I think part of what he's talking about is just the plain old lack of forward progress. Here we are 11 years after the UO beta, like what... 14 years after Master of Orion... and we are still recreating practically every thing for every game. I'll give you an example because it's been on my mind.
Ever since I started posting in another place about what games have cooperative campaigns and skirmish modes, I've been thinking about why so few do and why the AI is so often bad in them. It's because it takes a lot of time and costs a lot of money to do correct AI in a game. Well, all this time and we haven't done a dern thing to improve that.
What's needed is some sort of open resource library somewhere of unit AI behavior that would define how units should act in a game. Let all the devs be able to use it. Most of these game companies don't have the money or time to hire a dedicated AI designer/programmer and with every game, they've got to research and define and code how every unit would act in every situation. Good luck with that. Units could be regiments or companies or individuals depending on the game, RTS, or FPS or MMO etc. Each unit would have to be identified by by "epoch" or "age" I guess, ie: like Ancient, Medieval, Napoleonic, Modern etc. Then by type of unit, Cavalry, Infantry, Line Infantry, Artillery, Horse Artillery, magician(s), priest, healer, Sloop of War, patrol boat, frigate, bomb ketch, ship of the line, etc. Then the activities, what to do if attacked by X and movement is Y and you get the picture. Can't just jump into it though, you have to have a group of AI savvy experts define the methodology and what categorizations would be generic enough to be useful. This is just meant to think about, I'm not an AI codemonkey - better brains than mine would have to do it.
That's the reason a lot of these games don't have intelligent AI even for single player, and that's the stumbling block for co-op multiplayer against the computer. It would take someone just working on this one aspect of it, for each developer for each game, a loooong time to define these things correctly. And every developer has to do it over and over.
Now that's just an example because I was thinking about that particular topic a couple days.
That is the "kind" of advance to which R. Garriot refers I believe. Doing the kind of things like this which would free up all the game companies to be more creative. Standardization and availability of those things that further the industry freeing the developers up to be more creative and honestly, let the smaller ones make more money so they can make more games.
