Fabio Filasieno ❄️

2.2K posts

Fabio Filasieno ❄️ banner
Fabio Filasieno ❄️

Fabio Filasieno ❄️

@FilasienoF

On compiler and database system implementation. On system programming pragmatics. I share papers that I love.

Milan, Lombardy Katılım Mart 2022
151 Takip Edilen931 Takipçiler
Sabitlenmiş Tweet
Fabio Filasieno ❄️
Fabio Filasieno ❄️@FilasienoF·
@netreinstein This is my flag. The EU does not represent me. Long live a Europe of free and sovereign nations.
Fabio Filasieno ❄️ tweet media
English
2
3
25
1.3K
Rohan Paul
Rohan Paul@rohanpaul_ai·
Chamath on how AI agents are making the "10x engineer" distinction disappear because the most efficient "code paths" are now obvious to everyone. Just as AI solved chess and removed the mystery of the best move, AI is doing the same for coding, making the process reductive and removing technical differentiation. "I'm going to say something controversial: I don't think developers anymore have good judgment. Developers get to the answer, or they don't get to the answer, and that's what agents have done. The 10x engineer used to have better judgment than the 1x engineer, but by making everybody a 10x engineer, you're taking judgment away. You're taking code paths that are now obvious and making them available to everybody. It's effectively like what happened in chess: an AI created a solver so everybody understood the most efficient path in every single spot to do the most EV-positive (expected value positive) thing. Coding is very similar in that way; you can reduce it and view it very reductively, so there is no differentiation in code." --- From @theallinpod YT channel (link in comment)
English
255
93
945
738.7K
Fabio Filasieno ❄️
Fabio Filasieno ❄️@FilasienoF·
@oost_marcel Yeah… then try to pay taxes, open a bank account, etc., in Italy 😂 These are idiots. You cannot do things like that without a nation because of taxes and local laws, and the EU is not a nation. Italy, France, Spain, etc., are. The EU will fall. It's unsustainable.
English
1
0
14
992
Marcel van Oost
Marcel van Oost@oost_marcel·
🚨𝘽𝙍𝙀𝘼𝙆𝙄𝙉𝙂: European Commission President Ursula von der Leyen unveiled EU–INC, a new framework that lets you launch a company in 48 hours for under €100 Starting a company across the EU today = 27 legal systems, 60+ company structures 🤯 That might be about to change… The European Commission just introduced 𝗘𝗨 𝗜𝗻𝗰., a new optional corporate framework designed to make Europe actually function like one market. Here’s what stands out: → Set up a company in 48 hours → Cost: < €100 → Fully online, no minimum capital → One single framework across all EU countries → Easier share transfers & fundraising → EU-wide employee stock options (huge for talent) Especially the EU-wide stock option plans, taxed only when employees actually sell (instead of when granted) is huge. This makes it far easier for startups to attract and retain top talent, finally putting Europe closer to the US playbook. Source/More info: ec.europa.eu/commission/pre… In short: This is Europe trying to compete with the simplicity of a Delaware C-Corp 🇺🇸 And honestly… it’s long overdue. For years, European founders had 2 choices: 1. Stay local and deal with fragmentation 2. Move to the US to scale 𝗘𝗨 𝗜𝗻𝗰. is trying to remove that trade-off. If executed well, this could be one of the most important structural changes for European startups in decades. What do you think?
English
566
960
6.7K
884K
Fabio Filasieno ❄️
Fabio Filasieno ❄️@FilasienoF·
@wbmosler It's like a trauma. They are all scared about debt and inflation on all sides! I wonder: 1. If they know that they are effectively suggesting raising taxes and cutting spending. 2. Why there is no such thing for full employment in the USA; like unused workforce hours that could have been bridges, or roads, or parks, or services. I have no explanation.
Thomas Massie@RepThomasMassie

The debt passed $39 trillion today. Paying a trillion dollars of interest annually on this debt causes hardship for tax-payers and robs us of resources that could otherwise be used for infrastructure or national defense. And ultimately, this debt will enslave our grandchildren.

English
1
0
1
81
Fabio Filasieno ❄️
Fabio Filasieno ❄️@FilasienoF·
I agree and the experience is terrible. dspace.mit.edu/handle/1721.1/… this is a better model. Where each agent is a propagator, and feedback loops are naturally supported. Cells allow partial information accumulation and contradictions. Naturally designed for speculative execution. Right now agents are like the first cars: a chariot without horses.
English
0
0
0
69
Oskar Wickström
Oskar Wickström@owickstrom·
Right now agentic coding feels too slow. I get what I want faster typing it myself. "But parallelize!" you exclaim. Parallel work means context switching, and that is the mind killer, the little-death that brings intellectual and computational obliteration.
English
11
1
57
2.3K
Fabio Filasieno ❄️
Fabio Filasieno ❄️@FilasienoF·
È quello che temono. Fai passare un decreto liberticida comprimendo il mio diritto di parola? E io non ti voto più, ma non per questa volta, non ti voto mai più. Mi parli di debito pubblico e inflazione invece di piena occupazione? E io non ti voto. Non voto più te, ma continuo a votare altri, ovvero i tuoi concorrenti. Mi scontenteranno anche questi? Allora non li voterò più e ne voterò altri. Ma quando arrivano le elezioni devono tremare e subirle quando invece di ascoltare i lavoratori ascoltano il capitale. La difficoltà è che i media raccontano il contrario: “non votate” - ovvero non opponetevi - “tanto sono tutti corrotti”. Incidentalmente i servi del capitale ci vanno sempre al voto, ma dicono agli altri di non andarci. Una battaglia impari.
Italiano
0
0
1
23
Sabrina®️🇮🇹
Sabrina®️🇮🇹@SabrySocial·
@FilasienoF @leilamazzocco @ZombieBuster5 Da una parte infiltrano tutti i grossi partiti, dall’altra boicottano le nuove iniziative politiche sul nascere. Si muovono con l’inganno. Ma non possono fermare la scintilla che esplode nella determinazione popolare. Arriverà quel giorno.
Italiano
1
0
1
42
Fabio Filasieno ❄️
Fabio Filasieno ❄️@FilasienoF·
Astenersi: il sogno di ogni élite liberista. Il danno provocato dall’antipolitica è incalcolabile: l’assenza di rappresentanza politica è una politica precisa m, con un none. Decidere di non decidere è una decisione con effetti precisi. L’unica forma politica dove il popolo non vota è la dittatura: il capitale vuole che il lavoro scelga di non occuparsi di economia liberamente. Prima con l’uomo forte, poi con l’ordoliberalismo e oggi con un ingrediente ancora più efficace: l’antipolitica. Con l’antipolitica qualunque soluzione politica è sotto attacco; distrugge qualunque possibile futuro in cui il capitale è subordinato alla politica (perché questo è ciò che teme il capitale: la subordinazione alla politica). L’antipolitica è la vittoria definitiva del capitale e il non votare è la sua massima espressione; è come partecipare a un torneo e decidere di non giocare, ma la sconfitta a tavolino implica la schiavitù a vita per noi e i nostri figli. Oltre al danno, la beffa: i lavoratori lo scelgono liberamente.
Italiano
1
0
1
21
Sunny Bains @TiDB
Sunny Bains @TiDB@sunbains·
The real art of engineering isn’t writing clean code in a vacuum — it’s managing entropy. The best developers aren’t the ones who avoid ugliness entirely (impossible), but the ones who know where to let the mess live and how to keep it contained.
English
3
5
50
4.2K
Fabio Filasieno ❄️
Fabio Filasieno ❄️@FilasienoF·
I think that when you merge narcissistic and psychopathic tendencies, it's impossible to see reality as it actually is. One could think it's incompetence, but an incompetent must actually understand and make the wrong decision. It's mental illness: simply, it is impossible for that individual to understand the pain and suffering he is causing and that he might be wrong. Being worried you might hurt someone and making sure you are not wrong is the minimum a decision maker should consider. “It is easier for a camel to go through the eye of a needle…” for a reason; privilege correlates easily to psychopathic and narcissistic disorder.
English
0
0
1
8
Lewis Campbell
Lewis Campbell@LewisCTech·
@FilasienoF @witte_sergei It's also just flat out incompetence and childishness. The people influencing his decisions knew exactly would happen - Israel would have a free hand ethnically cleansing Lebanon while the US provided the blood and money (always they demand our blood and money) to attack Iran.
English
1
0
0
64
Big Serge ☦️🇺🇸🇷🇺
On the one hand, you’d think that unblocking the strait would be one of the most war-gamed and heavily studied scenarios in the American playbook. On the other hand, the President is tweeting “maybe China will send some ships to help.”
Big Serge ☦️🇺🇸🇷🇺 tweet media
English
130
368
2.9K
68.6K
Lewis Campbell
Lewis Campbell@LewisCTech·
@witte_sergei What makes you think Trump and his inner circle even have the intellectual capacity to even comprehend what their military experts say? I am positive they all knew the risks of this very well and advised against all of it.
English
1
0
3
139
Fabio Filasieno ❄️
Fabio Filasieno ❄️@FilasienoF·
@BenjDicken Excellent stack Maybe replace Tanenbaum with “Three easy pieces …”; or replace Cormen with Silberschatz; but honestly we are talking about details. Excellent stack. Commented just because these books brought up old memories.
English
1
0
2
662
Ben Dicken
Ben Dicken@BenjDicken·
Name a better tech stack
Ben Dicken tweet media
English
95
135
2.1K
84K
Fabio Filasieno ❄️
Fabio Filasieno ❄️@FilasienoF·
@lemire @EricRichards22 When I was a student, it was fun running them… for one moment you felt like the big guys and you started racing both with yourself and others.
English
0
0
2
46
Daniel Lemire
Daniel Lemire@lemire·
@EricRichards22 I once taught (a very long time ago) a data structure cours in C++ using STL. It was fun getting students to run benchmarks!
English
1
0
24
1.4K
Eric Richards
Eric Richards@EricRichards22·
I think that if you're trying to teach computer science, you need to go high, or low, probably both, and not muddle in the middle. I always think about the terrible algorithms class I took that was 100% chalk and blackboard. If you did the same thing, but with implementations in C and Haskell, you'd get 100x more out of it and really filter people who just can't do it
Daniel Lemire@lemire

Which programming language should you choose for teaching? Though many schools use Java, C#, C or C++, many others have adopted Python. The upside of Python is that it is somewhat easier to get going (helps motivation). The downside is that Python makes it harder to think about low-level issues such as data structures since everything is abstracted away. My own view on the matter, is that students should become polyglots. It is a strategic mistake to focus on a single programming language. But what about learning outcomes? Hott tells us that it does not matter. « there was no statistically significant difference in overall outcomes or struggle between students who complete their programming assignments solely in Python, solely in Java, or a combination thereof. Additionally, there was no statistically significant difference in overall scores on programming assignments, written problem sets, or quizzes from the course based on the language students chose when implementing their solutions. From these results, we conclude that providing students with a choice of programming language, including allowing students to program in a language they are more familiar with, does not appear to dramatically improve student outcomes. Additionally, the use of Python over Java (or consequently Java over Python) in an upper-level algorithms course does not improve performance overall, even though it may provide some benefit in isolated assignments. Therefore, educators need not worry about how the programming language chosen for their courses may impact student outcomes. » Hott, J. R. (2025, August). Student Outcomes When Provided Programming Language Choice in an Algorithms Course. In Proceedings of the 2025 ACM Conference on International Computing Education Research V. 2 (pp. 26-26). Hott has interesting research... engineering.virginia.edu/faculty/john-r… Coming back to what kids should learn, I largely agree with @lzsthw and his essay « AI Didn't Kill Programming, You Did ». Instead of worrying about which programming language we should use, we should turn things around and tell kids how to start a business, how to become independent from tech trends, and so forth. The very idea that you should standardize on one programming language should be a red flag. You can learn programming with anything. Start with Logo, Ada... Do it all! Heck!!! Invent your own programming language. learncodethehardway.com/blog/39-ai-did…

English
11
0
50
4.5K
Fabio Filasieno ❄️
Fabio Filasieno ❄️@FilasienoF·
The dual of this is that SaaS products can potentially use custom co-designed data layer-infrastructure-hardware and increase performance massively (think ultra-restricted domain DBs like Tigerbeetle). But your position from one point of view does not change the outcome. Knowing about databases, infra, and hardware is as important as ever.
English
1
0
1
110
Sunny Bains @TiDB
Sunny Bains @TiDB@sunbains·
Al is compressing the application layer, making some SaaS products easier to replace, and shifting value toward outcomes and infrastructure.
English
2
1
13
1.8K
Fabio Filasieno ❄️
Fabio Filasieno ❄️@FilasienoF·
Experienced developers are more important than ever. When writing specifications, focus heavily on data formats; consider memory as a storage medium just like disk. This is nothing new; it’s a key recommendation that Mike Acton suggests. Document your data formats both in memory and on disk! Why focus on data formats? Because that's the key to understanding cache utilization. Effective developers care greatly about how they lay out data—something that JS, Java, etc., developers often don’t even know exists as a problem. On the other hand, game dev folks know this very well. Using AI to define data formats for both memory and disk is a very powerful tool; and experienced developers who know what works and what doesn’t specifically in data format design will provide incredible value.
Vjekoslav Krajačić@vkrajacic

And for a prompt to contain enough information to reasonably tightly constrain the generated output, the person writing it needs to understand the field well enough, both to direct the solution and to verify it afterward. That's why for experienced programmers it can become a boost, while for novices it can create the false impression that they know what they're doing. Often times the exact opposite is what gets marketed, the less you know the better.

English
0
0
4
432
Lewis Campbell
Lewis Campbell@LewisCTech·
I am trying a revolutionary software development technique where I think about things before I code them. I call it Thinking About Things. Please join me in a mountain retreat so we can all sign the Thinking About Things manifesto.
English
33
50
819
14.8K
Fabio Filasieno ❄️
Fabio Filasieno ❄️@FilasienoF·
photonlibos.github.io/blog/stackful-… I think this is valid only if you fix semantics. What happens if you add colored routines? Blue: can call red and blue and can be suspended. Red: can call only red (standard routine). Now whenever you switch, the stack is empty; it is used for unsuspendable operations. Blue routines define context and are generated by the compiler using tail-recursive continuation-passing routines. Other issue: 1M coros… stackful… what is the minimum memory required to run it? Stackless colored seems still better as a design -> irif.fr/~jch/research/…
English
0
0
1
119
Fabio Filasieno ❄️
Fabio Filasieno ❄️@FilasienoF·
Non c’è nulla da correggere va rimandata e poi buttata nel dimenticatoio per sempre. Una compressione dei diritti inaccettabile a favore di Israele nel momento meno opportuno. “L’antisemitismo è una certa percezione degli ebrei che può essere espressa come odio verso gli ebrei. Le manifestazioni retoriche e fisiche dell’antisemitismo sono dirette verso le persone ebree o non ebree e/o verso i loro beni, e verso le **istituzioni** e le strutture religiose delle comunità ebraiche.” La definizione IHRA chiarisce che anche alcune critiche allo Stato di Israele possono essere antisemite, ad esempio: ∙Applicare a Israele doppi standard non richiesti ad altri paesi ∙Paragonare la politica israeliana al nazismo ∙Negare al popolo ebraico il diritto all’autodeterminazione Adottate per intero la definizione dell’ IHRA che include le loro interpretazioni (“international…” e già sappiamo la qualità di ciò che comincia con “World” o “International”). Ma il resto del testo è allucinante: monitoraggi, registri, … ma siete matti? Questa è una merda peggio di quella di Draghi: state sottovalutando. Io voglio poter dire quello che penso senza finire in nessun registro. Leggete signori, leggete lo schifo: senato.it/leg/19/BGT/Sch…
Italiano
1
0
1
20
Fabio Filasieno ❄️
Fabio Filasieno ❄️@FilasienoF·
Exactly. But you focus directly on the spec. You can change ideas much faster: discard everything and rewrite by altering a few sections instead of manually rewriting code. The feedback loop is quicker. It requires more skills, not fewer. That's my viewpoint, and the world is beautiful because there are many.
English
1
0
0
14
Imad
Imad@imadr_·
@FilasienoF So, write a spec that is precise, correct, transform data in an efficient way using hardware and math, iterate to make it better ...or in other words: code Sorry I can't see it in another way
English
1
0
0
13
Fabio Filasieno ❄️
Fabio Filasieno ❄️@FilasienoF·
It mentions only two paths, but there is a third one very few are mentioning. First, the author is very insightful on how developers feel about their work. I understand the struggle, but I’ve never felt more energized, as coding is mostly drudgery. I care about hardware, data formats, and specification clarity. To me, AI is a new world to help explore strategies that previously would have taken an impossible amount of time. If you want to keep the joy of programming—that intimate feeling between craftsman and work of art—focus on specifications. Rule of the game: AI must one-shot it! If it doesn’t, your spec is wrong. So what is our contribution: we transform data in the most efficient and effective way using two key tools: hardware and math. What is in between, if specified correctly, can be done by AI. From my point of view, little has changed; I’ve just removed drudgery. Look 👇
Mo@atmoio

I was a 10x engineer. Now I'm useless.

English
2
3
14
4K