uses AI to create “works of art.” original art ”- based on a prompt text.
That’s great, Great simple: you just describe what you want him to paint – say “A terrifying tree” or “The worst sandwich in history” – choose a style from the selection offered (Mystic, Baroque, Fantasy Art, Steampunk, etc.). ), or choose “no style” and press create.
Then, literally within seconds – I counted
You can’t even get bored during those few seconds of creation, because you get a glimpse of the AI at work:, to arrive, practically breathless, at another finished composition.
Some of these generated artwork looks pretty impressive. Some… not so much.
But of course two prompts don’t generate the same image. So you can keep requesting a new image from the same prompt until you like the look of what you see.
In short, Christmas card artists and pulp fiction illustrators can probably retire now.
Everyone is an “artist” now.
That said, real artists should have less to worry about. Particularly because the art created by a human brain and body will only gain value once the world is inundated with “mechanical art”. (Just as every NFT struck dilutes the meaning of the expression “digital art” …)
The quality of the “art” of the Dream app is definitely variable. Longer and more complex prompts seem to confuse him. So the quality of the output may depend on what you ask it to draw.
While his “style”, if it can be considered as having a single global style in the midst of so much pastiche, tends more towards the abstract and the distortion than towards the specific and the precise. Thus, portrait requests will not be rendered photorealistic. And it is generally more comfortable to describe the fantastic than the real. (A “Madonna and Child” prompt served a work closer to an infamous Spanish church restoration failure than a crypto-Botticelli, for example.)
But the production speed is impressive. Terrifying slash.
As soon as a new work of art appears, the app wastes no time trying to sell it – by showing an option to “buy a print”, which links to its online store and looks like a convenient way to turning a visual trick into real income. (He offers “Custom Wombo Dream Print[s]”which start at $ 20 for a matte poster or $ 45 for a framed print.)
If the startup can turn around 20 seconds of processing into over $ 20 in revenue, that could be a good little pipeline of cash.
While most people have limited wall space on which to hang any type of art, let alone images generated by, um, some insane machine – so most of these random creations will remain firmly virtual. (“The Art of AI” might be a perfect NFT fodder, though…)
Where will “the art of AI” fall in the fashion and cultural value issues is certainly an interesting question to ponder.
This is of course superior to clip art or stock photos. And the release of the Dream app may also be more interesting than the average “art” print you might buy from IKEA. But the results can also be rather smelly – or derivative – or void – or just plain weird.
Well, is that art? Or is it just a visual output of a mathematical process? An abstraction of human creative competence that cannot translate a real emotion or a sense of identity or soul because the code does not have any of those things? He just does as he’s told.
And do you really want to hang a coded abstraction on your walls?
I mean, maybe? If it’s above all aesthetic. Anyway, is that art or just wallpaper? Maybe Wombo should sell Dream AI wallpaper rolls or mouse pads and printed t-shirts (merch), rather than “art” prints …
Much to think about.
Some things are clear: AI generated art is amazingly fun to play. It’s kind of visual catnip. A toy for the imagination.
It is also, without a doubt, here to stay. And AI models will continue to improve – depending on what we mean by “better” around a subject as subjective as art. (Perhaps generative art models will give more successful results by bringing the user more fully into the creative process – giving them tools to customize and manipulate the machine’s outputs so that they can be (end ) adjusted until they are closer to what the person imagined, or otherwise to feel more personally unique and meaningful. Or, in other words, a more hybrid creation process can produce more powerful and moving artistic outputs.)
There will also be dozens of these artistic AIs, each producing different visual output “flavors” and “characters” – derived from their training data. Or even – if you like – artistic AIs with different “styles”. (But maybe “specialties” is closer to the coded mark.)
There are a number of other GAN-based image generation AI tools – and I admit to being A big fan Pixray’s system (pixel art outputs are particularly cute) although its processing speeds are much, much slower – but Wombo seems to have been the fastest to apply and monetize this technology.
The next decade of machine learning that distorts reality will be quite a journey.