You do not need to understand how it works to use it well
You do not know how your car engine works. You probably do not know how your accounting software calculates depreciation. You do not need to. You need to know what the tool does, when to reach for it, and what its limits are. AI is exactly the same.
The mental model that helps most people is this: think of a conversational AI tool like Claude or ChatGPT as a capable, tireless assistant who has read an enormous amount and is available at any hour, will never be too busy to help, and will never judge you for asking a question you feel you should already know the answer to. That assistant is not perfect: it makes mistakes, it can be confidently wrong, and it needs good direction to do its best work. But for a long list of things that take up your time and mental energy right now, it is genuinely useful.
The rest of this post is about the specific parts of leadership and management work where that usefulness shows up most clearly. Every one of these requires nothing more than typing in plain English.
Thinking through decisions before you make them
Leadership is largely a series of decisions, many of them made with incomplete information and real consequences. One of the most underrated uses of AI is as a thinking partner when you are working through something difficult and do not yet know what you think.
Describe the situation. Lay out what you know, what you do not know, and what you are weighing. Then ask for the strongest case for each option, the risks you might be underestimating, and the questions you should be asking before you decide. AI will not make the decision for you: it does not know your people, your culture, or the history behind the situation. But it will surface angles you had not considered and push back on reasoning that has a weak spot.
This is particularly useful for decisions you are too close to. When you have been thinking about something for days and your thinking has started going in circles, an outside perspective, even an artificial one, can break the loop. Ask it to argue the opposite of what you are currently leaning toward. Ask it what a skeptic would say. Ask it what you are probably not accounting for. The answers are not always right, but they are often useful.
Preparing for hard conversations
This one came up in post two of this series as one of the things that surprised me most about using AI in my own work. It deserves more space here because it applies directly to anyone who manages people.
Hard conversations: performance issues, delivering news no one wants to hear, pushing back on a decision from above, addressing a conflict between team members, are where leadership is most demanding and where preparation makes the biggest difference. Most people walk into these conversations less prepared than they should be because preparing feels uncomfortable, so they procrastinate until it is time to have the conversation and then improvise.
AI makes the preparation step easier. Describe the situation, the person, and what you need to accomplish. Ask it to help you anticipate likely reactions and how you might respond to them. Ask it to help you find the clearest way to say the hard thing without softening it into meaninglessness. Ask it to point out where your planned approach might land poorly.
You are not looking for a script. Scripted conversations feel scripted, and people notice. You are looking for the preparation that lets you stay composed when the conversation goes somewhere unexpected, because you have already thought through more of the territory.
Writing things that need to land a certain way
Leaders write constantly: emails to staff, messages to clients, updates to boards and investors, performance reviews, job postings, responses to complaints. Much of this writing has to carry weight: it needs to be clear, it needs to strike the right tone, and it needs to say exactly what you mean without accidentally saying something else.
AI is excellent at this kind of writing when it is given proper context. The key is specificity: telling it not just what you want to write but who you are writing to, what you need the message to accomplish, and what you specifically want to avoid. The February post on prompting covers this in more detail, but the short version is: more context produces better output, and the extra thirty seconds to provide that context is always worth it.
A few types of leadership writing where AI consistently earns its keep:
- All-staff communications during uncertainty: a reorganization, a difficult quarter, a policy change. Getting the tone right between honest and reassuring is hard. AI helps you find it.
- Performance reviews that need to be constructive, specific, and fair, especially for employees where the situation is complicated.
- Rejection messages to candidates, vendors, or proposals where you want to say no clearly and leave the relationship intact.
- Board or investor updates where the news is mixed and the framing matters more than most people admit.
Getting up to speed on things you need to understand
Running a business or leading a team means constantly encountering territory you are not expert in. A new regulation that affects your industry. A contract term you have not seen before. A technology your team wants to adopt. A financial concept your CFO mentioned that you nodded at but did not fully follow.
AI is a patient, non-judgmental explainer. You can ask it to break down anything in plain English, ask follow-up questions without embarrassment, and ask it to use a different analogy if the first one did not click. You can ask it to explain something assuming you already know the basics, or assuming you know nothing at all. It will not make you feel slow for not knowing something. It will just explain it as many times and in as many ways as you need.
One caveat that belongs here: AI is very good at explaining things but not always perfectly accurate on specifics, particularly on legal, financial, or regulatory questions where the details matter. Use it to get oriented and ask better questions. Use your lawyer, accountant, or relevant expert for anything where being wrong carries real consequences.
Turning a long pile of information into something actionable
Leaders are often the person in the room who needs to have read the report, absorbed the meeting notes, reviewed the proposals, and formed a view, all before the next conversation. The reading alone can consume hours that do not exist.
AI compresses this dramatically. Paste in a long document, a meeting transcript, a chain of emails, or a set of competing proposals and ask it to pull out what matters: the key decisions, the open questions, the disagreements between stakeholders, the recommended path forward, and the risks worth flagging. You get the digest in two minutes instead of spending forty-five with the source material.
This is not a substitute for deep reading when deep reading is required. A contract you are about to sign, a report with significant financial implications, something where the nuance in the original document is part of the point, those get your full attention. But the large volume of material that comes across a leader’s desk just to stay informed? AI handles the first pass well, and you direct your attention to what actually needs it.
Thinking through hiring, team structure, and people decisions
People decisions are among the most consequential a leader makes and among the hardest to think through clearly, because they are almost always emotionally loaded. AI is useful here precisely because it is not emotionally loaded. It has no stake in the outcome, no history with the person in question, and no discomfort with the hard version of the question.
Ask it to help you think through whether a role needs to be refilled or redesigned. Ask it to help you structure an interview process for a position you have not hired for before. Ask it to generate interview questions that surface the specific qualities you most need. Ask it to help you think through whether a performance issue is a skills gap, a motivation gap, or a fit gap, because those three have very different solutions.
AI will not tell you whether to promote someone or let someone go. Those decisions require judgment that belongs to you. But the quality of your thinking going into those decisions can improve significantly when you have spent time working through the angles with something that asks the uncomfortable questions without flinching.
The one adjustment that makes all of this work better
Give it more context than you think it needs. This is the lesson from the prompting post, and it applies especially here because leadership situations are inherently contextual. The history of a relationship, the dynamics of a team, the unspoken thing everyone is working around: AI does not know any of this unless you tell it. The more of that context you bring into the conversation, the more relevant and useful the output becomes.
You are not oversharing. You are briefing. Think of it the way you would brief a trusted advisor before asking for their perspective on something sensitive. The quality of the advice is proportional to the quality of the briefing. That is true whether the advisor is a person or an AI.
The leaders getting the most out of AI right now are not the ones who use it for everything. They are the ones who identified three or four places in their actual week where it genuinely helps, built the habit of reaching for it there, and let everything else follow from that. Start narrow. Go deep. The breadth takes care of itself.
A good place to start this week
Think of one decision you are currently sitting on, one hard conversation you have been delaying, or one piece of writing you have been putting off because you cannot quite find the right words. Open Claude or ChatGPT, give it the full context, and ask for help with that one thing. Read what comes back. Use what is useful, ignore what is not, and notice whether that felt different from doing it alone.
For most people who try this with a real situation rather than a test case, the answer is yes. And that first real use is usually the one that changes the habit.
What is coming in May
Next month: how to automate repetitive business tasks with AI in an afternoon. We will get practical and specific: real workflows, real tools, real time savings, for businesses that are ready to move beyond just using AI as a writing assistant and start using it to actually remove work from their plate.
This is post eight of a two-year series on AI for real people doing real work. Post one covers what AI actually is. Post two is how I use these tools day to day. Post three covers the five free tools worth trying first. Post four is about taming email with AI. Post five covers the AI and security landscape. Post six is prompt engineering without the jargon. Post seven covers building a phishing awareness program on a budget. Something on your plate you want to think through? Send a note.
