-
When the Consultant Gets Consulted by a SQL Injection
McKinsey’s AI platform Lilli was reportedly breached by an autonomous agent in two hours, exposing a familiar truth: digital prestige is not the same thing as security.
-
Nobody With Claws
A wry look at Nvidia’s reported NemoClaw move, tracing the name from classical “nobody” to the modern dream of open-source AI agents with enterprise claws.
-
Your AI Writes Tests That Pass. That Is the Problem.
AI-generated tests often create false confidence by checking that code runs, not that it is robust, correct, or secure.
-
The Fragmentation of Knowledge Work
Anthropic’s data shows AI isn’t replacing jobs yet—it’s fragmenting them. For programmers and analysts, the deeper risk may be the slow erosion of expertise
-
When Donald Knuth Lets an AI Do the Math
Knuth meets AI: Claude helped crack a combinatorial puzzle about Hamiltonian cycles that resisted neat constructions—prompting Knuth to reconsider generative AI.
gekko
