The agent kept finishing things, and I kept giving it more to do. Eventually, pressing Enter had become easier than deciding what should happen next. I had built a faster way to work and lost sight of where my own judgement belonged.

This talk follows the search for a way back. Slowing everything down gave me space to think, but left another problem unsolved: how to work effectively when expectations had already changed. I looked for other people facing the same tension and built a way of working around decisions I could actually make and results I could verify.

The journey moves from juggling tasks to managing a flow of work, then beyond my own desk. An efficient developer still needs an organisation that can decide, coordinate and keep work moving. We finish with the question behind the whole story: when an agent can execute the task, what does someone trust you to decide?

For developers, engineering leaders and organisations adopting AI. A personal journey connects the satisfaction of rapid output with the harder work of choosing, verifying and retaining professional judgement. The audience can recognise the pressure to do more while questioning where that pressure belongs. The organisational perspective makes this a useful bridge between technical AI sessions and discussions about how work changes.