There’s no shortage of advice on how to use AI. New tools, new features, new prompts. Every week, there’s something else we’re told we should be trying.
But using AI more often doesn’t necessarily mean we’re getting more out of it.
From our experience helping teams adopt AI, we’ve seen that the real value comes from being more thoughtful about where and how it’s used. Three simple habits can make a difference.
1. Start by finding your Time Thieves
One of the easiest ways to lose interest in AI is to use it without a clear purpose.
You open an AI tool, ask it to generate some ideas, play around with a few prompts, and walk away with nothing particularly useful.
A better approach is to start by looking at how you spend your week.
We all have what we call Time Thieves. These are the recurring tasks that quietly eat away at our day. They’re not necessarily difficult, but they take longer than they should.
Maybe it’s pulling information from multiple reports, summarizing meeting notes, preparing presentations, or spending an hour drafting something that should take 15 minutes.
In our AI training, we ask people to identify one of these Time Thieves and see whether AI can help them tackle it.
The exercise is simple. Pick a recurring task, understand how much time it takes today, and see what changes when you bring AI into the process.
Not every task will benefit, and that’s fine. The point is to find something useful.
And when you find a task where AI saves you time or improves your work, you’re much more likely to keep using it.
Start with one Time Thief. Solve it. Then find the next one.
Explore our Time Thief framework to see how we help teams put this into practice.
2. Know when to use AI and when to step away
Once you start seeing where AI can help, it’s tempting to use it for everything.
But not every task needs AI, and not every decision should involve it.
There are some important distinctions worth making.
Sensitive information: Before sharing company documents, financial data, or personal information, make sure you’re using a tool approved for that type of data. Not every AI tool has the same privacy and security protections.
Decisions that affect people: AI can help you prepare for a difficult conversation or organize your thoughts around a performance review. But it shouldn’t be making decisions about people’s careers, health, or relationships on your behalf.
Your own voice and experience: AI can help you find the right words, but it shouldn’t replace what you think or feel. Sometimes the best thing you can do is write something yourself, even if it’s not perfectly polished.
Facts that matter: AI can be a useful starting point for research, but when accuracy matters, especially with legal, medical, regulatory, or technical information, verify the details against reliable sources.
In these situations, know where AI can help and where you need to stay firmly in control.
3. Have a conversation with AI
Even when you’ve found the right task, how you work with AI makes a big difference.
But some of the best results come from the conversation that follows.
Start by giving AI enough context to understand what you’re trying to accomplish. One approach we teach is the RISE prompting framework, which provides a simple structure for communicating what you need.
You don’t have to memorize complicated prompts. The idea is to be clear about the task, provide relevant context, explain what a good result looks like, and then work through the response.
Tell AI what’s missing. Push back when something doesn’t make sense. Ask it to explain its reasoning or explore a different approach.
Think of it more like working with a colleague. You wouldn’t expect someone to understand everything from a single sentence, and you wouldn’t accept every suggestion without thinking it through.
The first prompt starts the conversation. It’s what you do next that often makes the difference.
The difference is how you use it
AI is going to keep getting better. The tools will become faster, more capable, and easier to use.
But that doesn’t mean every task needs AI or that every answer it produces deserves your trust.
Start with a real problem. Work through it. Question the results. Know when to use your own judgment.
