Thoughts on AI, engineering leadership, and building products that matter.
A knowledge firm's moat used to be what its people know. When frontier intelligence is rentable, that drains, and the moat relocates to what the firm values, and to whether those values are expensive enough to be real. Software can hold the answer. It cannot hold the hand.
Read moreNon-developers are already shipping software inside regulated firms. Lock it down and you forfeit the productivity transformation; open it up and every department becomes shadow IT. There's a third path, and during a narrow 12-to-24-month window, your developers are the ones who build it.
Read moreIf AI handles the code, the engineering discipline doesn't disappear. It migrates — to specifications, tests, architecture, and oversight. Most organizations haven't noticed yet, because they're measuring the wrong thing.
Read moreEvery large organization rolled out AI tools to make people faster at what they already do. That's the problem. Giving everyone an AI assistant is fixing one step on a staircase that still spirals around outdated processes.
Read moreThe cost of building software is collapsing toward zero. In 2026, AI agents aren't a forecast—they're a $10 billion market reshaping how work gets organized. Here's how to navigate the explosion.
Read moreAfter two years of pilots and proof-of-concepts, asset management firms face an 18-month window to cross the threshold from AI experimentation to production-scale deployment, or risk watching competitors capture advantages they can never match.
Read more