Comparto mi viaje en la era de la superinteligencia. Tips y recursos para devs de la nueva realidad. ◎ Founded Macch: AI for ISPs | Hacking the way.
Argentina
Joined December 2025
- Tweets479
- Following202
- Followers86
- Likes1.7K
Hey @karrisaarinen, it'd be very cool to have Linear Diffs Guide experience, but for a set of commits.
E.g., I sometimes stop reviewing, my agent fix some comments, and I want to see exactly what changed, but with the luxury experience of Guide, not rawdogging the diff.
21
Facundo García Martoni retweeted
We will all be learning in coming weeks & months, that most human-made software:
▫️ Was really slow
▫️ Had many security vulnerabilities
▫️ Was under-specified and incorrect
AI will rewrite it all. Many times. It will make it fast, cost-efficient, and secure. It will think about failure modes we couldn't wrap our heads around.
It will prove, formally verify, load test, harden. It will turn memory-unsafe foundations into robust, optimized ones. If you ask, it'll thoroughly document it, teach you about it, make diagrams, animations, and slide decks.
In many ways, we are witnessing the true birth of software engineering.
Facundo García Martoni retweeted
If you understand code, you can spot-check most languages — even if you don’t know them. Mid and high level structure are pretty obvious in virtually all languages. Names are self-evident.
The captain knows how the ship works at a high level. He does not know the torque rating of every bolt.
Facundo García Martoni retweeted
There’ll be times where you over-guide the AI. It knows better than you and is more intelligent than you.
There’ll be times where you under-guide the AI. It can go off-track, have the wrong biases, or miss context.
How well you manage this superpower is the #0 skill I look for.
Facundo García Martoni retweeted
As you invite more agents to our work, and create software factories to be more productive than ever, you’ll soon outsource everything to an agent, even the thinking.
The work you will do will feel frictionless, it will feel amazing on the surface, but once you go a bit deeper, it’ll feel extremely shallow. It won’t reflect the struggles you went through, your decisions, and ideas.
It will no longer remind you of you and the work you aspire to.
You will turn into just another extremely efficient producer of things, but not a creator like you were before.
Invite friction to your work.
Si no fuese por @mattpocockuk y su visión del agentic engineering, todavía estaría pasando días enteros peleándome con Codex o Claude para integrar un sistema ajeno al mío, con todas sus intricacies y detalles. Hoy, luego de varios /retro, el proceso es una skill 💙
1
1
156
Creo que Orca puede ser, si sigue haciendo las cosas bien, el nuevo Visual Studio Code. Ya estaba cansado de levantar agentes en la terminal, en Codex, en Cursor, etc. Todos peleando por la capa de aplicación en lugar de unificar. Qué lindo volver al IDE + Browser, y nada más ❤️
1
1
149
Hey @Answeror @claireszhou, could we be able to get the Muse Code usage stats via API or something? I'd like to integrate it with projects like ai-usagebar.
1
163
Recién estuve reflexionando sobre el tema del code review (que a fecha de hoy, para mí es sagrado) en relación al agentic engineering y al imparable avance en inteligencia de los modelos frontier.
A día de hoy, cuando yo hago code review, me enfoco casi exclusivamente en las reglas de negocio de lo que estoy haciendo (algo que skills como grill-with-docs de Matt Pocock logran muy bien de antemano).
Al principio leía con enfoque en la implementación, pero poco a poco esta implementación se va haciendo más y más perfecta, e incluso cuando no se hace "a mi gusto" porque el agente o grupo de agentes decidió usar capas de abstracción distintas, es iguál de fácil y rápido corregirlo y seguir adelante, o cambiar la arquitectura en pocas horas.
Creo que en el futuro cercano, con un AGENTS.md lo suficientemente preciso y con las preguntas correctas, podremos evitar el code review en el 99% de los casos. El valor ya no estará en nuestro entendimiento del código (los modelos nos superan ampliamente), sino en nuestra capacidad de crear y entender sistemas.
A final de cuentas, movernos un nivel de abstracción más arriba. Lo que muchos llaman "Strategic Programming". La pregunta está en cuántos niveles de abstracción podremos subir sin perder Pareto.
2
134
Facundo García Martoni retweeted
People will spend $50 to get a mediocre lukewarm burrito delivered to their door, but won’t pay for the maxed out version of AI tools that give them access to infinite intelligence.
This is your competition.
A esto no lo vas a tener nunca con Anthropic, OpenAI ni Cursor. Un equipo humilde que te resuelve el problema y se preocupa hasta que esté solucionado.
1 millón de puntos para el Muse Team.
1
3
324
Facundo García Martoni retweeted
Just realized my Supergrok and Cursor subscriptions are expiring tomorrow.
As much as I enjoy Grok @bot, the whole @spacexai stack is just too much of a mess right now to commit to spending more money on:
- Grok 4.7 late and Elon has lowered expectations
- No unified desktop app
- Grok @bot usage limits comically low
- Individual subscriptions across all services
I'll just continue daily driving my ChatGPT Pro 20x (humble brag) until the fine folks working on Grok improve the whole experience.
@Answeror @claireszhou Hey, can you help me? I was suddenly kicked out of Muse Code, and I didn't do anything wrong!
1
1
234
Oh, I think it's actually a bug @Answeror @claireszhou, apparently the error is misleading: I exhausted the usage.
2
143
Cansado de tener que ir la web de Depot cada vez que quería ver como iba el CI de mis proyectos, creé DepotBar: una app para macOS que permite ver rápidamente en el menu bar el estado de los últimos 5 workflows, usando Depot CLI under the hood: github.com/facmartoni/DepotB…
