AI promised less work. I've never worked more

Despite AI tools promising to reduce workload and free up time, the author—a VP of Technology—found himself working more than ever in 2026, with GitHub activity matching his busiest years as a full-stack engineer. The problem isn't that individual tasks are harder, but that running multiple AI agents in parallel creates constant context-switching demands, as each agent periodically requires his attention and approval, making work mentally draining despite being theoretically more efficient. Rather than going home earlier, the productivity gains from AI have simply expanded the scope and pace of work, replacing deep focus on a few tasks with shallow multitasking across many simultaneous projects.