Why this series. And why the books are free.
Knowledge grows when we can approach a problem, understand it and share what we have learned. This is where Mattone dopo Mattone begins.
Read the articleTHE SERIES BLOG
Questions, examples and reflections to continue the journey of the books. A small space to understand how we learn and what we build.
Follow the articles via RSS ↗Knowledge grows when we can approach a problem, understand it and share what we have learned. This is where Mattone dopo Mattone begins.
Read the articleA document comparison is interrupted. Resuming needs memory, notes and a record of steps: follow what Velia preserves and how she chooses where to continue.
Read the articleI imagine AI that can return to a project, reconsider what it has learned and choose how to proceed. That vision guides Velia and Little Velia, one brick at a time.
Read the articleBuilding an assistant that uses tools began with continuity: a request, an outcome and a fresh decision. This chapter follows the agentic loop.
Read the articleIdentity, memory and initiative are the first bricks. Monte Carlo enters the design as a tool she will be able to choose when exploring alternatives.
Read the articleA new part of the series follows two projects: an assistant that coordinates tools and a small agent whose interests, initiative and evolution we want to explore.
Read the articleComputing power keeps growing. I want the care we take with it to grow too: in our code, our resources and the attention we bring to programming.
Read the articleI found something zen in working with language models. It became a book, a few questions about myself and an apology owed to those who truly know Zen.
Read the articleFrom the Commodore 64 and the Amigas to the DGX Spark: different eras share a table. Along with the same urge to switch on, understand and explore.
Read the articleThe code arrives in seconds, with a flawless explanation. Before celebrating, I ask it a few awkward questions.
Read the articleA well-written answer can easily win me over. Before agreeing with it, though, I want to know what it actually knows about my problem.
Read the articleThe book’s program works. Good. Now I try changing something: that is where it starts getting interesting.
Read the article