Precision vs. Illusion: Rethinking Creativity in the Age of Generative AI

One of the biggest challenges with generative AI is that it creates the illusion of governance. When you prompt, you think you know where the points and pixels fall—you don’t. Within this illusion, one cannot build buildings, cars, or shoes that truly work. One simply cannot paint.

?? I often wonder how the next generation will develop an internal sense of precision—not merely outsourcing it to machines, but understanding things even when ultimate precision is delegated to them. Sure, procedural thinking is great, but how much of the procedures will we remain aware of as we move forward?

This question led me to the intersection of design, computation, and the arts—where I’ve been playing with ideas for over two decades.

Five years ago, I took an initiative to merge my oil paintings with machine learning. The results were strikingly washed out, their essence lost in translation to the multi-dimensional vector space of GAN-generated imagery. More on that -here-.

So, I sat down and developed an algorithm to bring that precision back—to know exactly how every point and every color was generated. Because governance is not just control; it’s knowing where all your points and pixels are, and making them appear—or disappear—at will.

? (Magic) ?

?? Watch the fifty-year version of Cloud Tectonics in 4K on my YouTube channel. More is on the way.