At the beginning of the COVID-19 pandemic, I started feeding the birds on the deck of my condo. In a couple of months there were quite a few coming and going. I found it quite enjoyable to have them around.

I had been doing digital drawings of people, and the birds became integrated into the works.

I found the symbolism of a bird being comfortable on a person's head intriguing. It could easily fly away but has chosen to stay. It was a delicate and fragile relationship - somewhat puzzling and humorous at the same time. The relationship seemed to challenge our society's contemporary view of nature. There was a partnership between the bird and the person. I liked it.

Moving ahead almost exactly two years, in the autumn of 2022, I was experimenting with artificial intelligence (AI) to produce images. I found what DALL-E could do interesting and somewhat amusing. A person gets 15 free credits a month with each credit producing four images. I tried several different text prompts and then moved on.

The next month, I produced a series of images using the text prompt “a woman with a cat on her head”.

Then it came to me.

An image of a woman with a bird on her head done in a photographic style. Yes! This would be very intriguing. So I went for it.

I found that what DALL-E had trouble doing was just as interesting as what it did very well. I quite like the mistakes. They add to the complexity of meaning in the piece. I kept the text prompt simple and watched how DALL-E would embellish it, adding details that weren't asked for.

So… the text prompt used was:

“Portrait of a woman in her twenties with a small bird on her head in a photographic style.”

Which was then refined to:

“Portrait of a woman in her twenties with a small bird on her head in a minimalist photographic style in color.”

None of the women in the images actually exist and none of the images will ever be duplicated. That's the nature of AI.

I chose a woman because of the connection to Mother Earth.

I chose a young woman because the majority of her life is still ahead of her.

I also found that specific races needed to be added to the text prompt to get a broader representation of humanity.

In the spring of 2023, my friend Sergio Soares helped me set up a Discord server so that I could use Midjourney.

I was astounded at what Midjourney could do.

I decided to go around the world creating a representation of every type of woman. So… by using Wikipedia and Midjourney, I spent three months downloading images produced by Midjourney, ending up with 3,600 images.

I started a series of recordings/albums using the images for the album covers.

I didn't like how slick the images were, so I came up with this prompt.

“cropped aged color Polaroid photograph of a young contemporary [insert race here] woman in her mid 20s, with a slight grin on her face and a single small bird on her head, outside in British Columbia”

I was very happy with these. I produced a little over 6,000 of them, all different.

It's 2025 now, and to be honest, I can't think of a good reason to generate AI images anymore.

So, what do I do with these? I print postcards and tiny 4x4 cm cards. I leave stacks of postcards on coffeehouse shelves, free for the taking. The tiny cards go wherever it feels fanciful: pinned to bulletin boards, tucked into product boxes in stores, left under a restaurant tip, or handed to a random stranger. You get the picture. It's fun! :)