My mind has always been chock full of images just waiting to be expressed. I’m able to convey many of these thoughts in English but my ability to illustrate what I see in my imagination has been woefully inadequate. I’ve recruited artist friends and my husband to put my visions into pictures and I’ve been grateful for their assistance over the years. That said, being unable to satisfactorily make my own pictures had left me a tad unfulfilled.  Enter generative AI.

I started Chemory Lane when I was first diagnosed with stage IV incurable cancer. Through sleepless nights and while struggling with the side effects of intense chemotherapy, I was desperate to get my stories out for the world before I was no longer able to do so.

Anxiety over disruptive technology is natural and I’m not here to debate the overall merits of generative AI. I am here simply to say that using these tools has been therapeutic for me. They have given me something to pass the time while trapped in a hospital bed and have helped me feel accomplished while lying mostly still.

Several friends and colleagues have inquired about the images used on Chemory Lane. I was recently emailing with Thomas Matta about my process and thought I might share the process in my journal.

I created the images on Chemory Lane using a technique called prompt engineering. Using prompts helps ensure the style across different journal entries remains consistent. It sometimes takes multiple attempts for me to get the “right” image. I also use Photoshop to make the image more to my style or correct an inaccuracy.

Chemory Lane has two primary sections. One section contains stories about the chemicals I’ve encountered over my lifetime and I create these images using a quasi-photographic style. The rest of the stories are about my cancer journey and I prompt the AI to use illustrations for those. All images, of course, are based on styles popular in the mid-twentieth century because those eras are my favorites.