Menu
ai
Let's look at Jev, the AI that decides without ever writing a sentence
You have a feature that needs a bit of common sense: sort, detect, decide. The reflex is to call an LLM and repair its output. Let's look together at Jev,...
When code complexity lives in the prompt
A file that shows up 'green' in PhpMetrics, but gets patched every month: static analysis got it wrong. In an application that integrates an LLM, half the logic lives in...
The developer's craft has been dying for 40 years. And it has never been healthier.
An owned opinion: what AI destroys (producing code by hand) was never the heart of the craft. It re-splits the analyst from the programmer, hands the code to the machine,...
LLMs in Production: Why an 8-Word Prompt Beat All the Others
.callout { margin: 2rem 0; padding: 1.2rem 1.5rem; border-left: 3px solid #2d6a4f; background: #e8f4f0; border-radius: 0 6px 6px 0; } .callout p { margin: 0; font-family: 'Lora', serif; font-style:...
LLMs can code. But can they maintain?
/* Diagram comparatif */ .swe-diagram { display: grid; grid-template-columns: 1fr auto 1fr; gap: 0; align-items: stretch; margin: 2.5rem 0; border: 2px solid #e5e7eb; border-radius: 10px; overflow: hidden; background: white;...
Speech Vectorization Explained: Building a Local AI for Pronunciation Detection
Reading or writing has never been much of a problem for me, but as soon as it comes to speaking, I realize my pronunciation is not always clear.
© Jean-François Lépine, 2010 - 2026 · Flux RSS