Lo más parecido al nirvana para mi gusto. VimWiki: Personal Knowledge Management in Vim mkaz.blog/working-with...

Beyond Code: The New Role of the Developer | IBM www.linkedin.com/events/build...

Una generación vivió el paso del taller al terminal, de la mecánica al silicio, del oficio técnico a la cultura digital. En este texto reflexiono sobre esa travesía: la de los Jones digitales. Elogio de la Generación Jones digital cienciasucia.com/post/7966721...

Entrenar, ajustar o construir modelos de lenguaje: guía para perplejos cienciasucia.com/post/7961461...

"This legendary page from an internal IBM training in 1979 could not be more appropriate for our new age of AI." simonwillison.net/2025/Feb/3/a...

Hice el experimento de comparar las cuatro ediciones del libro "Artificial Intelligence. A Modern Approach" (1995, 2003, 2010 y 2020) de Russell y Norvig y he aprendido mucho: cienciasucia.com/post/7951135...

"We [developers] want to take all the precautions we can to avoid future problems, we don’t just take requirements at face value – we question them, sometimes to the frustration of stakeholders." timharek.no/blog/why-dev...

'Most experts do not believe today’s models are capable of consciousness. Why doesn’t that settle the matter? There's a philosophical question: When the simulation is near perfect, does that make it real? You have to engage with that reality.' www.wired.com/story/micros...

Floating-point arithmetic in GPUs exhibits non-associativity, meaning (a+b)+c ≠ a+(b+c) due to finite precision and rounding errors. This property impacts the computation of parallel operations: can yield different results based on execution order. thinkingmachines.ai/blog/defeati...

'We study whether generative artificial intelligence constitutes a form of seniority-biased technological change, disproportionately affecting junior relative to senior workers. We find early evidence of a seniority-biased impact of AI adoption and its mechanisms.' papers.ssrn.com/sol3/papers....

Investigadores del MIT proponen una forma de entender por qué los modelos de lenguaje sobre proteínas (protein language models) hacen ciertas predicciones, lo que aumenta la confianza de los investigadores en entornos de alto riesgo como el descubrimiento de fármacos. www.pnas.org/doi/10.1073/...

Cuando la IA parece consciente (y por qué eso no basta). Una reflexión sobre la era de la SCAI (Seemingly Conscious AI) cienciasucia.com/post/7930468...

Del espejismo determinista al software como naturaleza www.tumblr.com/cienciasucia...

"No se trata de abdicar del pensamiento, sino de elevarlo: dedicar menos tiempo a lo comoditizado y más a lo que no lo está, que es precisamente lo difícil de automatizar." www.enriquedans.com/2025/08/la-e...

UNIX: A History and a Memoir by Brian Kernighan youtu.be/WEb_YL1K1Qg?...

"This is a fantastic and horrifying demonstration of how widespread and dangerous these vulnerabilities still are, almost three years after we first started talking about them." simonwillison.net/2025/Aug/15/...

"I believe that AGI is not possible with the current regime of LLMs. The GPT-style autoregressive language transformer that was published in 2018 by OpenAI as GPT-1 - this style of AI, we shall call them LLMs from now - lacks the capabilities needed for AGI." kennethwolters.com/posts/no-agi/

From GPT-2 to gpt-oss: Analyzing the Architectural Advances magazine.sebastianraschka.com/p/from-gpt-2...

Here’s a hand-picked selection of articles on AI fundamentals/concepts that cover the entire process of building neural nets to training them to evaluating results. aman.ai/primers/ai/

Discover the innovative world of open-source development with Kovid, the mastermind behind the Kitty Terminal and Calibre. Gain insights into terminal protocols, open-source strategies, and the future of software design. m.youtube.com/watch?v=8PYL...

"Lo que se publicita como «AI-powered» es, en realidad, un conjunto de reglas condicionales básicas [...] que hace décadas nadie en su sano juicio se atrevería a calificar como «inteligencia» sin quedar como el más idiota de los ignorantes." www.enriquedans.com/2025/06/llam...

"The second version of the json package coming in Go 1.25 is a big update, and it has a lot of breaking changes. It adds new features, fixes API issues and behavioral flaws, and boosts performance. Let's take a look at what's changed!" JSON evolution in Go: from v1 to v2 antonz.org/go-json-v2/

“No es que los LLM estén reemplazando la programación, es que están convirtiéndose en un nuevo tipo de primitiva de programación”. www.tumblr.com/cienciasucia...

Magistral: How to Run & Fine-tune docs.unsloth.ai/basics/magis...

“Looking at his code was like looking at the ceiling of the Sistine Chapel. His code was remarkable. It is what made the Macintosh possible.” Bill Atkinson, Who Made Computers Easier to Use, Is Dead at 74 www.nytimes.com/2025/06/07/t...

What Is OAuth and How Does It Work? - fusionauth.io/articles/oau...

MCP vs ACP vs A2A: Comparing Agent Protocols with @seldo.com from LlamaIndex youtu.be/kqB_xML1SfA?...

"This paper provides an in-depth technical analysis and implementation methodology of the open-source Agent-to-Agent (A2A) protocol developed by Google and the Model Context Protocol (MCP) introduced by Anthropic." arxiv.org/pdf/2506.01804

This guide is for readers already familiar with OpenAI models and LLM agents, and want to see how to orchestrate a team of agents for a real-world, complex task. cookbook.openai.com/examples/age...

El Subsistema de Windows para Linux ahora es de código abierto hipertextual.com/2025/05/el-s...

Just a few weeks after OpenAI said it would adopt rival Anthropic’s standard for connecting AI models to the systems where data resides, Google is following suit. techcrunch.com/2025/04/09/g...

The path to open-sourcing the DeepSeek inference engine - github.com/deepseek-ai/... "Their approach of working on splitting out maintainable sublibraries and sharing info directly even if not integrated seems a really nice way of working with the community." (cont.)

Behind the 6-digit code: Building HOTP and TOTP from scratch blog.dogac.dev/how-do-one-t...

"Natural language is proving to be a more powerful interface than drag-and-drop WYSIWYG editors. And most importantly, people don’t want to be locked into proprietary runtimes. They want actual code. They want control. They want to scaffold, edit, and deploy anywhere." kenneth.io/post/no-code...

20 years of Git. Still weird, still wonderful. blog.gitbutler.com/20-years-of-...

Docker Desktop 4.40+ introduces `docker model` CLI as the first class-citizen. This means AI models are now treated as fundamental, well-supported objects within the Docker CLI, similar to how Docker already treats containers, images, and volumes. medium.com/@pawel.piwos...

'I noticed that Alto programs tended to treat the keyset as function keys, rather than a mechanism for typing with chords. David Liddle of Xerox PARC said, "We found that [the keyset] was tending to slow people down, once you got away from system programmers."' www.righto.com/2025/03/moth...

El artículo "The Bitter Lesson" de Ankit Maloo explora la idea planteada por Richard Sutton en la que se argumenta que el mayor progreso en inteligencia artificial (IA) ha surgido de aumentar el músculo computacional, no del desarrollo de algoritmos manuales complejos.

IBM’s CEO doesn’t think AI will replace programmers anytime soon techcrunch.com/2025/03/11/i...

Roadmap To Learn Agentic AI.

Random AI Tools. Discover Amazing AI Tools through Serendipitous Discovery. www.randomaitools.com

Martians on @t3chfest.es. The best introduction at the software conference this year. The robot song “you are smarter let's be friends” was amazing.

Aunque la repetición de "if err != nil" en Go puede ser tediosa y genera sobrecarga visual, obliga a los desarrolladores a ser conscientes de los errores en cada paso crítico, lo que puede prevenir errores silenciosos. Why Go's Error Handling is Awesome rauljordan.com/why-go-error...

The purpose of these posts and the resulting software is purely educational - to learn network and system programming at a deeper level. www.saminiir.com/lets-code-tc...

Starting next week, we'll open-source 5 repos – one daily drop – not because we've made grand claims, but simply as developers sharing our small-but-sincere progress. No ivory towers - just pure garage-energy and community-driven innovation DeepSeek Open Infra github.com/deepseek-ai/...

The new Granite 3.0 8B and 2B language models are designed for tasks such as Retrieval Augmented Geneneration (RAG), classification, summarization, entity extraction, and tool use. They're re designed to be fine-tuned with enterprise data. newsroom.ibm.com/2024-10-21-i...

Load Balancing Algorithms: How They Work and When to Use Them www.linkedin.com/posts/ninadu...

Is it really possible to encode arbitrary data in a single emoji? tl;dr: yes, although I found an approach without ZWJ. In fact, you can encode data in any unicode character. paulbutler.org/2025/smuggli...

A Minecraft server written in COBOL github.com/meyfa/CobolC...

Everything we actually know about the Apple Neural Engine (ANE) github.com/hollance/neu...