People think the trick is going to be the first to find where AI excells and to push that HARD. That's not where people will find success. People who succesfully use AI will find the thing THEY DO WELL that the AI can't do, and they will find clever ways that the AI does things that can work in tandem with each other. For example, you're not seeing video clips that enter the uncanny valley. The people who make quality AI videos know the limitations, they know not to enter the Uncanny Valley. So they make absurd animal clips with funny themes. The same will be true of gamedev. AI 3d art and animatoins aren't up to quality yet, but incrimental games are TOTALLY possible using prompts. But a succesful incrimental game needs a designer that knows how to make incrimntal games FUN. Everyone thinks AI is going to make the next Call of Duty absolutely next level. That's not where we're going to see AI absolutely shine, it'll be genres of game that can be fun in unique unexplored ways that the AI excells in, and AI is good at procedural things and math. Games in space, an earth sim with procedural terrain and incerdible new physics systems. AAA will be able to do more hybrid type stuff. Fantastical new physics sims with amazing character art and all that jazz. More important than knowing the emerging horizons, is understanding the limits. THAT is what will dictate the paths to success.