I think showing the (very small) source code as they have with the first example is key to allowing new people to appreciate these. Because we are exposed to so many amazing computer graphics art these days, but most of those take many man hours and are assembled with massive amounts of complexity and computation driving them. Essentially they are brute force art, people imposing their will on the visual output.

What's impressive with generative art like this is how minimal, effortless and elegant the process that created it is. And what i find interesting about it's natural origins is how the process of creating it is far more biased to exploration than forcing output from your own mind.

Shameless plug... although CA are one of my favorite things, I find attractors better for generative art, you get a lot for very little code (although I understand them far less than CA):

  Abstract:
  https://www.dwitter.net/d/15329
  https://www.dwitter.net/d/15367
  https://www.dwitter.net/d/17223
  https://www.dwitter.net/d/17272

  Planet by mapping the same technique:
  https://www.dwitter.net/d/17271

These are all < 194 chars. To see the code for the "eval(escape(unescape..." ones just replace eval with throw.

If I was to ever print something like this I would put the source code directly in the picture itself, seeing that is part of the art for me.