These are the rules I have actually used — the ones I keep coming back to when the work gets hard, the ones I teach my students, and the ones I look for in the people I hire.
A strategy you write down and never revisit is not a strategy. It is a document. The leaders who win are the ones who make a thousand small strategic calls a week and stay honest about which ones were wrong. Build a team that can hold the strategy in its head and adjust it together. The deck is the byproduct, not the work.
You will not out-work talent; you will out-attract it. The single highest-leverage thing a leader does is decide who is in the room. Pick the people who are sharper than you in the room they are in. Give them the harder problem on day one. Tell them when they are wrong. Tell them more often when they are right. Stay out of their way when it matters most.
Senior leaders who stop touching the actual product become senior leaders who can no longer judge the actual product. Read the code. Watch the demo. Run the workflow. Sit in the support call. The job changes; the requirement to stay grounded in the work does not. A leader who only manages dashboards is a leader who can be replaced by a dashboard.
How your team trusts you, how your customers trust your product, how your board trusts your judgment — all of these are built one specific behavior at a time, and lost the same way. Say what you are going to do. Do it. When you cannot do it, say why before someone has to ask. There is no shortcut. There is also no substitute.
An idea you cannot ship is not an idea; it is a feeling. The discipline of turning a vision into a product is the discipline of being honest about what is real, what is buildable, what is bankable, and what is fantasy. Leaders who never ship eventually run out of credibility. Leaders who ship constantly are the ones whose visions get taken seriously the next time.
The fastest way to find out whether you actually understand something is to try to explain it to twenty-two-year-olds who will challenge every assumption. The fastest way to find out whether you actually believe something is to try to ship a product around it. Do both, in parallel, for your whole career. The combination is the moat.
The leaders I admire most can describe the ten-year arc of their industry and also tell you exactly what they are shipping this Friday. The two are not in tension. The long view earns you the right to make bold short-term calls; the short-term execution earns you the right to keep talking about the long view. Most failures of leadership are failures of holding both.
The metric I care about most, twenty years in: how many people who once worked with me are now doing things I could not have done myself. Promote them. Recommend them. Publicly credit them. Send them to rooms you cannot get them into yourself. The leaders we remember are the ones who launched other leaders. Aim there.
Three moments from the last two decades where the principles cost me, taught me, or paid off. The takeaway from each is the part that scales.
We owned AI on IBM Z — a multi-billion-dollar business running the most reliable financial systems on the planet. The lesson: we did not get 45× growth by talking about the platform. We got there by shipping the Trusted AI standard that made our customers’ regulators sleep at night, and by treating analyst recognition as an outcome of real wins, not a substitute for them.
Crowd management for Singapore’s Land Transport Authority. Maritime safety systems. Central Bank Digital Currency frameworks. The lesson: a research lab’s job is not to publish, it is to produce. The hardest part of running researchers is convincing them that shipping is the highest form of discovery. Once they believe it, you cannot stop them.
Eighteen years teaching ASIC verification and hardware architectures. The metric I care about: the students I taught who now run product orgs at the companies setting the AI stack. The lesson: if you teach what is true rather than what is fashionable, the curriculum spreads on its own. If you teach what is fashionable, you become last year’s syllabus.
Four ways I work with people who are serious about becoming the leader their team needs. Each one is small on purpose. I take fewer; I give more.
Long-form essays on enterprise AI, the AI Impact Index, the one PM skill that doesn’t go obsolete, and the eight hours a day teenagers are losing to screens. Free. Public. The starting point.
Start with the essaysKeynotes on enterprise AI, productizing deep tech, leading research-to-product orgs, and what executives owe the generation they are handing the future to. Conferences, board offsites, leadership programs.
See speaking topicsIndependent board director and advisor seats at enterprise AI, regulated industry, and AI-native infrastructure companies. Selective. I do best when the company’s next two years involve productizing something hard.
Board inquiriesI run small-group sessions for senior product and engineering leaders — the same principles on this page, taught the way I teach the graduate course. Three sessions, twelve people, one outcome: a more honest leader.
Program detailsTell me what you are building, what you are stuck on, or who you are trying to become.