Newsletter

There is no secret sauce for data integrity.

It always starts the same way: “Data is the new [insert cliché here]!” “We must become data-driven!” “AI first!” At some point, someone will point out that this will require a lot of high-quality, trustworthy data on an ongoing basis. Nodding. Then … Read full article

Can AI clone your best people, finally “scaling expertise” like software? It’s a seductive question, but not quite the right way to think about the issue.

The marketing promise is simple: extract what the veterans know, package it into an algorithm, and suddenly every junior performs at a senior … Read full article

The purpose of a metric is to focus attention. That is also its greatest danger.

When you measure something, it doesn't just become visible. It actively pulls focus and resources from everything you don't measure. A hospital starts tracking patient throughput. The metric improves. Unseen: the nurse who no longer … Read full article

“Break the data silos” is a slogan, not a diagnosis. Most messy data isn’t caused by malicious territorialism, but by well-intentioned people solving the problems in front of them.

The siloed executive jealously guarding a spreadsheet is a convenient villain. More often, however, that’s a comforting myth we tell ourselves … Read full article

A carpenter doesn’t discuss with their client what chisel to use. Yet in data and AI, the tool tends to become the conversation.

A carpenter’s tools stay in the bag until the job is clear. It’s about measuring the room, talking about shelves that will hold, doors that will close, … Read full article