On AI-generated images: a reflection with myself

14 July 2026

Boy, that article about AI-generated images I wrote yesterday, the several blog posts, comments and opinions I have read here and there about that controversial topic… it sticks in my mind.

Since yesterday, it's like I'm having a passionate debate about it with myself. I'm weighing the pros and cons, and upon reflection, I'm now considering using my own cover images instead of the ones I used to generate with AI. I mean, this is pretty paradoxical to what I wrote and defended in my previous post, but it just made me think. Which is a good thing, isn't it?

There was a time, waaaaay before AI, when I enjoyed tinkering with Photoshop so much, playing with images (photos I shot myself or images grabbed on the web). When I was in my early twenties, I spent countless hours on it. It was my main hobby. That's basically how I learnt to use Photoshop, starting with version 6 back then.

2005 desk — The computer on which I spent an innumerable amount of time on Photoshop, just for fun. This was 2005, so I guess it was during the earliest versions of Photoshop CS. Can you notice the Sharp GX25?

Since that time, I have never ceased to use Photoshop (which became a heavy-laggy-bloated-and-frustrating-to-use-software, but that’s another story), but I’m not using it with as much excitement and innocence as I used to have. This has to change. Not for the sake of not using AI-generated images anymore, but just to spend more time enjoying and trying to create my own images.

I like the process of writing, of course, but I may have made it an excuse not to spend time anymore on that other thing I also enjoy, deep inside of me, which is tinkering with pictures on a computer. I should, and I will, allow myself more time to make my own cover images. I wonder how it will turn out, though.

I might still use the help of AI, but I should not delegate all the creation to it. In fact, I should do it the opposite way from how I’m doing it now, which is: writing a prompt, submitting it to the AI, going back and forth if necessary, and refining the final AI-generated image in Photoshop. I should try to: create the image by myself at first in Photoshop, and if necessary, use the AI to refine it (remove something, for example).

So. Yeah. Between my post from yesterday and this one, one might think I’m kind of a contradictory person. I am, indeed. I often call many things into question, sometimes adjusting or changing my tune after reading and hearing others’ points of view.

Starting today, I will accord myself time to create my own cover images for my posts, starting with this very one¹. I will surely continue to use AI as a tool (to generate images, code, or other things I will write about on a future /ai page), but I also want to revive my old self who was having fun on the computer.

I would like to thank the many people who wrote about that topic on their blogs. That’s the web I love.

The cover image of this article is an old film photo of myself from the late 1990s. I scanned and cropped it a lot (which explains why it’s so blurry and pixelated), to focus on me and the computer screen, on which I added a screenshot of Photoshop. It was fun to do, and guess what: no AI was involved in the process! ↩