How do I use AI safely and usefully at work?
Nobody trained you, the tool is just sitting there, and the advice online is either breathless or terrified.
This is the practical middle: what to actually do on Monday, what to keep away from it, and how to tell the difference. No technical background assumed — if you write emails and documents for a living, everything here applies to you.
New to this?
About an hour end to end, and you will have done one real thing rather than only read about it. Everything else on this page is browsable in any order — this is the route if you would rather be told one.
Pick a first task where being wrong is cheap and obvious.
Learn the thing that actually changes output quality. It is not prompt wording.
Get one real win today, on work you already have to do.
Stop retyping the same setup. Save it once and reuse it.
Answer the etiquette question you will hit in your first week.
The full set for this pillar, grouped by the question you are trying to answer rather than the date it went up.
Where to point it first, and how to get competent without waiting for your employer to train you.
A simple filter for choosing where to start with AI — one that steers you toward quick wins and away from the tasks that create real risk.
Clever prompts get all the attention, but the thing that actually separates people who get real value from AI is much simpler — and much more teachable.
Employers say they'll reskill workers for AI — 77% of them — but only about 13% of employees have actually received any training. A practical, do-it-yourself path to real AI literacy while you wait.
Calibration. Useful for deciding what to hand over, and for not being the person who oversells it in a meeting.
What can AI agents do? In plain English: carry out multi-step tasks across your apps instead of just answering questions. Here's what that covers today, what it still can't do, and how to start.
Google mapped 14.7 million AI conversations to real occupations. AI showed up in 68% of them — but in the median job, it touched just 21% of the tasks. Here's what that gap actually means for your week.
AI feels fast, but 'feels fast' and 'saves time' aren't the same thing. A back-of-the-envelope way to work out which tasks actually pay off.
The judgment calls: what you are allowed to paste in, when to say you used it, and how to avoid producing confident nonsense.
The report AI helped analyze, the email it drafted, the message it reworded — when does honesty require you to say so, and when is it just using a tool? A practical test.
Roughly half of employees use AI tools their employer never sanctioned — and most of it is harmless. A plain-English guide to the line between 'fine' and 'you just leaked something,' and how to stay on the right side of it.
40% of workers got hit with 'workslop' last month — AI output that looks finished but isn't. Each piece took nearly two hours to clean up. Here's how not to be the person sending it.
The parts that are about you rather than the tool.
AI at work isn't only a tool you use. Increasingly it's a lens your employer points at you — scoring productivity, ranking performance, even helping decide layoffs. A plain-English look at what's real and what to do about it.
Most AI users fear falling behind if they don't keep up, and new grads worry AI will swallow entry-level jobs — yet most active users say they're producing work they couldn't have a year ago. Making sense of the contradiction.