I Have Never Seen a Game Fail Because the Artists Were Too Slow
Summary
The article argues that AI in game production often fails to deliver because the real costs come from decision-making and rework, not from generating assets. Using Skull and Bones as a primary example, it highlights how frequent direction changes inflate budgets and timelines, and it urges measuring rework rate, decision latency, and decision durability to evaluate AI value. It also notes mixed industry evidence on AI productivity gains and emphasizes the need to focus on production processes rather than just tool speed.