A fractal spectrum of tales
- Firefox’s new hidden ad-blocker
Firefox 149 has quietly started shipping the Rust-based ad-blocking engine from Brave within the browser. For now it’s part of an internal experiment and is not exposed as a browser feature. So, before you do anything, keep in mind that you are already better off if you are using uBlock Origin for instance and keeping Firefox’s
Enhanced Tracking Protectionon.If you are still curious and want to try it out, it can be done by messing around with
about:config.
2 min read - May 1, 2026 - Does University Participation Change You?
Recently, a friend asked me a variation of the question in the title. I was caught off guard. There is a general impression that participating in university politics makes you more bitter and cynical. I don’t think it is true, yet. But I have to agree that it had an effect.
It is hard to understand precisely how it changed me; you would need to ask someone who knows me and knew me before. But, for instance, I can see that I used to jump at opportunities, excited to explore new ideas, and complain about how slow it is to make anything happen and how it is all the administration’s fault. Now I find myself repeatedly pushing back, putting things in perspective and trying to reflect on what kind of resources would be needed for something to happen before jumping to conclusions. In this sense, I think it made me a lot more realistic. A friend said that sometimes it takes the joy out of things. I think it is a good way of putting it :D
2 min read - May 1, 2026 - Elucubrations on Humanocracy at University
There is a particular kind of exhaustion that everyone who works in a university will know well. This isn’t the exhaustion of thinking hard; if anything, that kind is regenerative, even pleasurable. We expect to feel tired after wrestling with an intricate proof or brainstorming how to resolve some delicate technical issue. Even when our brains feel fried, we are still energized by it.
But there is another, “scarier” kind of exhaustion: the administrative one. It is the friction of a form that requires three signatures before you can make even the most trivial request. It is the demand to specify the future edition of your course that won’t run for over a year and is still under active redesign. It is the updated privacy regulations for your research project. It is the committee that exists primarily to schedule the next committee. The list could go on and on…
12 min read - Apr 16, 2026 - Workplace Cultures
The one thing I truly value of dialogue-based podcasts is that no matter the topics, the conversation often wanders into unexpected and insightful places. This is how this post came to be, specifically when listening to the Signals and Threads episode Why Testing is Hard and How to Fix it with Will Wilson.
In these times of international turmoil, financial crunch and decreasing numbers of students, there are important discussions about restructuring our University’s governance and processes, with the aim to make its decision processes faster and operations cheaper and more efficient. So it should not be too surprising that while listening to the episode my attention kept picking up all remarks about workplace culture. Maybe it is confirmation bias, but they kept coming back to highlight how collaboration, honesty, and shared learning together create a working environment that is both constructive and rapidly adaptable.
11 min read - Apr 12, 2026 - Students and AI resources
Over the past few years, I have tried to collect useful material to point our students to in my AI policy pages. The idea is to provide nuanced resources that explore uses and misuses and ethical concerns, and promote critical thinking and reflection.
This morning, when I came to add a new resource, I was absolutely convinced that I had a blog post about this already.
Turns out I did not. So let’s fix this once and for all.
2 min read - Mar 28, 2026 - Mechanizing Mathematics
The day before I finalized my previous post, I was attending our Institute’s annual public lecture: the Bernoulli Lecture, named after Johann Bernoulli, who worked for a while at the University of Groningen.
This year’s lecture, Can Machines Do Mathematics?, was organized with Studium Generale Groningen and given by Johan Commelin, mathematician at the University of Groningen, director of the Mathlib initiative and co-organizer of the recent workshop on Mechanization and Mathematical Research at the Lorentz Center (it ended with a public symposium which is well worth watching as well).
6 min read - Mar 26, 2026 - Mathematics, Beauty and the Shadow of AI
There is this interesting stereotype that mathematicians deal with the cold, hard truth. Yet, even the most technical mathematics, deep down, is highly warm and playful. Most mathematicians long for the moment a proof finally clicks. This is something beyond its correctness: when it clicks, it also feels right and looks right. It usually comes with a sense of elegance or beauty, be it because of some surprising connection or for remarkable simplicity. It is common that in those cases one wonders—after the fact and forgetting the long struggle that led there—how come no one has seen it before.
13 min read - Mar 19, 2026 - Maths and Assessment in the Age of AI
The commoditization of AI—and its rapidly improving ability to solve, explain, and generate mathematical problems—is increasingly getting in the way of our teaching and, even more significantly, disrupting our usual assessment methods.
For instance, many courses have seen an surge in near-perfect, overly elaborate solutions whose quality is rarely reflected in final exams. Discussion and active participation seem to be declining, replaced by a reliance on LLMs as convenient—albeit impersonal—tutor replacements. This is concerning, as research and experience consistently show that collaboration and productive struggle are key ingredients of learning.
7 min read - Mar 15, 2026 - Fixing an infinite recursion in OCaml applicative parsers
Nine years ago, while I was still a fresh OCamlr, I gave a presentation (later turned blog post) with title First steps with Category Theory and OCaml. The idea was to show how to implement the basic typeclasses of Category Theory in OCaml, and how to use them to build a simple parser combinator library, by following what I had learned from Haskell. At the time, not knowing better, I stumbled upon an annoying issue: if you look at the code examples for the
Alternative_Utilsmodule, you will find the following commented-out code.
22 min read - Feb 27, 2026 - Circular Time
’Forty-two,’ said Deep Thought, with infinite majesty and calm. – Douglas Adams, The Hitchhiker’s Guide to the Galaxy
The narrative on AI keeps swinging between panicked existential threats and overoptimistic hype. Last week, while flu forcefully made sure I sayed in bed, I had time to read some articles in my backlog and really found it disappointing how the discussion has stagnated and keeps revolving around the same points.
None of the ideas look anywhere new. In The Hitchhiker’s Guide to the Galaxy, Douglas Adams already brilliantly satirized on the quest for superintelligence and what will happen with it.
3 min read - Feb 17, 2026