The teacher who walks in the shadow of the temple, among his followers, gives not of his wisdom but rather of his faith and his lovingness. If he is indeed wise he does not bid you enter the house of his wisdom, but rather leads you to the threshold of your own mind.

— Kahlil Gibran in “The Prophet”.

Early on in my leadership journey, I used to have a very directive leadership style. I am obsessed with efficiency. So during team discussions I would often jump in with my opinions, make quick decisions and basically guide the team quickly to the right path (from my perspective). I thought this is what good leaders are supposed to do, and I was amazing at this. Move things forward as fast as possible, and maximize team productivity and efficiency, right?

It took me years to get rid of this bad habit.

This is still my weakness, to be honest. But often I am able to catch myself doing this, and correct myself. Now, when I get into a team discussion, I lead with open-ended questions. I try not to bias things one way or the other. I find that often things come up that I had not thought of, and I actually learn a lot more about all the aspects of that decision. Most importantly, when the team makes a decision, it is not something forced down from the top. As Dwight D. Eisenhower once put it: “Leadership is the art of getting someone else to do something you want done because he wants to do it”.

It’s not a democracy & free for all, for sure. Sometimes I am not in agreement with something, so I will ask more questions, poking holes in things. And sometimes it’s a gut call and I may just have to put my foot down and make a decision that nobody wants, or nobody wants to make. But it is the last resort. And I learnt that this is the only way to scale to big high performance teams, instead of being the only superstar in town.

Turns out, all these years of leadership training is coming very handy when I work with Claude.

I am relatively new to AI tools. So in my early days I would tell Claude what I want to get done: like implement this piece of code, or research this. Whatever. But now, when I am architecting a high level system with Claude, I don’t start by telling it what I want and ask it to fill in the micro-details. I start by asking very open ended high level questions, like how to architect this system, what to think about. And often I find myself pleasantly surprised by seeing it come up with things I had not thought of. It makes many mistakes for sure. Like an overeager, overconfident intern. But then I poke holes in what it produced, and we iterate together. At the end of many iterations we come up with a product that is much better than what I could have imagined in the first place.