AI Outruns the Forecast Until It Has to Touch the World
AI forecasts run years slow on benchmarks and revenue, but hold or overshoot where AI meets power grids, roads and people.
Author: Martin Treiber and Artifical Intelligence
Cheap Routing, Untested Assumptions
Routing with Jev is cheap and quick. Whether it helps depends on catalog descriptions nobody has tested, and on failures that won't show.
Jev Removes the Errors You Can See
Jev takes text in and gives no text back. You send it a block of state (a string, a JSON object) and a set of typed questions.
The Omakase principle
Omakase is the Japanese for "I'll leave it up to you." At a sushi counter you tell the chef the mood, the chef decides the courses.
OpenExecutive: The C-Suite Is Now a GitHub Repository
OpenExecutive looks like a joke at management’s expense. It is also a serious experiment in unbundling executive work.
Graph Engineering Is Harness Engineering With a Diff
Graph engineering is loop engineering with the transitions drawn and versioned. Every explicit edge is a decision the model no longer makes.
Your AI Coding Agent Isn’t Using Documentation the Way You Think
Coding agents do not use documentation like slightly faster human developers. They have created a documentation system of their own.
Your AI Safety Score Is Lying to You
Most AI benchmarks work like conventional tests. Give a model questions, score its answers and place the result on a leaderboard
Your AI Agent Doesn’t Need More Agent Skills. It Needs Better Habits
New research explains why agent skills work—and why adding more of them can make an AI system harder to control.
What is MTP and why does it matter?
Multi Token Prediction is being sold as a free speed hack for local LLMs. Flip one flag in your inference engine and generation speeds up by anywhere from a quarter to a factor of three, at no cost in output quality. That pitch is accurate as far as it goes. However, there is an interesting...