Dave,

> AI-powered text-to-image generators such as Dall-E 2 are amazing
> demonstrations of the power of deep neural networks trained on
> massive datasets.

The latest results on image generation are often impressive and
sometimes humorous, but it's worth noting some earlier AI systems
that have generated creative visual art.

One example is AARON, which Harold Cohen developed over decades
by incorporating ideas from the knowledge-based systems movement:

- https://www.wired.com/2001/05/aaron-art-from-the-machine/
- https://ojs.aaai.org/index.php/aimagazine/article/view/2695

More recently, Simon Colton has developed another system that he
calls The Painting Fool:

- https://www.newscientist.com/gallery/painting-fool/
- https://www.youtube.com/watch?v=m2KWQ47LBXQ

Cohen's effort on AARON started in the 1970s and Colton's work dates
back 15 years. Both are worth presenting to students so they realize
that DALL-E and its relatives are not the first of their kind.

Best, -Pat

P.S. Before anyone claims that deep learning lets us develop artistic
       systems with much less effort, remember the new results come
       from large companies with many resources, not just data.