Even with all the advantages they had (including, arguably, being single), they still had their own deathmarch (100 hour weeks) at the end. I wonder if it's at all possible to make video game development sane when you have a publisher?

Maybe when we create an AI that can test games for us.

Is testing the biggest time sink in the process? I’ve always wondered, despite having obviously complex architecture, why game development cycles had such crunch times. I knew it was based on publisher deadlines and such, but figured the bottleneck would be more in development time rather than testing. If the latter is the case, given the progress of OpenAI, I could see that being a plausible solution.

Agreed, just look at the amount of innovation in indie games compared to AAA titles.

I agree. I was surprised by not only how few but how young. To me it seemed the work of a large group of industry veterans.