Automation

Automation is not about doing less. It’s about doing the right things with the time that would otherwise go to repetitive execution.

The case for automation is obvious in theory and surprisingly hard to act on in practice. The friction of setting up a workflow always feels larger than the friction of just doing the thing manually — until you’ve done the thing manually forty times and the cost compounds. Most automation that doesn’t happen isn’t blocked by technical complexity. It’s blocked by the difficulty of recognising recurring work as a pattern rather than a series of one-off tasks.

AI has changed the automation calculus in one specific way: the category of things that can be automated has expanded significantly. Workflows that previously required custom scripting — or human judgment at every step — can now be handled by an agent that navigates tools visually, adapts to context, and stops when something requires a decision. The bar to automate is lower. The responsibility to automate the right things, and stay in the loop on the ones that matter, is higher.

Posts here cover automation at the practical level: workflows, tools, prompts, and the judgment calls about what’s worth handing off.