The author argues that the focus on developing frontier technology is not just technical but also human, and that the discipline required to pursue it has been underestimated. He argues that developing a system that works in isolation is not enough and must be deployed, integrated, and sustained in real-world environments. The author also notes that the challenge is to maintain the discipline to develop something properly over time, rather than creating something new. This discipline requires an obsessive level of precision, architectural decisions that prioritize long-term adaptability over short-term convenience, and resistance to introducing shortcuts to demonstrate progress. Despite this uncertainty, the author acknowledges that progress in frontier innovation tends to occur even in uncertain conditions.