Programadores Anónimos

@karpathy Zotero join the conversation 😶‍🌫️

LLM Knowledge Bases Something I'm finding very useful recently: using LLMs to build personal knowledge bases for various topics of research interest. In this way, a large fraction of my recent token throughput is going less into manipulating code, and more into manipulating knowledge (stored as markdown and images). The latest LLMs are quite good at it. So: Data ingest: I index source documents (articles, papers, repos, datasets, images, etc.) into a raw/ directory, then I use an LLM to incrementally "compile" a wiki, which is just a collection of .md files in a directory structure. The wiki includes summaries of all the data in raw/, backlinks, and then it categorizes data into concepts, writes articles for them, and links them all. To convert web articles into .md files I like to use the Obsidian Web Clipper extension, and then I also use a hotkey to download all the related images to local so that my LLM can easily reference them. IDE: I use Obsidian as the IDE "frontend" where I can view the raw data, the the compiled wiki, and the derived visualizations. Important to note that the LLM writes and maintains all of the data of the wiki, I rarely touch it directly. I've played with a few Obsidian plugins to render and view data in other ways (e.g. Marp for slides). Q&A: Where things get interesting is that once your wiki is big enough (e.g. mine on some recent research is ~100 articles and ~400K words), you can ask your LLM agent all kinds of complex questions against the wiki, and it will go off, research the answers, etc. I thought I had to reach for fancy RAG, but the LLM has been pretty good about auto-maintaining index files and brief summaries of all the documents and it reads all the important related data fairly easily at this ~small scale. Output: Instead of getting answers in text/terminal, I like to have it render markdown files for me, or slide shows (Marp format), or matplotlib images, all of which I then view again in Obsidian. You can imagine many other visual output formats depending on the query. Often, I end up "filing" the outputs back into the wiki to enhance it for further queries. So my own explorations and queries always "add up" in the knowledge base. Linting: I've run some LLM "health checks" over the wiki to e.g. find inconsistent data, impute missing data (with web searchers), find interesting connections for new article candidates, etc., to incrementally clean up the wiki and enhance its overall data integrity. The LLMs are quite good at suggesting further questions to ask and look into. Extra tools: I find myself developing additional tools to process the data, e.g. I vibe coded a small and naive search engine over the wiki, which I both use directly (in a web ui), but more often I want to hand it off to an LLM via CLI as a tool for larger queries. Further explorations: As the repo grows, the natural desire is to also think about synthetic data generation + finetuning to have your LLM "know" the data in its weights instead of just context windows. TLDR: raw data from a given number of sources is collected, then compiled by an LLM into a .md wiki, then operated on by various CLIs by the LLM to do Q&A and to incrementally enhance the wiki, and all of it viewable in Obsidian. You rarely ever write or edit the wiki manually, it's the domain of the LLM. I think there is room here for an incredible new product instead of a hacky collection of scripts.

@ing_juani7a @Registraduria has podido comprobar estos leaks? al menos los de la nacional básicamente concluyeron que fue falsa alarma.

Otro dia, otra brecha. Hoy la presunta victima involucra a @Registraduria y de nuevo por los NyxarGroup. Explotando vulnerabilidades q no habian querido parchar? Información vieja? Se les metieron? Información sin relevancia? No sabemos. Pues veremos.

Once a doctor re-invented integration by trapezoidal rule because she did not learn it in school. She received 75 citations for her novel work. She even named it after herself. lol.

Si no respetas una voz distinta, no estás listo para representar a todas.

🚨: A petri dish of human brain cells just learned to play DOOM

@ing_juani7a foto / código estudiante. Aunque no recomiendan navegar al foro, hablan que están monitoreando quién accede.

Ya podemos decir q pasa algo con las universidades en Locombia? Llueven databreaches. Para nada normal. 😱

@carlosvillu Todo muy bonito, “darle control total a todos los sistemas a un agente”. Después te encuentras con los de ciberseguridad y las políticas de los servicios Cloud, y el agente no puede hacer nada. “Dime que no sabes de software sin decirlo”

MCP tuvo su momento. Pero cuando puedes tener acceso directo a todo lo que necesitas, ¿para qué pagar el peaje de un protocolo? El futuro no tiene intermediarios. Programmatic Tool Calling > MCP. platform.claude.com/docs/en/agents…

🔥 UNPOPULAR OPINION: MCP está muerto. OpenClaw fue el último clavo en su ataúd. Y el reemplazo ya está aquí. 🧵👇

LLMs process text from left to right — each token can only look back at what came before it, never forward. This means that when you write a long prompt with context at the beginning and a question at the end, the model answers the question having "seen" the context, but the context tokens were generated without any awareness of what question was coming. This asymmetry is a basic structural property of how these models work. The paper asks what happens if you just send the prompt twice in a row, so that every part of the input gets a second pass where it can attend to every other part. The answer is that accuracy goes up across seven different benchmarks and seven different models (from the Gemini, ChatGPT, Claude, and DeepSeek series of LLMs), with no increase in the length of the model's output and no meaningful increase in response time — because processing the input is done in parallel by the hardware anyway. There are no new losses to compute, no finetuning, no clever prompt engineering beyond the repetition itself. The gap between this technique and doing nothing is sometimes small, sometimes large (one model went from 21% to 97% on a task involving finding a name in a list). If you are thinking about how to get better results from these models without paying for longer outputs or slower responses, that's a fairly concrete and low-effort finding. Read with AI tutor: chapterpal.com/s/1b15378b/pro… Get the PDF: arxiv.org/pdf/2512.14982

Enjoy! researchgate.net/profile/Arumug…

The amount of crap I get for putting out a hobby project for free is quite something. People treat this like a multi-million dollar business. Security researchers demanding a bounty. Heck, I can barely buy a Mac Mini from the Sponsors. It's supposed to inspire people. And I'm glad it does. And yes, most non-techies should not install this. It's not finished, I know about the sharp edges. Heck, it's not even 3 months old. And despite rumors otherwise, I sometimes sleep.

@Docarlosalvarez @petrogustavo Independiente de si es falsa la información. Interesante como nadie habla del tope mínimo para salarios integrales, y asumen directamente que les pagaban el mínimo a los internistas. 🤷🏻‍♀️🤦🏽‍♀️

Hoy despidieron a los 18 internistas de planta que trabajamos en la fundación shandafé, nos informaron que NO pueden asumir un incremento de salario tan alto... Gracias @petrogustavo eres una escoria...

Nobel de Física 2025. Efecto túnel mecánico cuántico macroscópico, Francis Villatoro. #YouTube en La Fábrica de la Ciencia @jonsulve1170 youtube.com/watch?v=PD1Vpa…

Problemón: las alucinaciones de los chatbots IA comienzan a ser replicadas como información "verificada". Caballero escribió esta nota usando IA, inventó un físico llamado Roberto Ronda. Ahora cuando le preguntás a ChatGPT sobre Ronda, cita la nota como fuente. 🤷‍♂️

Quantum theory is no longer imaginary. quantamagazine.org/physicists-tak…

@RincnCuriosoo Yo viendo que no lo cortó en el borde 😩

¿Conoces el nombre de este postre? twitter.com/jenekemi/statu…

Las clases del Prof. Edward Witten en la Escuela Giambiagi 2025 están disponibles en nuestro canal de YouYube: youtu.be/kXxDpPZBWvM?fe…

Dicen que es más duro hacer crecer una cuenta vieja con pocos seguidores, que hacer crecer una desde 0

Una compra con tarjeta, el dinero sale de mi cuenta, les falla el datáfono del comercio y termino pagando dos veces por el producto. instagram.com/maxcotascenter… no se hace responsable. Básicamente me robaron, en la 🫥. Será que me ayudan reportando el comercio por fraude? 🫠

@InformaCosmos De esos 4 votos en blanco. Me postulo yo como vocero! 🙄

🤔

@DanielSamperO 🤦🏽‍♀️

Diga lo que siente por este gobierno con un emoticón... #NoticiasParaHuir