Ideas
For most of my career I’ve been handed problems described as people problems. That was rarely the whole story. What I usually find is a system doing exactly what it was built to do. Every business problem is a people problem. Every people problem is a systems problem. Three questions come out of that, at three altitudes.
The ideas
-
The Future of Work
We’ve spent a decade getting better at how work gets done, often losing sight of what it’s for. Technology facilitates the middle of the value chain. People start it and people finish it. The future of work isn’t something we inherit. It’s something we design.
-
The Human Side of AI
Your job is not to manufacture certainty. It’s to architect trust. What gets called resistance is usually a threat response to three specific things, and most implementations neglect it.
-
Engineering Performance
Your performance system was designed for people who don’t exist. One process is running two incompatible jobs, evaluation and development, and the brain, the math, and the manager all strain under the contradiction.
The Future of Work
The question isn’t how fast. It’s toward what.
We’re rebuilding work faster than we’re asking what it’s for.
Work is being redesigned right now, at speed, against a variable few of us chose out loud. Faster. Cheaper. Fewer. Those are answers. The question underneath them usually goes unstated.
Here’s the question: For the sake of what?
Technology facilitates the middle of the value chain. It rarely starts it and it rarely finishes it. Work starts with the creativity to notice a problem worth solving and ends with the judgment to serve a real person on the other side. Lose sight of those two ends and you can optimize a chain that no longer connects to much.
Watch where the efficiency goes and you’ll learn which one you’re building. Much of the time AI saves gets recaptured as additional output. That’s closer to overclocking human capacity than optimizing it. And it’s hard to move people toward meaningful work before the work is decent: in our study of the US workforce, 44% showed signs of workplace angst.
Efficiency is a legitimate goal. Efficiency without a stated purpose is how organizations drift into places nobody wants to work. That tension doesn’t need resolving. It needs designing for.
The future of work isn’t something we inherit. It’s something we design.
The Human Side of AI
The question isn’t whether they’ll use it. It’s whether they trust you.
Your job is not to manufacture certainty. It’s to architect trust.
Leaders are under real pressure right now to have the answers. To know how AI changes every role. To promise nobody gets left behind. Promise a future you can’t see and people tend to sense it. Trust goes first, and the transformation usually follows it out.
People don’t need certainty to move. They need trust.
Start with what resistance actually is. When employees hear “AI transformation,” their brains scan for three threats. Am I safe and treated fairly? Can I still advance, or does the machine take the stepping stones? Does my contribution still matter? Security, growth, significance. Much of what gets labeled resistance is a threat response. It isn’t laziness, and it isn’t lag. In our study of workers across the US and Europe, 61% reported significant fear and uncertainty, 70% cited clarity barriers rather than skill gaps, and 45% are concealing their AI use from the companies paying for the licenses.
A second gap sits on top of that one, and it runs vertically. Executives see the horizon. The people expected to implement the change see the cliff. That’s altitude sickness: the higher you climb, the clearer the horizon and the harder it gets to read the ground. Two different movies on the same screen.
Trust is the antidote, and it’s not a feeling. It’s a psychological state with three conditions. Credibility: do you know what you’re talking about, and are you honest about what you don’t? Integrity: do you do what you said you would, especially when it costs something? Benevolence: do you care about my well-being, not only the organization’s? Miss one and the other two get harder to hold. Trust in leadership on AI is associated with a 32% performance edge.
Which is why honesty beats clarity. You often can’t tell people the destination. You can usually tell them what you don’t know.
Engineering Performance
The manager isn’t the problem. It’s the architecture.
Your performance system was designed for people who don't exist.
The annual review assumes a manager who remembers a year accurately, rates without bias, and holds one candid conversation every twelve months. That manager doesn’t exist. We built an entire system around a person we invented.
Then we gave that system two incompatible jobs. Evaluation allocates pay and promotion. Development requires someone to say out loud what they can’t yet do. Ask a person to be honest about their gaps in the same conversation that prices them and you’ve asked them to choose. They’ll choose the money.
That’s not a character flaw. It’s arithmetic. Managers inflate to protect their people, employees perform certainty instead of learning, and 62% of the variance in performance ratings ends up reflecting the person doing the rating. Then we attach pay, promotion, and someone’s sense of their own worth to that.
Your people aren’t bad at performance. Your performance system is bad at measuring it.
The fix is architectural, not procedural. Separate the tracks: one system allocates, one develops, and they don’t share a meeting, a form, or a season. Then change what the system is for. Most performance management is built to appraise the past. It should be built to drive the decisions ahead. Stop appraising. Start deciding.
None of it runs without managers, who account for roughly 70% of the variance in engagement and receive almost no development for it. The system is made of conversations. That capability is learnable, and it’s the argument at the center of The Coaching Shift and the courses I teach at Georgetown University.