Human-in-the-loop has become a catchphrase that means everything and nothing. Systems are described as having human-in-the-loop when a human reviews an AI output, when a human approves an AI decision,

Low-code AI platforms promise to put machine learning in the hands of people who cannot write Python. The pitch is compelling: connect your data, configure a pipeline with drag-and-drop components, de

A specialty insurance firm underwriting commercial property policies received submission packets as PDF documents. Each packet contained an ACORD application, loss runs from prior carriers, a statemen

When organisations adopt AI, they plan for the first-order effects: new tools, new skills, new models. They almost never plan for the second-order effects: how AI reshapes the relationships between pe

Data pipelines break in production more often than they should, and the breakage is expensive. A pipeline that silently produces wrong data for three days before anyone notices has corrupted downstrea

Partnership announcements in AI have become a quarterly ritual. Two companies issue press releases about a strategic alliance, exchange compliments about each other's technology, and promise integrati

A reference librarian organises a library by subject, not by format. A book on butterflies and a photograph of butterflies and a scientific paper on butterfly migration all live under the same subject

E-commerce search has a problem that keyword matching cannot solve. A user searches for "light summer dress for beach wedding" and the system returns results matching those exact words. The product ca

Post-mortems are useful. Pre-mortems are cheaper. A post-mortem tells you why a project failed after the money is gone. A pre-mortem tells you why a project might fail while you can still change cours