Above: Detail of “beautiful female in futuristic bodysuit (teddy) with transparent plastic hood” generated using Stable Diffusion by Ian Reid (@theironicmnemonic on Instagram).
Two Trinidad and Tobago artists, Ian Reid, art director of Reid Designs and brand strategist/creative director Anthony Burnley agreed to answer questions about their exploration of art generated by Artificial Intelligence (AI).
What led you to try AI driven art creation?
Reid: I can’t pinpoint the exact moment but I remember a retouch photoshop artist by the name of Pratik Naik, who came down to Trinidad for a seminar a few years ago started posting Midjourney renders and I was intrigued. I immediately realised my job as a graphic artist was coming to an end. I was also put on the path by another local graphic artist called Sherwin Marcelle, he and I have been exploring this tech together.
Burnley: I like to keep abreast of what’s going in the international creative world. You need to know what’s happening in order to adapt, and what better way than by trying things as they develop.
Which services/apps/portals have you used to create AI art and how would you describe the results?
Reid: I started with Midjourney but due to financial constraints I couldn’t pay the subscription free. So I signed up for Dall-E and open.ai. Eventually I realised I could use Google Collab for a few hours a day using the Disco Diffusion tool and thats when I really started seeing good results. Now I am using invoke.ai locally on my Mac to use the Stable Diffusion 1.5 model.
Burnley: I tried midjourney.com, after reading some reviews. It seemed to be the more dynamic while having high usability for first-timers. Which is exactly what I wanted.
What inputs did you use to create the works?
Reid: With Dall-E and open.ai its simple stuff. I am interested in rendering images that clients continue to request that we can’t find in [stock image agency] Shutterstock. “Young Indian woman from Trinidad using a phone” for example. My goal is to see if I can slot into a niche where I can provide a service to clients who can’t find a solution anywhere else.
For Disco Diffusion I have been using it to create largely original paintings so that would be “Painting of Maracas/Mayaro/Chaguaramas in watercolor/oil etc. at various times and “various points of view.” I rarely use “in the style of” an artist and let the diffuser interpret it how it wants. No matter what I do, I get some interesting results.
Burnley: I used simple inputs, such as, futuristic Egypt, Victorian robots…things like that.
Did you find the experience complex or costly?
Reid: On a scale of 1-10 for complex: 5. Once you figure out the learning curve you’re off like a sports car. However to TRULY get the results you want you have to know how to use the language and prompts. So the learning curve goes up exponentially.
Price 1-10 I would say 3. MJ is cheap but I am not willing to pay for this until I can make some money first. Stable and Disco are free with caveats (you need to pay Google to run the server all day) and for Stable Diffusion you need a very expensive graphics card to really generate images fast. Until it starts generating income it’s still just a “hobby” that I won’t be putting money into until I’m sure I can recover the cost.
Burnley: The experience was a bit challenging, as with midjourney, you don’t go directly to their site and begin experimenting. You first have to create a Discord account, figure that out, (as I never used Discord before) and then find the MidJourney server. Then, you’re off to the races.
As for cost, I used the free option, however, I think the cost is minimal. Something like US$35 per month I think.
As an artist, what do you think about this new technology development?
Reid: Twenty-five years ago I started doing graphic art on a Mac. I witnessed paste up artists become obsolete and redundant very quickly because a new tool arrived to change the landscape. In the ensuing years graphic art has largely become digital and with Canva everyone can “do it”. The tool is no longer reserved for a tiny elite. So the market and the landscape has changed again.
AI is a natural progression of this. You can either rail and rant and say that the hand held drill is making people skilled at using manual crank drills obsolete. Or learn to use the new drill to create things.
I think in five years sites like Fiverr and Upwork will find themselves in a very precarious position because new sites like Canva and the like will replace them since all you have to do is say “Build me 10 512 x 512 [pixel] posts for my coffee brand showing men and women in the shop.” And boom, you got a layout template generated ready to use with copy. Would cost you a few cents to Generate every month. No need for the Fiverr artist.
Burnley: I am very excited about the development of AI tech. It blows my mind every time. It’s crazy what is being developed and how the resulting tech seems to speed up exponentially when the masses engage with it. It’s a great time to be alive.
Do you believe it will impact the creative community? Would you make use of AI created art?
Reid: Yes, I will be making use of AI art. What the new landscape will look like by the end of 2023 is a guess. But the greatest impact I see for local designers and photographers is you will no longer need to pay a mas designer to render costume ideas. I will do that for you by being the prompt engineer.
Right now I am learning everything there needs to know about prompting so I can get what is requested. I also plan to use the tech to start producing my own illustrated books. Something I didn’t have the resources to do before. Its scary, exciting and raises so many philosophical questions. It’s a good time to be alive I think.
The whole question of copyright and “is it right to use other people’s images as a basis for these new images” is a very interesting thing. It will take a while before this settles and I’m pretty sure that in time the AI will no longer need to scrape existing images anymore, it will know what a “Mark Lyndersay photo” is and if Mark Lyndersay has said “no to AI” then it would not render an image in that style.
BUT can you truly copyright a style? No idea. Personally I would be very upset if I saw “paintings” that LOOK like my work but isn’t being sold by me, but what can i do? Change the style? Maybe. Things to think about and have a debate over. The possibilities are still almost infinite…
Burnley: For my personal art, I will never use it. Because it’s not my art, it’s otherwise generated. Even if I am clever with the prompts I come up with, it’s not generated by me.
As for my commercial profession, by all means. I see it as a form of clip art on steroids. Or stock art, or stock photography, they all come from a similar place and seeks to solve similar problems. The challenge is, who will use it the best. This is where creativity will drive the outcomes.
I think it would be used in a novelty way at the beginning, (just like clip art) and then it would begin to be about who uses it best. So no, I’m not too concerned about it in that sense.
Also, it would take a while for before local brands/clients even understand what’s going on, far less accept it in local advertising.
Our weekly newsletter lets you know what’s new in technology on our website.
Check your inbox or spam folder to confirm your subscription.
TT Computer Society
Copyright TechNewsTT, 2022