You can use AI at work to practice a skill, review an idea, and improve a piece of work you already started. When you're in your first job or early in your career, it's worth choosing uses that also help you understand the work better.
That might mean asking for feedback on a short update, practicing how to ask a clear question, or learning why a spreadsheet formula works. You still bring the context, check the facts, and choose what happens next.
The aim is to finish with better work and more skill of your own. A useful AI session leaves you able to explain a choice, spot an error, or take on more of the next task yourself. You can build that into the way you work from the start.
Choose how to use AI at work around one skill
Pick a skill you want to strengthen before you choose an AI task. This gives you a way to judge whether the tool helped you learn.
In a first job, that skill might be writing clear updates, checking a spreadsheet, planning a small project, or explaining a choice to your team. Name something you can practice and see in the work.
Start with your manager's expectations and your workplace's approved tools. Rutgers' guide to AI and your career advises checking both before using AI at work. Ask what information you may share and what needs a person's review.
For practice, use made-up details that don't contain work information. Removing a name may still leave enough detail to identify a customer or project. Keep private files and personal details out of any tool your workplace hasn't approved for that use.
Then choose a small, clear goal, such as making your next progress update easier to follow. You can check that with your manager after you've written it. The tool's praise isn't your only measure.
Make your first attempt so you have something to improve
Write down your own ideas before asking AI for feedback. Even a rough attempt gives you a view of what you know and where you need help.
For a project update, note what's finished, what's next, and what needs a decision. For a recommendation, name your choice and the reason behind it. For a spreadsheet, say what you want the formula to calculate before you ask how to build it.
Then try: “Here is my first attempt at a work update. Tell me where a reader may need more context or a clearer next step. Ask me questions before suggesting changes. Don't rewrite it yet.”
Answer those questions yourself. You may notice that you left out a date or never said who needs to act. Those are useful things to learn for the next update, too.
You don't need to work through every task alone before using a tool. For a new skill, though, making an attempt helps you choose the kind of help you need. When you're planning the work, turning a to-do list into a plan gives you clear steps to bring to the conversation.
Ask for feedback that teaches you why
Ask AI to explain a suggested change so you can decide whether it fits. A polished sentence is more useful when you understand what made it clearer.
Try this prompt: “Review this practice email for clarity. Name the main request and tell me whether it's clear who needs to act. Suggest one change and explain why. Keep the tone friendly and direct.”
Look at the reason before you accept the edit. Moving a request near the start may help a reader know what you need. Adding a deadline may help them plan. Adding a deadline you never agreed to would change the message, so you need to catch that.
You can use the same approach with a presentation or report. Ask which point needs evidence, which term needs explaining, or which chart label may confuse the reader. You make the final choice based on the real audience and task.
Keep one useful note from the feedback and apply it on your next attempt. “Put the request near the start” is something you can practice. “Write better emails” is harder to act on between meetings.
Practice a conversation before you have it
Use AI to rehearse what you want to say, then have the real conversation yourself. Practice can give you a place to try the words out loud.
You might want to ask which task comes first, explain a delay, or ask for more detail on an assignment. Choose an ordinary situation with made-up details, and tell the tool what part you want to practice.
Try: “Help me practice asking my manager which of two tasks should come first. Use a made-up work situation. Respond as the manager, one reply at a time. After I answer, tell me whether my question was clear and what information I left out.”
Say your response aloud before you type it. You may find a sentence that looked fine on screen has far too many words once you try to say it. Work language can do that.
Treat the role-play as a rehearsal. AI doesn't know your manager's mood, your team's history, or what happened before the conversation. Its reply can't predict the response you'll get.
When you talk to the person, listen and adjust. Asking a clear question is part of the skill. So is making room for an answer you didn't expect.
Use another viewpoint to check your ideas
Ask AI to question an idea you've formed, then judge its feedback against real evidence. This gives you another set of questions to consider before you make a choice.
Try: “Here is my recommendation and the reasons behind it. What assumptions am I making? What information would I need to check? Give me another option and explain when it might fit.”
Check whether its concerns apply to the task. A suggestion may sound sensible while ignoring your deadline, budget, or team's way of working. You know details the tool may never have.
A 2025 study of AI and critical thinking at work found that greater trust in AI was linked with less reported critical thinking. It was a survey of workers, so it doesn't prove that AI caused a loss of skill. It does give you a reason to stay involved in checking and choosing.
Write down why you accept or reject a suggestion. You don't need a report about the report. A short note can help you explain your choice when someone asks how you reached it.
Learn the steps behind the work you keep doing
Use AI to explain a process you want to understand, then check it against the way your workplace does that work. Learning the steps helps you handle the next version of the task.
For a spreadsheet skill, ask the tool to teach it with made-up data. Have it explain what each part of the formula does. Test it on a small example where you can check the answer yourself before using it in a real file.
Try: “Teach me how this formula works using a small practice table. Explain each part, then give me a similar task to try. Wait for my answer before showing the steps.”
For a work process, use your team's instructions as the source. Ask a colleague when the written steps don't cover a real situation. AI can explain a general idea, but it can't set your company's policy or tell you what your manager has agreed to.
You can also use a tool for routine steps your workplace allows it to handle. Choose which parts still need your practice and attention. If clear writing is your goal, keep writing and reviewing some of the messages yourself.
Give practice a place in a weekly plan that fits your life. Pick a work task you already do so you can try the skill again in context.
Check the result and bring a person into your learning
Review AI-assisted work against real sources and the needs of the people who will use it. Your final check is part of doing the job.
Confirm names, dates, numbers, links, and promises. Read the original source when a claim matters. NIST's guidance on generative AI risks includes the risk of false information presented with confidence. Asking a chatbot whether it's sure doesn't replace checking the evidence.
Then ask whether the work makes sense for your team. Does the update give your manager the decision they need? Does the plan include the person who has to approve it? Can you explain how you checked the total?
Ask a colleague or manager for feedback on the real skill you're building. You might say, “I'm working on making my updates clearer. Was the next step clear in this one?” They can respond from the context of your job and the work they received.
Over time, notice what you can do with less help. You may ask better questions, catch a missing detail sooner, or write a clearer first draft. Keep a few notes about what changed and try that skill again on the next task.
For your next AI session, choose one work skill and bring your own attempt. Leave with one change you understand well enough to use yourself.
You might also like
Planning when each day looks different
Making room for life after work
