Maintaining Mental Altitude

Reflecting on past experiences where I felt AI was genuinely helpful and enjoyable to work with, I noticed a common pattern: Many tasks involve multiple levels of mental altitude. High-level thinking and low-level implementation. Working solo, you have to constantly climb to high altitude, then drop back down, climb back up, etc. This can be mentally exhausting because you're context switching. Not between totally different tasks, but between different ways of approaching the task.

That's where AI can help, by making the switch less of a pain. You can stay high-level thinking about what needs to be done, then have the AI do the implementation piece. Much in the way that driving a rally car at racing speeds has a driver and a co-driver, allowing the driver to apply total focus on the immediate road ahead, pairing with an AI that way allows you to maintain a stronger focus on what you really want to accomplish without getting pulled out of that focus by the minutiae.

For me, a good example is re-organizing things in a codebase. I have a high-level vision for how things should fit together, and can execute on that vision without having to stoop to the line-by-line rearranging of functions, making sure that imports point to the new destination, and all those details.

I can see similar gains in other domains. Using AI-first design tools (such as pencil.dev ) a designer can stay totally honed in on their vision without interrupting their flow by having to click through menus and drag around objects on the canvas.

Given that context-switching is bad for your productivity (and leads to a fried brain at the end of the day), thinking about AI use as a means to reduce it should be a top priority.