There has been a lot of talks about the advances of artificial intelligence, especially in the field of machine learning in recent years. Instead of bringing any real intelligence to the table most of the tools are mostly big, empty promises that can produce some kind of statistical results at best.
The results have been more entertaining than useful. There has been many recreational outcomes from the tools. A great example is the AI dungeon, a text based story driven game along the good old games long before stunning graphics took the mainstage instead of storytelling. Those AI models turn out to be quite good at coming up with made up stories.
The latest trend in this field is the AI generated imagery. DALL-E, Midjourney and Stable Diffusion can each produce stunning artistic images (some even winning art contests). But when it comes down to producing something more real world like they still hit the uncanny valley. Even the best images give out the feeling that there is something wrong.
On the recreational side again, these images have found creative uses for example in game development. Producing graphics for a game can be quite a job, but with these new tools, and a bit of downscaling the pixelated retro look games have become easier than ever to produce.