Couldn't help but focus my post on the thread title itself, so my post will revolve around that, if you may forgive that approach. Let us take a step into the future and ponder how games development itself could be.
A new way of making AAA games would mean new technology and approaches. You can't change a running system, a lot is simply dependent on time windows and money and the realistic constraints (development taking time with finite resources and an eventual delivery date).
Imagine in the moderate future, advanced technology aiding you in game development rapidly. Maybe even with AI support. A world where you don't need to develop such major titles over several years and still be unable to finish all original ideas (different apartments, way back years ago, possibly pickable classes and factions or jobs, etc) due to the aforementioned constraints and priorities, because development itself becomes faster and easier.
I see issues in both the "mainstream industry" and also more independent indie developers. Notable titles often get changed scopes and some downgraded package because you eventually have to deliver as the market or investors demand it, or simply because you cannot afford to develop a title longer. Some titles outright fail and never see the light of day.
Then, you get indie developers who don't have the same financial security. Some fail, some make it but their scope remains limited as they lack the bigger resources. And then there is crowd-funded mega projects like Star Citizen. Star Citizen is, depending on how you look at it, both a bad and yet also good example.
You're not subject to certain investors as far as I understand how it works - or not so much like in the big industry. That's where such more independent developers or studios pride themselves - yet, a huge mammoth project takes time. Star Citizen will easily see the next decade in continued development until it is probably somewhat or relatively near where it was pitched to be, without being finished for good.
Long story short, no matter the route or approach and depending on project size, you are always subject to higher powers and limitations. Money, time, luck, scope, people to work with, experience, skill, and so much more. Some of that, you cannot simply wash away or change radically: How "the system" works.
But I feel you can make a lot along the way easier with ground breaking technology aids, like AI. Where games can be made or tweaked faster, thus, more content can be offered. Where a lot of the work doesn't become menial or hands-on tasking and coding, but can shift to a creative process as technology takes the parts that take time and prone to mistakes in coding or fine-tuning and does the job in almost no time. The coder becomes the artist who simply oversees and shapes the process as needed by the creative process or director.
These are the games and virtual (vast) worlds I look forward to and it should still be within my time frame on this earth. Maybe at the end of this decade, maybe in the 30s, 40s.
It's quite exciting to think what could be around the corner.
But let's stop dreaming and look at the near future: Curious to see what they will make with DLC and the next Cyberpunk game.