Art Prototyping with AI code assistants

21 Sep, 2025

There are two types of people that buy 3D printers: the people that like to tinker with 3D printers and have nothing but hundreds of benchies printed out in every imaginable filament, and the people that just want the 3D printer to make stuff. The second group isn’t interested in calibration tweaks or firmware updates—they just want results; they just want to use the printer as a tool to help them create their ideas.

I think this is a good analogy for AI-assisted coding. Some people love programming for its own sake—constantly learning new languages, building Fizz Buzz in every syntax imaginable. Others just want to implement their ideas, and coding is simply the means to that end.

With AI coding tools, you can spend more time prototyping your ideas and less time wrestling with syntax, libraries, or frameworks. You can no longer procrastinate by learning “just one more” language or library before starting. Now, there’s nothing standing between your concept and your first working prototype.

Critics often point out that AI-generated code doesn’t scale well to enterprise-grade systems. That’s fine—my goal isn’t to build massive production architectures. I’m here to experiment, to explore, and with AI, I can try more ideas than ever before. In that sense, I can become a 10x artist.

Bellow, I created an SVG of isometric cubes in affinity designer, and then worked with a coding assistant to create a css animation over the stroke width:

Animated Cubes

one note, I know css and html, so once the llm spits out code, I am able to go back and modify it to suit my needs.