The 2023 Microsoft Work Trend Index report, “Will AI Fix Work?,” landed simply weeks after the launch of Bing Image Creator, a brand new design instrument powered by a sophisticated model of DALL-E. When it got here to creating artwork for the report, the staff noticed a novel alternative to discover how a human artist would incorporate it into their course of. Would it assist generate concepts? What prompts can be best? And how would utilizing AI as a co-collaborator influence the ultimate product?
The artist for the job: Los Angeles-based Jon Han. Han makes a speciality of creating thrilling futuristic worlds with out the same old dystopian tropes. Because the report information confirmed that most individuals are looking forward to AI to assist ease the burden of labor, his model felt like precisely the appropriate strategy to convey the findings of the WTI report back to life.
Jon began by feeding prompts into Bing Image Creator, testing how he might collaborate with it. Using AI impressed a extra iterative strategy than what’s typical, as a result of taking part in round with it was fast and straightforward. “The process for this project was an experimental one,” he says. “Just trying to see what it’s capable of.”
Here, Han takes us inside that course of.
Prompt 1:
Person wanting on a bridge of panels exhibiting totally different landscapes
Jon Han: In my regular course of, I learn a inventive temporary, take into consideration the idea and visualize it in my head. So within the early stage right here, I wished to see how the instrument works with concepting concepts, and distinction that with my regular course of. I wished to see how the picture creator would visualize what I’m pondering.
You can see that the instrument is utilizing plenty of photo-reference pictures, however typically my work is just not as stiff—it’s extra summary. So I began attempting to push it to go to the specifics that I used to be searching for visually.
Prompt 2:
Person within the foreground + in entrance are geometric panels which are materializing digitally
Jon Han: Here I shifted away from bridges and tried to push for a extra summary picture and a digital impact. I preferred the shapes of the panels and the digital really feel. But it appeared an excessive amount of like a sci-fi film. I wanted a happier temper.
Prompt 3
Person within the foreground + in entrance are geometric panels which are materializing digitally + panels come collectively to create a lush panorama
Jon Han: This time, I added in ‘lush landscape,’ which made the picture extra atmospheric, not so darkish and digital. But the temperature of the image was a little bit too cool. I put within the phrase “warm,” and it got here out like all the pieces was on fireplace. Much like some other instrument, like Photoshop or oil portray software program, there’s methods to some of these things, the place sure prompts create sure varieties of pictures. So there’s a studying curve to determining easy methods to do these minor changes to get the specifics you need.
Prompt 4:
Person within the foreground + in entrance are geometric panels which are materializing digitally + panels come collectively to create a lush panorama + within the model of jon han
Jon Han: This time I added my identify to the immediate to see if it will be capable of reference my specific model based mostly on my revealed illustrations. Adding my identify made the particular person smaller and obtained the appropriate viewpoint, seeing the particular person from the again. I do a number of compositions the place I do a small determine in a panorama. But it didn’t really feel hyper-referenced to my work, in order that was good.
There are a number of combined emotions amongst artists on AI-generated pictures, issues whether or not these methods could possibly be referencing artists’ work with out copyright. I attempted to steer away from something that appears like an artist I do know. But one factor about this instrument, in comparison with different AI instruments I’ve tried, is that it didn’t look like it was referencing somebody’s work. This was a superb endpoint for the concepting stage.
Concept sketch created by Jon Han
Jon Han: After these rounds of experimentation I had a transparent picture in my head of what I wished, so I created this idea sketch by hand. The full image I wished to create was too advanced to speak to the picture creator. So quite than convey AI-generated pictures straight into this sketch, I made a decision to make use of the instrument [or the AI] in later levels to generate choices for specific parts of the illustration, like timber or sky.
Normally, you would possibly do an web seek for photographs of issues like that, for reference. An artwork instructor as soon as taught me: by no means use the primary few pictures from an web search as your reference, as a result of everybody’s going to make use of these. So with the picture creator, it’s good that these are going to be generated a distinct manner.
Final illustration created by Jon Han
Jon Han: Before creating the full-color illustration, I used the picture creator to check totally different sky colours. I additionally tried pixelated timber and wireframe timber, searching for one thing fascinating that doesn’t have an excessive amount of of an AI really feel to it. This was all experimentation, discovering limitations, and experiments with management.
Han says that for lots of people, AI picture creators are “kind of like a surprise box.” In his expertise, they’re most helpful if you don’t know precisely what you need. In a case the place a creator already has the picture they need of their head, a picture generator may be good for iterating particular parts.
We are solely simply starting to see how next-generation AI can match into the inventive course of, and it’ll definitely look totally different for each artist and creator. From the prompts individuals use to information the instruments to the diploma that AI does—or doesn’t—affect the ultimate product, all the pieces is up for exploration.