The world of all things creative, whether photography, painting, digital art, product design, cartoons, etc. is about to be blown to pieces by a tornado of unprecedented proportions: the emergence of AI.
For the record, I created all the images in this post with the help of one of the AI’s mentioned below*. I then enlarged them because the current output sizes are low, but no other editing was applied. Those tools have been trained to compare the relationship between text and images across a dataset of billions of web images and are able to create a graphic rendition of any text prompt which includes a description and stylistic preferences. I have noted the description part of these prompts in each caption.
After finding the tip of an iceberg that had eluded me until a few weeks ago, I have been exploring the fascinating emerging trend of AI-driven art creation tools. There are probably some which remain off my radar but the major contenders I have toyed with are DALL-E 2 (named after Disney/Pixar’s Wall-E and Salvatore Dali), its sibling Craiyon, the art-centric and marvelous Midjourney, and newly released and mind-blowing Stable Diffusion by Stability.ai via the DreamStudio app.
I will obviously explore all these and the possibilities they offer in depth, but I can already summarize and make a bold statement: right here, right now, under our noses, these tools are in the very rapid process of changing the creative landscape forever, and whether it is for better or worse, I cannot say yet.
We are talking about rapid change here, not occurring over decades or even years, but months and weeks! And while I adore my cameras and all the gear that goes with them, I suddenly wonder how long those prized possessions will retain their value and purpose before plunging into obsoleteness.
After all, other than forcing one to move, explore, travel and document, a camera basically does the same thing these new AI’s do: it renders a vision of the world. If today I go on a walk to photograph butterflies, I have the option to then edit my images in advanced software programs, tweaking composition and exposure, removing unwanted or distracting elements, even replacing the sky with something more dramatic (the latter being a recent trend in photo editing tools for the last few years, already inching towards an “AI” deep-learning approach). This has existed for a long time now. But photography is still based on recording the so-called reality.
But tomorrow, if the weather turns nasty and I decide to hunker down at home, I will be able to fire up my AI, write a polished description of what I wish to render and have the AI generate it for me while I lounge in my pajamas, in infinite variations, until it hits the nail on the head.
Deep fakes, highly-realistic forgeries, convincing approximation, AI creations will take the world by storm. These images in here are not perfect by any means, but consider that we are in the infancy of this new medium.
Generating successful art is a game of trial an error at this time, of course, and it is likely that none of the images in here were output as they are on the first run. I had to tweak descriptions and styles repeatedly. The digital art category clearly wins the first round. Photorealistic images will be harder to master, but they are coming.
The creative industry is going to be unraveled, and the stock photography world along with it. Who will want to hire a graphic designer, a photographer, or subscribe to a stock agency’s content when they can ask an AI to create what they need in a few minutes and for a fraction of the cost?
Sure, the nimblest graphic designers and photographers are going to pivot and adopt AI too, to try and survive the tidal wave. How exactly they will achieve that remains to be seen. Here are my suggestions for new specialties and professional titles: “Senior Prompt Engineer”, “Stock AI Art Distributor”, “Text-To-Image Artist”, “Dream Rendering Technician”—for the latter, when you wake up from this horrible nightmare you struggle to put words on to describe to a loved one, you can now write a sentence about it and have a DRT issue a fully rendered image of it, and images and worth a thousand words as we all know (thanks to Marie for that idea).
The ethical, legal and commercial implications are going to be mind-boggling. Who owns what, what claims can be made, how much money is such a creation worth, does the creator’s identity even matter anymore?
Make no mistake about it, hauntingly beautiful imaginary is going to be created with absolutely no graphical nor design work being done by creators who will not even need to know how to draw, paint or photograph. This changes everything. It’s history in the making. Or breaking.
I have created a showcase of the art I am generating through the three major Ai’s and uploaded it here for documentary purposes. I decided not to host this on my own website since I feel this content really isn’t my own, even though I pretty much have the right to do whatever I want with it, as per the respective TOS’s. I will be adding to the showcase along the way, as both my prompt engineering skills and the AI themselves evolve. Do visit it again!
Our hope is that DALL·E 2 will empower people to express themselves creatively. DALL·E 2 also helps us understand how advanced AI systems see and understand our world, which is critical to our mission of creating AI that benefits humanity.
OpenAI
AI by the people, for the people – Designing and implementing solutions using collective intelligence and augmented technology. Stable Diffusion is a text-to-image model that will empower billions of people to create stunning art within seconds. It is a breakthrough in speed and quality meaning that it can run on consumer GPUs.
Stability AI
Midjourney is an independent research lab exploring new mediums of thought and expanding the imaginative powers of the human species.
Midjourney
• ♦ •
*The term AI for Artificial Intelligence has been trending and is used in many ways across the technoloy and image editing industries. Whether a sophisticated script or program with deep learning capability can be called AI is up to anyone’s interpretation.
Comments
Marie
Germaine Versailles
Frank@Mound