Camilo

software engineering, music, cryptography, infosec, languages, psychology. ex @amazon. EN/ES/PT/DE

[Camilo @camilojd]

@vikingog @hiperclonaso1 @FlorG_Py no solo eso, usan la misma contraseña para todo

[Casey Muratori @cmuratori]

[1/2] I don't normally make commentary videos, but after seeing the entirety of Eric Schmidt's University of Arizona commencement speech, I felt like there was a lot more going on than just "CEO mentions AI, gets booed". So I made a video to explain what upset me about it.

[Camilo @camilojd]

@keremk @edzitron sounds too much like cryptocurrency propaganda

[Camilo @camilojd]

@LRH_Superfan pure BS, only the software development jobs that pay the least are the ones that require German, technically competent people don't struggle to get a job with English

[Gary Bernhardt @garybernhardt]

It simultaneously feels like 1) I can use agents to build anything quickly, but also 2) they have no sense of aesthetics when writing code, so everything comes out a mess, never unified by any vision. It's such a bizarre combination.

[Mario Zechner @badlogicgames]

listen to be people who say "i don't know, but i'm excited to find out". seek them out. ignore the rest.

[Camilo @camilojd]

the way to steer properly a glorified autocomplete is with more determinism, not less

[Kent Beck 🌻 @KentBeck]

Dario demonstrating that he doesn't understand software engineering. The human side of what we do has always been the heart. I can empathize with *wanting* to never talk to an engineer ever again, but engineering becomes more important with better tools.

[Славян @SlavAesthetics]

You will never be her

[François Chollet @fchollet]

One of the most jarring things about current AI is its lack of introspection ability and metacognition. It doesn't know what it doesn't know, how it knows, or how it could find out. It's a one-way system.

[Armin Ronacher ⇌ @mitsuhiko]

One of the many reasons we need real competition. Two large labs is not enough and we need to ensure we have competitive open models.

[Ali Spittel @ASpittel]

cron jobs are going through the most epic rebrand

[Camilo @camilojd]

@dvassallo I don't applaud what he did, but good CPAs are worth their time and expertise. Delegating something so delicate to a dumb LLM loop may cost you dearly

[Patrick Senti @productaizery]

We software / IT people have a lot to learn from the airline industry.

Lesson 1: Male-bro-blame culture won't make your field safer. It will just lead to more cover ups.

Lesson 2: Safety first means transparency trumps blame. Focus on outcomes and root causes, not people. Use checklists (& automated checks).

Lesson 3: Responsibility and accountability in a complex system is not easy to attribute. Let the process work it out, not the team. Every insight counts, not who reported it.

[Andrej Karpathy @karpathy]

I'm not very happy with the code quality and I think agents bloat abstractions, have poor code aesthetics, are very prone to copy pasting code blocks and it's a mess, but at this point I stopped fighting it too hard and just moved on. The agents do not listen to my instructions in the AGENTS.md files. E.g. just as one example, no matter how many times I say something like:

"Every line of code should do exactly one thing and use intermediate variables as a form of documentation"

They will still "multitask" and create complex constructs where one line of code calls 2 functions and then indexes an array with the result. I think in principle I could use hooks or slash commands to clean this up but at some point just a shrug is easier.

Yes I think LLM as a judge for soft rewards is in principle and long term slightly problematic (due to goodharting concerns), but in practice and for now I don't think we've picked the low hanging fruit yet here.

[Diario de un picateclas @devruso]

por si os encontráis con el término "deuda cognitiva" y os da pereza leer sobre ello:

es cuando has vibe-codeado el proyecto entero sin siquiera entender las decisiones estructurales que gepeto ha metido.

"¿es malo?"
si.

"¿por que?"
porque cuando hay un error (y gepeto no sabe solucionarlo), no sabes ni por donde te está soplando el viento.

"¿es diferente a la deuda técnica?"
sip. la deuda técnica es "se que este código o esta decisión que estoy tomando es mala y lo tengo en cuenta para el futuro". La "deuda cognitiva" es "no se lo que estoy haciendo, no tengo ni idea del estado del proyecto, yo solo hablo con gepeto y aparecen cosas en la pantalla, le doy que 'si' a todo y rezo para que funcione".

"¿pero esto no sería un problema si gepeto pudiese resolver el 100% de los problemas, no?"
correcto. de la misma manera que no tendríamos colas por las mañanas en el bar para pedir un café si existiesen las vacas esféricas que diesen café en vez de leche.

"sigo sin entenderlo"
tranquilo, ya lo entenderas ;)

[Cindy Sridharan @copyconstruct]

You know what’s more sobering?

Amazon *already* has some of the most stringent/sophisticated CI/CD game in the business. See https://t.co/D7Upm1VTZD

Most embracing “vibe coding” don’t have the guardrails Amazon already does and are going to relearn reliability first principles.

[Camilo @camilojd]

@noampomsky people drink and smoke way less

[Camilo @camilojd]

@DamianCatanzaro especulación, ninguna es profitable

[brecht apologist @madison_tayt]

i’m tired of the learned helplessness in general!
no one can do anything anymore. you can’t cook, can’t read, can’t go out, can’t make friends… stop making excuses and start trying!
“but capitalism!” ok sure but fight back. reclaim even the tiniest portion of your life, my god!