VP of Tech reports working more with AI agents despite lower role demands

The author attributes this surge to the removal of friction in implementation. While their role shifted from full-stack engineering to leadership, reducing expected coding tasks, AI agents made it easy to take on additional projects, fix minor bugs, and pursue new ideas. This created a cycle of "bursts" where the author starts multiple parallel agents, leading to high mental load from context switching and constant review needs.

The post argues that while AI increases capacity to execute, it does not increase the capacity to decide what to execute. The result is a higher baseline of work and periodic burnout, as the author finds themselves constantly taking on more tasks than in previous years, even during slower periods. The author notes that AI helped draft the article and analyze the data, but the opinions are their own.