Field Notes — AI at the desk, not on the slide

The adoption news tracks what large names deploy. This is the other half of the story: the small, concrete ways AI already helps on an ordinary desk — the wins that don't make a conference slide because they aren't big enough to announce, only useful enough to keep.

Each note is one practical move, written the way practitioners talk about it: the boring task, the small AI assist, what it saves, and the gotcha that keeps you honest. The theme underneath them is the same — AI at work is graduated improvement, not a magic agent that does your job. It compounds through people who learn to use the tools they already have a little better.

What these are, and aren't. Field notes are practitioner experience and general industry practice — the standing reference pages in the use-case library grade their evidence source by source; these don't pretend to. They're anonymised on purpose: the people who do this work are in regulated firms and can't disclose much, so no employer, client or deal appears here. The practice is real; the identifiers aren't in it.

The notes

• You don't need a new tool — you need AI to teach you the one you already have

  Point AI at the vendor's own documentation to learn your task the lean, reusable way — then flip it and ask what useful features you're not using. From Power Query to the feature already sitting in the app.

For what firms have put into production, deployment by deployment, see AI adoption in the fund industry; for the job-by-job assessment with graded evidence, the use-case library.