Skip to content

Blog

The world is changing fast. Post-Covid, and with climate change on our minds, we have to ask ourselves: Do we really need more machines? More physical stuff? Do complex hardware systems actually help us communicate better? I’m becoming more skeptical. Maybe language, shared through the platforms we already have, is all we need.

Memory, data, latent spaces, and digital assistants—these are the things I’m obsessed with.

  • 8 min read

Explorer plus loin, comprendre ensemble

A note on language

I chose to write this essay in French, my native language. This is a deeply reflective piece, and writing in French lets me express the nuances, tensions and uncertainties with the precision they deserve.

Read it in your language: copy the prompt below using the copy button, paste it into your favorite AI, and add your preferred language at the end before sending.

Please read the article at https://xbasset.com/blog/2026/09/09/exploring-further-prospering-together/ and translate it in full, preserving the author's voice, nuances, questions and distinctions between aspirations, hypotheses and established facts. Keep the headings and links. If you cannot access the article, ask me to paste its text rather than guessing its contents. Translate into the following language:

Nous cherchons à comprendre, à créer, à soigner, à vivre mieux. Nous développons des connaissances et des techniques pour repousser certaines limites et ouvrir des possibilités nouvelles. Nos motivations diffèrent : curiosité, besoin, désir d’émancipation, reconnaissance ou pouvoir. Elles nous mettent en mouvement sans nous donner une même définition du progrès.

L’intelligence artificielle prolonge ce mouvement. Nous construisons des systèmes capables de participer à nos recherches et, potentiellement, d’accélérer leur propre amélioration. La lecture d’An Alien Mind, de Jakub Pachocki, m’a conduit à interroger la direction de cette puissance : quelle aventure pourrait donner aux humains des raisons de la développer ensemble, et quel avenir voudrions-nous partager avec ces intelligences ?

Je propose d’explorer une ambition : comprendre davantage l’univers et rendre son exploration progressivement accessible, afin d’élargir nos possibilités de prospérer. Un projet capable de réunir des aspirations différentes, tout en prenant au sérieux les rivalités qui continueront à nous traverser.

La Terre entourée de trajectoires entrelacées qui s’ouvrent vers l’espace, sans figurer la forme des futurs explorateurs.

Ouvrir des possibilités au-delà de notre domaine actuel, tout en gardant la Terre vivante au cœur de l’aventure. Illustration conceptuelle générée avec une IA.

  • 5 min read

The Git Reporter: How a morning automation became a team of AI reporters managing a newspaper

It started as a Clawd automation project.

The first idea was almost naive: what if an agent could wake up every morning, look at what changed in code I was interested in: agents and understand why it mattered, write a small article, and post the result back to me?

Not a dashboard. Not a changelog. A tiny article made from source code. I wanted to read something I cared about rather than scrolling click-bait social media and news

The Git Reporter homepage showing the latest daily edition

  • 3 min read

The Rise of Middle-Stage Artifacts in AI Coding

I recently came across an article in HBR attempting to make sense of how people use AI. Honestly, it felt a bit superficial—like feeding a few reports into an LLM and calling it a day. It missed the deep behavioral analysis that we really need right now.

But it did spark a conversation with a friend about what’s actually changing in the daily life of a developer. And there is a lot going on.

Middle Stage Artifact

  • 4 min read

The Way We Want to Work with AI: Empowering Care and Connection

New technology is constantly reshaping how we spend our time. Right now, the world of "work" is on the cusp of a generational shift, especially when it comes to tasks that demand our focused attention. The catalyst? A cognitive intelligence tool most of us simply call "AI."

alt text

I've had countless conversations about this, and one thing is clear: this transformation is undeniable. The technology is accessible to everyone, remarkably user-friendly, and often free to get started with. So, the logical step is to simply begin.

  • 3 min read

The Evolution of My Coding Muscle

I've spent the last 20 years coding.

My brain thrived on programming languages: C, Java, Python, Dart/Flutter, JS/TS.

I learned tricks, optimized algorithms, and debated design patterns.

Crafting solutions was a meticulous skill—like building precise machinery.

Then AI-powered coding tools arrived. GitHub Copilot. Cursor.

Suddenly, my routine transformed.

calculator

  • 6 min read

Dynamic Confidence Estimation for LLM Task Execution

How do you measure the reliability of LLM outputs in production?

Run tasks multiple times, track dominant results, and calculate dynamic confidence with statistical methods (LLN, CLT). Practical guidelines and code in this post.

As developers, we often use Large Language Models (LLMs) in real-world systems where not only the answer to a task matters, but also how confident we are in that answer. In an experiment, I ran a simple task—"How many R's are in 'strawberry'?"—103 times sequentially and independently on the DeepSeek-R1-Distill-Llama-8B model. This experiment illustrates two key aspects of a production system:

  1. Task Result: What is the output of the task?
  2. Probabilistic Confidence: How much trust can we put in that output?

result

When we observe successive executions, both pieces of information evolve. The outcome of the task can be inferred using principles like the Law of Large Numbers (LLN) and the Central Limit Theorem (CLT). But if the results do not converge to a stable answer, it might mean that the task is inherently ambiguous for the model. On the other hand, the confidence is evaluated dynamically based on the number of executions and the distribution of outputs.

In other words, you want to ask, "Now that I've done X executions, and I see that Y of them point to one particular result while the rest differ, how confident am I that this result is indeed the correct one?"

  • 3 min read

I've saved 13% disk space on my Mac today

Today, I've cleaned 65GB safely and I feel better now. That's ~13% of my disk space.

alt text

See this big gray "System Data" space? That's where I've been cleaning a lot stuff.

What's annoying as a developer, is that it keeps growing and that the built-in MacOS storage space optimization tools don't really get it.

And I never really took time to understand it before it was my last viable option before buying a bigger Mac.

Yesterday, even the 'magical' Store in iCloud didn't change a thing.

alt text

So I had to figure out what was going on.

  • 4 min read

How Realtime Voice AI is Transforming Digital Interactions

This post is part of a series of reflexions on the upcoming transformation of digital interactions:

  • Use Cases
  • Technical (coming soon)
  • Responsible Design (coming soon)

Early November 2024, I spent a day with the OpenAI team in Paris. A few weeks earlier, the announcement of an API to allow developers to create multimodal assistants—voice + text—in real time had made waves. I was fortunate enough to have time and receive firsthand advice to explore this new Realtime API with the team that designed it.

This technology is what powers the ChatGPT app in voice mode, and it's now available to developers for creating new applications.

I was blown away.

In 6 hours, I prototyped an interaction for a cooking coach. 100% voice-based, with beautiful sound quality, expressive intentions in the voice, and emotion. User-driven interactions, assistant-driven responses, contextual information references—all orchestrated with very simple code for a highly encouraging result.

  • 3 min read

Rico LeBot • Real-time Voice Interface Toolkit

In November 2024, I've had the opportunity to participate in the OpenAI Builders Lab in Paris, where I was able to explore the potential of the Realtime API. I was amazed by how quickly I could build a prototype for a real-time web-based voice interface that uses function calls.

The initial prototype was focused on a cooking guide where you could ask for a recipe or ask for instructions on how to prepare a certain dish. It was an amazing and inspiring experience to me because I've spent the last 10 years building a similar technology. And it just works.

To go further, as I explored the possibilities of this prototype, it became clear there were some challenges to overcome before it could become a deployable product. That's why I created an open-source toolkit to address those problems.

Challenges and Solutions

  • WebSockets not well suited for longform connections: The official OpenAI Realtime API toolkit relies on WebSockets, which proved unstable for long-term sessions over HTTP. The OpenAI team suggested using WebRTC bridges for better stability. The toolkit implements WebRTC with a LiveKit integration.

  • Dynamic UI: I wanted a dynamic UI that could respond to user input in real-time. This meant connecting the function calls from the model to the front-end functions using remote procedure calls (RPC) over WebRTC. This brings the voice interface to life, allowing users to interact with different functions of the app seamlessly.

  • Architecture: A clear separation was needed between the web app's backend, the AI agent's backend, and the front-end. The toolkit achieves a lightweight design that is modular, has few dependencies, and is easy to use.

  • Roles: To quickly iterate on the user experience, you need to refine the prompts / instructions of the agent very frequently. That's implemented through an architecture where the 'roles' are separated from the code, and allows to add and modify them very quickly

Rico LeBot – Screenshot

  • 4 min read

Evolution of jobs

I remember when I was a kid in the 90s, my mother, a pharmacist in the hospital, telling me about her struggles with the new computers she had to use. She felt like these computers were stupid tools because they didn't understand her, and she often came home disturbed and uncomfortable with this new way of working. It was not a choice of her to use them, it was kind of forced with limited bandwith to adapt to it, and that was a bad feeling.

At the same time, I remember my own passion for tech growing, wondering why my mother was so bothered by it while I found so much joy in it.

Jobs have always evolved over time. One notable example is the blacksmith, a profession that saw significant growth from the Middle Ages with the development of the population and the strong need for the agricultural world.

Maréchal Ferrand