Photo by Alexas Fotos.
The Peter Principle crowd is clear. You get promoted for being a super-IC and then you step into management and it’s an entirely different job. No offense, but you’re terrible at it. And we can’t promote you again but we also can’t unpromote you. And that’s why there’s sadness in the world forever and ever and nothing gets better and work sucks and then you die.
We’re not huge fans of the Peter Principle. In part because that’s a dim-ass view of work and sorta defeatist as an approach. And that’s before you reckon with the fact that we spend a third of our waking hours at work. So not only is it hopeless, it’s long-running and hopeless. No thank you.
Our sunnier take starts with the idea that humans are quite good at learning, even bosses. And you can take someone who has excelled at individual craft and help them get better at managing a team of people who do that craft. It’s also lovely cause when the person running the team used to do the work, there’s built-in empathy and credibility.
And that’s how it goes some of the time. But not always…
Sometimes you step in to manage a team and the work they are doing is work you’ve never done. You not only haven’t picked up those tools yourself, you’re deeply fuzzy on the details. Goodbye, built-in empathy. So long, credibility. This head-spin moment is deeply uncomfortable. You are at best an imposter. At worst a fraud. And everyone will soon know it.
Faced with searing ignorance and a deep, primal instinct around self-preservation and intragroup dynamics, you do what anyone would do. You read a book about design the night before your first meeting with the team. You’ll be managing a team of designers and yes, sure, you draw stick figures. But they don’t need to know that.
Spoiler alert: they notice
We’ve heard this story a bunch. We’ve starred in this story at least once or twice. It doesn’t go well. The mortifying, terrible thing you are trying to avoid is the precise thing that happens. You enter a room of experts and you feign expertise. And experts are experts. So they don’t buy it. How would they? You and your single evening of reading about design up against SCAD grads? FFS, you never stood a chance.
Here’s the thing. The designers were already skeptical about reporting to someone without a design background. And you just proved their point. How will they get feedback on their work? How will they grow in the role? And, crucially, how the fuck are you going to represent the needs of the team at meetings they aren’t in when you not only don’t understand their work but have shown a willingness to fake it and hope no one notices?
This is a disaster. And worse than a disaster, it’s offensive. Because, and again, this part is important so listen close. Experts are experts. There’s a difference between hard-won wisdom and a hastily consumed glossary of design principles. The moment you conflate the two, you devalue the distinction. And for folks who have spent their entire lives getting good at a thing, that shit is downright disrespectful.
You might be growing worried that we’re back to the Peter Principle and everything is just going to be awful forever. We would never do that to you. But if we’re gonna talk about the solve, we need to acknowledge that the fake-it-and-hope-no-one-notices approach to management is having a real moment right now. You don’t even have to read the design book. Just generate a tl;dr of the corpus, some talking points for the meeting, and a recommended change in strategy after uploading the last 6 months of meeting notes.
The tools are getting faster but the core issue was never that you were bullshitting slowly.
Management by centaur
There are people who will tell you that the current version isn’t even faking it. Not really. The metaphor of the moment is that the leaders of the future are centaurs — half human, half-machine. You do the human parts, let the machine do the machine parts. Sure you can have a bot summarize, but that’s not very agentic of you. Just have Chat watch for new designs being uploaded, and write the email directly. You just focus on the human parts. Which parts are the human parts? Well, the bot companies benchmark themselves based on how many of the human parts they can do. So, you know, whatever’s left.
We’ve spoken with execs lately who feel like every 1:1 with their CEO has become a 1:1+1, because there’s an LLM copilot listening in and offering private suggestions. If you’re Sam Altman, maybe you imagine that that goes over really well. You might imagine that this person feels like they’re getting better management from their centaur-boss than they were ever gonna get from the all-natural kind. People can imagine all kinds of silly things.
What actually happens, any time we hear it brought up, is that the person feels disappointed and disrespected. Not a lot. Not enough to make a big stink over. Just a creeping sort of shittiness about the interaction. It’s not that the content of the conversation was so terrible, per se. Or even that the LLM said stupid things, though they were kind of shallow. It’s just like you had a whole conversation with someone who was looking at their phone instead of you — nodding and pretending they were paying attention. Like they didn’t care enough to just engage directly and undivided.
If you’re really committed to the centaur bit, this is irritating. Everyone should love a boss with a backstop. They should be grateful that their boss has a tool to give more domain-specific feedback. This is way better than reading a design book the night before — the bot has stolen every design book ever written! The technical pieces are handled by the machine so the boss can focus on the human parts. They should be happy, damnit.
The history of management is full of people with intuitive-to-them theories of how organizations should work and how humans should feel. But any successful management toolkit starts by meeting people where they are. Where people actually are is that small slights can have significant impacts on workplace productivity, engagement, and performance. Where people actually are is that when you bullshit them, they know it. And it’s still fucking offensive.
Management is the human part
We get that the faster things go, and the more senior you get, the louder the imposter syndrome grows. The more it can feel like whatever you would write on your own, or say on your own, or think on your own, maybe you should just have the bot polish it up first. There might be other places in your work where that impulse makes sense. But when it comes to your people, it’s gonna work out better to arrive as your human, not centaur, self. Management is almost entirely the human part.
Because here’s the thing. Your job is not actually to be smarter than your team. Your job is to get brilliant work out of them, to keep them growing and engaged and supported and sustainably driving the organization forward. The more senior you get, the more you should expect your people to know more than you do about their domains. That’s not your failure, that’s your team’s strength.
So read the book. Actually. Yourself. But read it as a sign of respect for their craft, not as an attempt to usurp it, or to pretend that you’re suddenly peers on the subject. Design (or engineering, or sales, or copyright law, or carpentry) is their craft. Management is yours. There are lots of ways as a leader to take accountability for your team’s growth and development without delivering it all yourself.
And as for your craft, a big part of your job is to listen. To the actual conversation, not a summary. Listen to what’s bothering them or getting in their way. Listen to their ideas for things they’d like to try, or opportunities they see for the team. And, without delegating it off to a bot, actually sit with that and figure out what you think. That critical thinking informs the decision making that every org looks to their leaders for. Doing it well is a deeply intuitive, empathetic act.
Wanna know what’s better than a centaur that can bullshit about skills that take 20 years to learn? A human manager that doesn’t.
— Melissa & Johnathan
