Articles
The Work Was the Teacher
Producing raw work was never only about the output - it was how judgement got built, and how a manager read who had it. As AI takes over the doing, we're asked to verify more with expertise we're building less, unless we put the formation back in on purpose.
Hand It Over, or Think It Through?
Being good at AI is usually framed as fluency with the tool. The real test is whether the work can be cheaply checked - hand that over and let AI run it - or whether it turns on judgement you could never verify, where you keep the pen and use AI to think against rather than decide for.
You can't manage what you can't measure
Adoption is easy to measure and capability is hard, so most organisations count usage and treat it as a proxy for what their people can actually do - but the two come apart, often dramatically, and the gap is wider than the usage numbers suggest.
Before You Decide Who Owns AI, Answer These Two Questions
The debate over who owns AI is premature. Two questions come first - what AI is actually for in your business, and what you say when someone asks how you use it - and answering either one well depends on the leader having worked with AI directly.
Trust nothing and verify everything: Why the best AI policy in law wasn't enough
Sullivan and Cromwell had the best AI verification controls in the legal profession, and they still filed 40 fabricated citations in a federal court - this is what the failure tells us about how human judgment interacts with AI output, and what firms can do differently.
The Glorified Google Trap
A leader's belief about what AI is determines what they attempt with it. For most, that belief is based on an experience they had six to twelve months ago.
What separates leaders who get value from AI?
Most organisations invest in AI training. The research suggests the skill that actually matters is one they already have - but have not connected to AI.
AI Is Not a Convenience Technology
The gap between those who use AI well and those who don't is compounding. Basic use isn't good enough (unlike for a microwave).
Experience and the AI Adoption Gap
The adoption gap is shaped less by the technology and more by the experience leaders have with it.