2026-03-09
There are so many things I have learnt in the past year and I can't wait to see what the future holds. This is an overview of what I have been doing and what I have been thinking lately. A yap-sesh to re-light my website.
It’s been a minute since I used my website and I feel bad for neglecting it for such a long time even though it took me a lot to build it. Even though it’s already March, I feel like I need to tell so much about what happened in the last year. I don’t know if it’s just me but sometimes I want to do so many projects, that I start preparing them and then I realize after some time that I want to move to something else: my attention span - or in this case, “attention span for projects” - diminishes every single day. I would like to pursue a long-time project, without my brain wanting to switch for some reason. But anyway, I would like to wrap my 2025 in a few words so that I can start talking about what I have been thinking lately (some of these thoughts may be a hot take, so I would suggest you take a look if you are interested) and some of my projects I have been working on. In 2025 I have:
This is certainly the era of AI. It’s weird time. Almost weekly a new ground-breaking model comes out. Everyday theres news about stupid biased benchmarks. Imaginary tech stacks and non-existent packages being installed into one’s machine. Big companies AI CEO’s talking about the world’s future in a very dystopian way. In 2026, perhaps we can say, that we are already in the 2nd era of AI (from just a web experience through prompts to AI completely taking control of your PC like openclaw), and the speed at which this technology continues to evolve day by day, is honestly creepy. I have never felt this much discomfort in coding, when I use LLMs. A lot of content creators decided to stop with the vibecoding content because it was slowly making them choke on something they didn’t wanna do. And to be fair I completely understand them, because I would have done the same. I almost fell into the propaganda: some time ago I decided I was going to develop a totally vibe-coded app just for the sake of it. But was I really eager to give everything to a clanker and lie to myself that I was the owner of that piece of code? Would that really matter, in a sea of completely vibe-coded apps? See, the fantomatic illusion of choice. There was a post I saw on X made by @dax (an opencode maintainer and co-founder) and his words are the ones I stand by:
Is it not crazy seeing how workers at one of the biggest open-source coding agents says about vibe coding? In fact, I want to clarify: I am not against AI per se. Matter of fact: progress cannot be avoided anyway, so we better accept the many positive feats it can bring in our daily life and society. History is repeating itself, just like the other big revolutions (where the most recent one is definitely the Internet revolution) and we just have to adapt, especially the oldest generations. Instead, vibe coding is what I fear, exactly the level where the neurons stop firing and we rely on a machine to base our entire codebase. But the change I fear the most, is not the one in the machines but in the people’s mind. Seeing how people are no longer interested in what’s behind the code that the LLM write, instead, delivering and shipping becomes more important than everything else. Although I get the enthusiasm of being able to build anything you want even without coding skills or knowledge into the field (and this applies to the artists side of the complaints) I feel like is important to still get to know the stuff you do. The laziness. The instant-result drug. The unconsummated desire to have what you want, right away. Man, I really do hope I am not becoming a boomer and my children won’t call me lame for the way I talk. Reader, what do you think about this? What do we need to expect about GenAI in the future? Will humans be able to tame the beast? I know one thing for sure though. It is comical.
As a conclusion, I think studying Computer Science still makes perfect sense in this era, if that’s what you want to specialize in. Even though the gap really got closed for people to dive into the field, years of studying is what makes the difference and even companies have started to realize that. Coding, even if assisted, even if automated, still need to be checked by a human being to make sure everything goes the right way. So I am confident and I will still continue to pursue what I like and strive to become better.
I always thought that being able to work in FAANG would be insane and would symbolize the epitome of one’s career. And for some extent, I still do.
But the more time passes, the more I reflect about what are the half measures to get a job and into which environment would I like to work. Would I challenge myself and do something more than just “coding”? Would I like living in a very stressful place where I could be fired at any time? Is every smile on social media actually real? How much money would a job require me to work in some questionable conditions? And some of these questions are very easy to respond while others may require some time to think about them.
As a child, I never really cared about stipend. All I wanted is to do what I loved most. Playing videogames. Yeah, that was my first love. A love so deep that whenever I hear some of the OG pokémon medleys or Super Mario Galaxy’s I instantly cry. And I grow up saying to me “I would like to make something that make people feel the same way, because I will be forever grateful for the people who created this for me”. Creating, making something from zero, was always the hobby I spent the most time with. It could be drawing, writing or whatever was going through my wacko mind. Putting this thought through code helped me realize I could really build whatever I wanted If I ever had enough time and enough desire (or just create something with whatever tool is Turing-complete). And when I learnt through some of my first friends that you can actually study this in school, I was flabbergasted. Finally something that was not as boring as maths (I learnt to love maths during High School) and I really liked at school. And the more you study Computer Science the more you learn is not about Computers. Which got me just more hooked in.
“Computer science is no more about computers than astronomy is about telescopes”
- Edsger W. Dijkstra
So, at the end of the day, I arrived to the conclusion that I don’t really care what kind of job I end up getting if I do what I like (I should have arrived to this way before, but I was blinded). I would like to experience and contribute to some open source projects, for sure, but I’ll talk about this more later.
In January, I finished my internship here in university. Here in Italy it’s mandatory to do (at least) 150 hours of internship that will then be approved and become credits. I decided to ask one of the professors I really liked who taught Logic during my first year, if I could do my internship with her. I realized little by little how Computer Science is just a fusion between Mathematics and Philosophy: what can we compute? what can we solve? Many mathematicians like Hilbert were asking this question way before machines and calculators were a thing: then Turing proposed a formalization of a machine that can do calculations. Seeing how Logic is such an important piece in this field, made me fall in love with it, making me see stuff I couldn’t really see before. As for the internships, she gave me a project to make so that I could insert it into my Bachelor’s thesis. This project aims at the implementation, in the proof assistant Agda, of a deductive system in natural deduction for modal logic S4.2. The work lies within the fields of mathematical logic and theoretical computer science, adopting a formal approach to the representation and verification of logical systems through dependent types.
The project follows the deep embedding methodology - that is, representing one logical system within another (Agda) using inductive types - with the goal of studying its metatheory and proving key properties such as normalization. Apart from Agda (which was a pretty difficult journey to embark on, because of its very difficult and different programming paradigm, fully using the Curry-Howard isomorphism) the project was really rewarding and I liked to develop something so different from what I ever have done before. I also wrote a little Agda resource to get started with the programming language (so of course I suggest the read and of course advices are welcome!), because I also needed it myself to NOT forget the language and perks if I ever focus on something else temporarily (which, by the way, I am doing right now). In this moment, I still have to start my thesis but I am really eager to continue this work.
Now that the second-last exam session has finished, I can finally get some rest and I can resume my little chores I have been doing here on the web:
I am surprised you read until the end. Thanks for staying. That’s all for now, still trying to push through. Comments or discussions are welcome! Cya :) !
Back to blog