Michael Byers
As artificial intelligence becomes an ever-more-capable learning companion, we are outsourcing to it more and more tasks. We ask AI to summarize meetings. To manage calendars. To map out projects. Every day there’s a new opportunity to use it—for example, as you’re facing a difficult conversation with an employee, could you use AI to help you prepare?
Here’s how I think about this question and related ones: Delegate to AI only the tasks and skills you are willing to lose. AI tools can be useful companions, but do not let them replace spaces where you develop your leadership skills and capacity.
To understand, let’s envision a day when we all have AI-powered robots that can do for us all the unpleasant things we’d like to avoid. I am unenthusiastic about going to the gym, so in theory I would love to send a robot in my place—but obviously working out is not something I can outsource. The point of exercise is that your body adapts to effort over time. I can send a robot to lift weights all day, but my muscles will not get stronger as a result. The same is true in leadership, where it takes hard work over time to develop the skills that you need to be effective.
In my courses on leadership, I always cover learning how to learn. Each student brings to my classroom a unique worldview, set of experiences, and capacity for reflection.
The most important leadership skills—framing problems, exercising judgment, and learning from feedback—require you to stay in the practice yourself.
I don’t just hand them a book with all the rules needed to be an effective leader. Instead, they learn to write the rules. Leadership education is most effective when the students create their own knowledge rather than simply consume prepackaged insights that are curated for them, whether by me or an algorithm.
Armed with the knowledge they generate, they also practice the skills required to apply that knowledge. They learn to interpret their own behavior, construct personal frameworks, and distill their own lessons. In short, they do the hard work of making sense of their experiences, and through this, they master the ability to coach themselves—an essential capability that enables them to translate classroom insights into behavioral choices and improved outcomes long after any course has ended.
Let’s say you want to use AI to prepare for your difficult conversation. You could ask AI to generate a script for you to take into the meeting, and it would likely give you something that is coherent, structured, and persuasive, so you’d head into the meeting feeling prepared. As technological tools are designed to minimize effort, your preparation would have been fast and simple. After all, it is far easier to receive bullet points—whether from a lecturer, slide deck, or chatbot—than to generate them.
But simplicity is a warning sign that you likely would be underprepared and overconfident, not ready for a conversation that could veer off from the path you expect it to take. Think about what you do during that preparation time when unassisted. That’s when you wrestle with and process information, turning it into something meaningful and useful. It’s when you think through data and reflect, getting to the gist of the issue at hand. Creating your own script for the conversation involves an important distillation process during which you are expanding your capacity for leadership. There is no shortcut for this type of skill building.
That’s not to say that AI should play no role in your preparation, and perhaps the best use of AI in leadership is to guide you through cycles of hypothesis, action, feedback, and revision. When preparing for the conversation, you might start by drafting your own opening and key messages. Then ask AI to critique them from three perspectives: that of the employee, a senior leader worried about risk, and a coach focused on long-term development.
Make AI work for you
Jonathan Joaquin Jiménez is the founder and CEO of RUNG, an early-stage company that uses AI to collect and analyze the data of experience, doing that through writing prompts inspired by my book, Choosing Leadership.
“The worry is that AI will quietly take the coach’s seat,” says Jiménez, who is a former student of mine.
But RUNG is taking a different approach, designing a structured leadership-development curriculum in which human connections and experiences remain central. Its AI-supported software platform connects employees with mentors who are either at the same company or part of an external network, and then records and transcribes discussions between the pairs.
This is AI that helps an employee archive the data of their experience, analyze patterns, and track their progress—and that’s a good use case. But whether you’re helping someone else develop their leadership skills or strengthening your own, you should provide judgment, set priorities, and bring to the table a willingness to revise your own thinking. Leadership involves recognizing what kind of problem you are facing, what kind of help you should seek, and how to interpret (and sometimes resist or reject) the answers you receive.
You might also have AI generate likely emotional reactions and questions you could ask instead of statements. These ideas can expand your repertoire, but your growth comes from clarifying your intentions, values, and fears, and then practicing in real time. If you delegate that sensemaking exercise to AI, you skip practice in precisely the skills you cannot afford to lose: reading nuance, managing your own emotions, and adjusting course midconversation.
AI can support your practice of building leadership skills. It can optimize for accuracy and engagement metrics, and it can serve as a fruitful sparring partner for ideas. It can help document any experiments you conduct as part of your learning, notice patterns in your behavior, and rehearse responses you might give in a meeting, presentation, or other situation. Over time, AI could help you deepen your reflections, expand your perspective, and increase your ability to handle ambiguity.
The danger is overdelegation. If you let AI choose your goals, define success, and interpret your experiences, you stop practicing self-observation and self-coaching, which is the core of leadership learning. The most important leadership skills—framing problems, exercising judgment, and learning from feedback—require you to stay in the practice yourself. Just as you wouldn’t send a robot to the gym, you shouldn’t ask AI to replace you in any situation where you build these crucial skills. If a task is hard, before outsourcing it, think about what makes it hard and how you grow from doing it.
Linda E. Ginzel is the Konstantin Sokolov Clinical Professor of Managerial Psychology at Chicago Booth.
Your Privacy
We want to demonstrate our commitment to your privacy. Please review Chicago Booth's privacy notice, which provides information explaining how and why we collect particular information when you visit our website.