A few days ago, Unsloth made Qwen3.8-27B available, and I downloaded it to run locally with Ollama. I had heard about the benchmarks for this model, so naturally it caught my attention right away. Not because running an LLM on your own computer is still something new, but because we are starting to see relatively … Read more
In seven days, Codex processed 65 billion tokens for me. That does not mean it produced 65 billion tokens’ worth of value, but it does say something about the way agents are changing both the duration and the intensity of work. A few weeks ago, I installed Who Burned More, a small open-source tool that … Read more
Quando usi un coding agent, la tentazione più naturale è partire dal risultato. “Costruiscimi un’applicazione che faccia questo.“ Descrivi la funzionalità, magari indichi linguaggio e framework preferiti, lasci che l’AI cominci a lavorare. Dopo poco compaiono la struttura del progetto, il database, alcune API, l’interfaccia, forse anche una prima serie di test. Il risultato può … Read more
Quando oggi si parla di sviluppare un’applicazione basata sull’intelligenza artificiale, quasi inevitabilmente la discussione comincia dal modello da utilizzare, dalle funzionalità che dovrà avere l’applicazione, dall’interfaccia e, naturalmente, dal prompt. È comprensibile. Sono le parti immediatamente visibili del prodotto e sono anche quelle che permettono di ottenere velocemente qualcosa da mostrare. Si collega un LLM, … Read more
L’illusione di costruire software con l’AI senza prima costruire le fondamenta Da qualche tempo, nelle aziende, si ripete una scena sempre più comune. Qualcuno presenta un’idea, magari ancora poco definita, e prima ancora di discutere requisiti, vincoli e impatto sul sistema esistente arriva la frase: “Che ci vuole? Con l’intelligenza artificiale possiamo farlo in pochi … Read more
Using large language models for personality-based job-fit assessment, and the calibration rabbit hole that followed. The Challenge My boss asked me a question that sounded simple: “Can you get an AI to generate a psychometric scoring formula from any text description?” I said yes before I fully understood what I was agreeing to. Here’s what … Read more
While building OtterForm, I had one of those moments where AI felt like a superpower and a limitation at the same time. At the beginning it is incredible with features that moved fast, boilerplate disappeared, and it genuinely felt like we had added another developer to the team. But as OtterForm grew into a real … Read more
This project began because I was fed up with constantly bookmarking website after website to remember where I saw that pesky little utility… Need to convert JSON? Google it. Need to format some XML? Google it again. Need to create a UUID? Once more, you got it… google, click through 3 ad-riddled websites, and discover … Read more
For a while, my workflow with AI looked like everyone else’s. Ask for a feature. Get an answer. Refine it. Ship it. Fix what breaks. Repeat. It’s the classic loop, it’s called the ralph-loop: one plugin, one model, one thread, one endless iteration. And to be fair, it’s great when the task is small and … Read more
I have been making pizza for the past 25 years and one of the most difficult things to do was accounting for yeast activity when room temperatures fluctuate and different flours behave unpredictably. For this reason, based on a study published by J. White and D. Munns, I decided to build a tool that takes … Read more