I don't normally do videos -- too frustrating in how long it takes to transfer information versus text -- but if I'm going to be namechecked then I guess I have to make an exception.
Dr. Johnson's comments certainly do echo a lot of my own thoughts about how to generate content, including the importance of forms flowing plausibly from innate motivations -- there's a reason why things look and behave the way they do. (For giggles, I'll note that I actually created the thread here for Ultima Ratio Regum
, the game Dr. Johnson's been developing.)
It's maybe also worth noting that this video was posted by "Games By Angelina
," the project of games AI researcher Michael Cook.
Finally, "speech impediment?" As far as I can tell, it's just the special non-rhotic pronunciation that some English (especially around Kent, I believe) learn to use. Jonathan Ross is probably the most well-known person using this kind of pronunciation, but I've actually used it myself for voicing the speech of a character in my audiobook reading of the Horatio Hornblower novels.
At any rate, I'd love Joshadamlindsey to listen to this video to get some ideas for how visual/functional game content can be procedurally generated based on core variables. Nice find!