Excellently put. I was really aiming to understand the internal conflict I was feeling, knowing that my involvement in the creation was minimal, and yet being able to say, “Look what I did!” Thanks for your comment, Johan.

Expand full comment

That illusion of accomplishment is a core psychological trap of gen-AI, and "text-to-image" is a dangerously misleading term in that reinforces that perverse allure.

They position it as if you, out of your creative genius, expressed through your masterfully crafted prompt, birthed an original image out of mere noise.

When all you did was to find a pretty stick in latent space. One random point inbetween a statistical average of all other pictures; one that a billion other people could have also found, and would be identical from the same prompt if not for the tiny bit of randomness added by the service.

That image was built on somebody else's creative choices, sense of aesthetics, craft and effort. You can give some direction, but can't add sufficient control to claim ownership.

Prompting is searching. Which makes claiming creative ownership and copyright over the output is ridiculous.

But the allure is so, so strong.

Expand full comment