Artificial intelligence (AI) programs that allow us to generate images from textual descriptions they are becoming an invaluable creative tool. Over the past few months we’ve marveled (and amused) at the creations of Craiyon, DALL-E 2, Midjourney, and StableDiffusion, to name a few, but it’s only now that we’re starting to see more complex usage.
Some are combining the power of these AI imaging programs with other tools to do a great job. William Buchwaltera former machine learning engineer at Microsoft Research has used the generative capabilities of StableDiffusion with a Photoshop plugin called Alpaca to create an amazing “digital painting” with his computer.
AI’s, creativity and challenges
Buchwalter is not a cartoonist or digital illustrator. In fact, confesses that he does not know how to use a drawing tablet. However, he does seem to have the ability for his chosen imager to be able to produce exactly what he wants. Instead of drawing “a house on a hill”, this young man asks StableDiffusion. He does the same with, for example, “a small farm” and other things that he imagines.
* Of which I’ve got about 0 as can be seen below. pic.twitter.com/3L1vteAEKA
— William Buchwalter (@wbuchw) August 26, 2022
As the image generator goes about its work, Buchwalter is in charge of placing the images in different parts of a Photoshop canvas and, using his editing skills, he is in charge of combining them. The result? An entirely new creation that human abilities or StableDiffusion alone would not have been able to generate.
The artist Martin Nebelong is another one interested in exploring and exploiting the capabilities of new generation tools powered by AI. In his case, he does apply drawing techniques, but he also uses StableDiffusion, but with a collaborative art platform called Artbreeder. Some of his most recent work can be seen on his Twitter profilelike the one below.
How to generate an owl?
Ai image generation tools are evolving fast these days, and here’s one of the latest ones. The new Artbreeder beta is powered by Stable Diffusion, and it’s a different take on what tools like this could become.#stablediffusion #artbreeder pic.twitter.com/iH4mQkYqiC
— Martin Nebelong (@MartinNebelong) August 25, 2022
With these kinds of possibilities strongly emerging, keep in mind that these are just a few examples, different opinions arise. There are those who are in favor, and ensure that the new tools will help democratize visual storytelling, others regret having studied artperhaps fearing that in the future it will be the machines that do their work.
But it is not the only dilemma with this issue. A key point that we have also explored here at EuroXlivehas to do with the copyright of the images. The AI’s used for the generation of images have been trained with huge data sets coming from the internet which can include work from professional photographers and digital illustrators.
Are we facing copyright problems? What about the combined images? The World Intellectual Property Organization (WIPO) points out that only people can be authors of creative works. So since an AI is not a person there would be no infraction. However, these organizations are considering adapting their standards to these new times where image generators acquire an increasingly leading role.
Over time we will see what role these image generation systems will play and they will become the norm for future art creations. The truth is that artificial intelligence is advancing by leaps and bounds. Would you have imagined such a scenario a couple of years ago? Possibly not. At this rate, the most surprising could still be to come.
In Xataka | AIs have a problem: they are opaque and closed. BLOOM is the great open source project that wants to change everything