A Year of Growth, Evolving Ai Text to Image Project Brief
The growth of art generation using Ai has grown drastically in just one year. This project has been documented by artist Tara Eveland since October 2022. Full series at Free The Jane via Substack.
The Beginning
I opened an email from Canva in October 2022, it asked me to Beta test their new “text to image” art generation feature. Up until this point I had rolled my eyes at the fuss from the media about this “Artificial Intelligence” that could create anything in the world with a few words, seconds, and direction. I decided to go ahead and help Canva out, check into this text to image concept, and see what it was all about.
First Impressions of Text to Image
Honestly? I was not impressed at first try. But… it was a beta program, and I have been with Canva since it started in like 2015 or so… and I have seen how much it has grown and believe in their company, SO… results aside from the beta model generation, I recognized the POTENTIAL. Because of this, I started documenting the prompts I was using, the exact words I put into the program and what it produced. This project shares those results. Over a year of “Growth of Ai” by artist, Tara Eveland.
I mean… haha, as you can see.. it was… lacking. Okay, I was being kind, text to image prompting was downright SCARY when I beta tested it. In such a very short amount of time though, like literally 2 months, it was creating magic. I grew, artificial intelligence text to image generation grew, and in a way, well, basically we grew together. The prompts got better and the technology kept updating over the year.
Same Prompt, Different Results
In the year I’ve been using Ai to generate images, the issue of art theft has been solved. When the same prompt used in that last image above, “Photograph of a women in white gown standing in the forest in the style of Brooke Shaden.”, is entered into Ai now (Nov. 2023) the outcome says in red, “Content Restriction: The term Brooke Shaden may be restricted find out more”. However, I simply can open another window, go to ChatGPT and do this… I must say, the results shocked me… how smart is this thing?!
Me to ChatGPT:
“hi! how can i describe "style of Brooke Shaden." to ai text to image generation WITHOUT using her name and trigger in the content thing?
ChatGPT
“Hey Tara! To describe the style of Brooke Shaden without using her name, you might want to focus on the elements that define her unique photographic aesthetic. Here’s how you could describe it for AI text-to-image generation:
Envision a world where dreamscapes and reality blur, characterized by a moody and ethereal atmosphere.
Think of compositions that are rich in storytelling, often with a solitary figure centered amidst a vast, often surreal landscape.
Picture a palette that leans towards muted tones with occasional vibrant highlights, creating a sense of vintage timelessness.
Imagine scenes that seem to be captured from a whimsical tale, often with a dark, fairy-tale-like undertone.
Visualize textures that add depth and a painterly quality to the images, as if each photograph is a canvas.
Consider the use of natural elements like water, fire, earth, and air, intertwined with the human form to convey complex emotions and narratives.
Reflect on the imagery that evokes a sense of introspection, solitude, and often, a haunting beauty.
Using this kind of descriptive language allows you to communicate the essence of Shaden's style without naming her directly, guiding the AI to generate images that resonate with her signature approach to photography.”
Results:
Well… they are nothing short of amazing, yet somehow Canva managed to updo this moment by asking me if I wanted to also see it as a VIDEO! They are Beta Testing Ai text to video generation, in partnership with another company called “Runway”. There isn’t a ton of movement, but WOW! I’d say the video nailed style of art and direction perfectly!
I did not keep a detailed ‘month’ record of this project, however, here are a few images that have evolved over these past 12 months and an update on the newest tech upgrade that just launched this evening via Chat GPT. Subscribe to Free the Jane for more creative project briefs and ai art updates as it continues to evolve.