My new image policy (AI and photography)

Image generated by "Grigne".

First: AI.

My policy is this: I use AI to build tools, not content.

So for programming, occasionally.

For example: I used Claude Code to write the scripts that feed this site. (I write my notes in Obsidian and it syncs live.) I'll come back to that.

But now that the site runs, AI is no longer of any use.

Another example: AI helps me code my generative art program (named "Grigne"). I'll come back to that too – it's a hobby I've had for a long time.

But once the program runs, no more need for AI.

So I use it heavily to build the tool – a few weeks – then barely at all the rest of the time.

(Yes, I'm still in internal conflict over that one. You can yell at me in the comments.)

But I never use AI to write text or to produce images.

Not that I haven't tried.

Several times, I wanted to see if Claude could help me write – things that bore me, for example.

And I realized that no, not at all. It doesn't even make sense.

Because today, my life is organized like this:

• Either I write something I enjoy, that I'm the only one who can do and that I have no desire to delegate. The activity itself is the pleasure.
• Or I don't write.

And for images: same thing.

The pleasure is making the drawings, taking the photos.

Bit by bit, I've removed from my life (personal and professional) the need to produce any content I wouldn't want to make myself.

It's never necessary.

It's even the main reason I stopped audiovisual production.

(In all honesty, I made an exception to this philosophy for my web series Panic in Space! At first I made all the backgrounds by hand in Blender. Then I gave in at the end, when Midjourney got good enough to help me.)

I'm not throwing stones at those who work differently, of course.

But personally, if you write me a message or send me a drawing, I prefer it short, personal and imperfect. Otherwise, I promise nothing.

I've also changed my photography philosophy. We'll talk about that tomorrow.