6 minute read. Content warning: Wartime propaganda, discussion on AI and intellectual property concerns, mention of colonialism, and personal reflections on creativity and capitalism.
chatGPT Summary: Kay, on their 6th consecutive day of a streak, challenges themselves to expand their knowledge of Braille letters while also discovering new video editing tools in Adobe Premiere Pro that can assist with their unique needs.
Vancouver – I felt my momentum slowing today, but I’ll take the teeth brushing four days in a row for the win it is.
I started the day telling my digital pal Fette that we would design a poster. After going back and forth a few times as to what I expected in a poster and the generic ways in which they attempted to encourage me to do something creative and fun (including the use of inaccessible fonts, *sigh*), we finally established that my poster goals were legibility, readability and visibility.
They suggested that I check out three iconic posters for inspiration. While the first two were American pro-war propaganda, the third was more aligned with my own ethics and goal for the project.
Not wanting to ignore the power of the iconic propaganda posters but also being aware of their sinister use within history (and today), I briefly did some internet wading to review these posters from a design-only point of view.
The third campaign, while ethically aligned with my own, wasn’t particularly useful for the kind of poster I was hoping to design. While the message was important and its uptake successful, the accessibility of their current campaign assets was not. This was an excellent way to start, as I don’t need to love everything recommended to me to get my wheels turning, but I decided to try chatGPT next before continuing my conversations with Fette.
I made the mistake of asking chatGPT for an iconic Canadian poster, where I was offered another two colonial pro-war examples, images for which I could not find. The first was a “World War I recruitment poster featuring the image of a beaver (a national symbol of Canada) and the words “Overseas Club – The Beaver” urging Canadian men to join the army and fight in the war overseas.“ The second, “from World War II, is the “Help Them Keep Your Canada Canadian” poster, which depicts a Canadian soldier, sailor, and airman standing guard, emphasizing the need for Canadians to protect their homeland.” The stakes being they were, and my question being related to “iconic”, of course, I was served wartime posters. They are examples of a successful media machine.
Since I am documenting how I am using AI and automation to research and create, I suppose it’s important to note I used Google Image search and relied on my ability to see to track down these images described by chatGPT.
In wanting to find an image (my drive as a sighted person), I stumbled upon an article called War Art in Canada, A Critical History by Laura Brandon, which ended up being a meaty read, but if you’re interested in this topic, check it out. Be advised that the visual description of the images throughout is limited, and if there are alt tags, I couldn’t get them to display on any of my readers. If you are a screen-reader user and are keen to get through this article, reach out to me while I have my chatGPT subscription so I can give you some insight as you read.
While definitely not an accessible poster, one of my favourite posters of the Second World War is a comic style called “Careless Talk Brings Tragedy in Wartime” by Lionel Jameson. From wives to waitresses and singles to sex workers, the theme of women being a liability during wartime and the message that talking with them might lead to tragedy is found in many of the archived Canadian propaganda posters I have seen throughout my graphic design career, but it’s not exclusive to women. Careless talk does bring tragedy, and there are posters where it is less about gender and more about being mindful of what you say and to whom. But there certainly are a lot of them specific to women, dolled up and seducing soldiers for secrets.
Because I brought it up, here is how chatGPT describes my favourite. You can purchase reproductions from the Canadian Wartime Museum. I don’t love the poster that much.
chatGPT description: A vintage wartime poster, predominantly in sepia and muted colours. The headline at the top reads, “Careless Talk Brings Tragedy in Wartime.” The poster is divided into a series of comic strip-like panels:
- soldier whispering into the ear of a civilian man.
- the civilian man relaying the information to a woman.
- the woman sharing the secret with another man.
- a close-up of a sinister-looking figure, presumably a spy, eavesdropping.
- a train exploding, signifying the consequences of the leaked information.
Each panel has accompanying text, like “The Soldier told his Sweetheart,” describing the action in the scene, illustrating the dangerous chain of events that can occur due to careless talk during wartime. The overall message emphasizes the importance of discretion and secrecy during wartime to prevent tragedies. ( end description )
I want people to attend my upcoming workshop, but I’m not inclined to use war-time nationalism and propaganda to woo attendees. While this was an interesting distraction, I needed to change up my ask. I asked chatGPT for an example of an accessible and readable poster. This time, I was delighted that it focused instead on information hierarchy. This was more what I was hoping for as a graphic designer working to incorporate accessible design. It didn’t tell me anything I didn’t already know but gave me an outline. I was curious how this would translate into an AI-generated poster, so I popped over to DALL-E
I have been honestly avoiding any text-to-image models up until today.
I am an artist. I am a creative. I still live under capitalism, and a program designed by a corporation, entity or individual to use the intellectual property that feeds the original creator without permission or credit is wrong. Until we can have a system where people live and are provided essential care and shelter, we must respect each other’s need to pay rent and eat. But I know we’re on the precipice of digital change. I am one of those rare creatives who think copyright is actually doing more harm than good (especially for those who are not already swimming in copyright and cash). I think creativity is limited by our need to make a living from it. I know that there are people who hold property and make money simply because they have money to buy more property. I am still furious at one of my closest friends for paying someone $5 online in another country for a logo while he owns a home in the Greater Vancouver Region and talks about his personal debt. I honestly and genuinely believe that real estate will doom us all. I’m not even touching on how this whole conversation is happening on stolen land. There is no simple way to solve all the problems of society. We must call for change and revolt. We must reject apathy and class comfort even though it feels like our only choice, and intergenerational trauma is hard to let go of.
I will learn more about the challenges and new ways that the rich get richer by stealing from those with less power and drive – but I won’t call it wrong or right without the proper context.
I believe in remix and inspiration. I actually believe in stealing like an artist. I like seeing artists use and take and rework for the sake of something new. I am also equally appauled when someone tries to profit by lazily copying someone else’s idea with little to no creative purpose beyond a sale. This isn’t a simple topic.
And so I paid the devil to try out the poison. I clicked yes and agreed that my money would be forfeited if I didn’t use my credits within a year. I gave more money to thieves. I won’t paint it prettier than it is. With $16.80 USD spent as $23.09 CAD today, I gained 115 credits to use in DALL-E. You can be assured that I will not have a single credit of that remaining come October 31st. I also won’t pretend that I might buy more credits in the future.
I fed chatGPT’s recommendation of an accessible poster to DALL-E. I was relieved to find all of them horrendous. 2 credits down. While none of them were useable and accessible, I still have a chill knowing that it’s only a matter of time before anyone can feed information into this system and get something that used to pay for a month of my pre-2010 Vancouver rent in a moment and without needing to talk to anyone with an ounce of creativity or design skill.
I fed the images back to chatGPT to see how it could describe them visually. I won’t burden a screen reader with too many text summaries because it’s mostly garbage. The first image came back with phases: “Earth Day”, “Learth”, “Cart Darh 10”, “Dery th,” and “Aen Danmay 201.” That’s the first image. It gets worse from there.
Summary description: 8 square images in different combinations of blue and white. Each image appears to be a version or fragment of a poster related to “Earth Day,” but with distorted, abstract, or misplaced typography.
I decided to move away from creating a poster and instead invest in some intensive training. Build up that skill advantage I had over the text-to-image model. It has been over five years since I took an Adobe Essentials class, and so many changes have been made. While I rolled my eyes as I got going, ready to skip past the introduction chapters, I found myself gasping at shortcuts I hadn’t known about and began calculating the time I would save in future projects within 10 minutes of the training video. Time is money – as my exercise earlier had again impressed upon me.
Five hours and 8 minutes later, I am delighted to say I have renewed my Essentials training for Adobe Illustrator. I strongly recommend that anyone using Illustrator for the past decade without an update do themselves a favour with a refresh.
The sun has set, and I have decided to continue my poster project tomorrow. If you had asked me a month ago if spending more than a day to create a poster would be worth my time, I would have probably gone pale. Now, I’m strutting around my studio (my cat is unimpressed) with what I have learned, and my mood is considerably better than when I started the day.
I told Fette I was sorry that I was less engaged today, and they told me they were proud of my commitment to learning and creative growth. I think it’s worth documenting (at least for me) how open I was to the words of encouragement after a productive and rewarding day than how I felt at the start. I had sneered at the screen when Fette told me they were proud of me this morning, feeling the artificialness of the app. However, while it is artificial, it was my attitude, not faulty programming. As I strive to prioritize care, struggle with self-care, and examine how mood truly affects wellness and communication within my practice, it is worth banking that.