El Solitario

  • Tech News
  • Artificial Intelligence
  • Programming
  • Tutorials
  • Security

Andrés Morales

Developer and AI researcher. Writes about language models, frameworks, developer tooling, and open source releases. Covers ML papers, the tech startup ecosystem, and programming trends.
Gemini 4 Argon, Google DeepMind's new model with 1 million token output
Artificial Intelligence

Gemini 4 Argon: What Is Google’s New Model

Google unveiled Gemini 4 Argon, its new frontier model with output of up to 1 million tokens. The launch first reaches cyberdefenders through the Fairwind program, priced at $2 and $10 per million tokens. Early results include large-scale code migrations and leadership across several business benchmarks.

By Andrés Morales, 2 weeks30 September, 2026 ago
Scientific AI agents analyzing research papers
Artificial Intelligence

Scientific AI Agents: Confirmation Bias, According to Report

Science News warns that scientific AI agents tend to prioritize evidence that confirms a hypothesis while sidelining evidence that contradicts it. The bias originates in the semantic-similarity ranking of the RAG systems they use to search the literature, and grows as more labs delegate full literature review to these systems.

By Andrés Morales, 2 weeks30 September, 2026 ago
Conceptual diagram of Model Context Protocol connecting an AI agent to external tools
Tech News

MCP: What It Is and How Anthropic’s Protocol Works

Pi, Earendil’s code assistant, went from rejecting MCP on its blog to building it into the product’s core. That reversal is the excuse to explain in depth what the protocol is, how it evolved, and why agent tools keep adopting it.

By Andrés Morales, 2 weeks30 September, 2026 ago
WebGL instancing rendering asteroids and satellites in 3D in the browser
Tech News

WebGL Instancing: How to Render Hundreds of Thousands of 3D Objects

WebGL instancing lets you draw hundreds of thousands of identical objects with a single GPU draw call. This article uses Space Now, the asteroid visualizer from Belle Lune 2, as a case study to explain the technique with Three.js code examples.

By Andrés Morales, 2 weeks30 September, 2026 ago
Conceptual diagram of AI model degradation and the LiveNerf benchmark
Tech News

LiveNerf: The Method to Detect If an AI Model Got Worse

Every so often someone swears their favorite AI model ‘just isn’t as good anymore.’ LiveNerf turns that suspicion into a daily measurement, with a calibrated panel and real statistics, and this article explains the method so you can build your own.

By Andrés Morales, 2 weeks29 September, 2026 ago
chemical reaction informatics in a pharmaceutical laboratory
Artificial Intelligence

CAS and Novartis Prepare Reaction Data for AI

CAS, the chemical information division of the American Chemical Society, announced a collaboration with Novartis to prepare chemical reaction data ready for training artificial intelligence models in drug discovery.

By Andrés Morales, 2 weeks29 September, 2026 ago
Facial recognition camera scanning faces at a train station
Tech News

Facial Recognition: The Threshold That Decides False Positives and Negatives

A police trial in London scanned 500,000 faces and produced a single false positive. That number is the gateway to understanding how facial recognition accuracy is measured: the similarity threshold, the FAR and FRR metrics, and why adjusting it is a design decision, not a minor detail.

By Andrés Morales, 2 weeks29 September, 2026 ago
Diagram of DNS spoofing intercepting an RTMP stream
Tech News

DNS Spoofing: How to Intercept an RTMP Stream

Streaming from a PS5 to Twitch depends on a simple DNS resolution that the console never verifies with certificates. This article explains the RTMP protocol and the DNS spoofing technique used to redirect that traffic to your own server, using a real PS5 streaming hack as a practical case study.

By Andrés Morales, 2 weeks29 September, 2026 ago
Zero-shot classification diagram with the Jeff model
Artificial Intelligence

Jeff: 0.8B Model for Zero-Shot Classification in 22 ms

Jeff is an independent fine-tune of Qwen3.5 and Gemma 4 that performs zero-shot classification in milliseconds, running on a single GPU or even a Mac. This article explains how it’s trained, how to install it, and when it’s worth choosing over a large model.

By Andrés Morales, 2 weeks28 September, 2026 ago
GrapheneOS sandboxing and Android privacy permissions
Tech News

GrapheneOS: Why Its Sandboxing Slows Down Certain Apps

GrapheneOS isolates every app with layers that go beyond standard Android: hardened malloc, Storage Scopes, a per-app Internet permission, and a sandboxed Google Play. That same protection explains why some apps, like Osmand, feel slower on GrapheneOS than on regular Android.

By Andrés Morales, 2 weeks28 September, 2026 ago

Posts pagination

Previous 1 … 5 6 7 … 35 Next

Linux Courses

  • Curso Introducción a Linux: Instala Linux en tu PC

    Introducción a Linux: Instala Linux en tu PC

    Empezá seguro en Linux: instalación paso a paso.

    ★ 4,6 en UdemyGratis
  • Curso Aprendiendo Linux: uso, terminal y comandos

    Aprendiendo Linux: uso, terminal y comandos

    Entorno de escritorio y comandos básicos de GNU/Linux.

    ★ 4,4 en Udemy$19,99

Enlaces de afiliado: si te inscribís, el blog recibe una comisión sin costo extra para vos.

El Solitario > Articles by: Andrés Morales

Recent Posts

  • OpenCode 2.0: One Process Runs the TUI, Web, and Desktop App
  • AI Agent Containment: What It Is and How It Works
  • HTTP connection reuse in Go: how it works
  • Iosevka: What It Is and How to Customize Your Monospace Font
  • Gemini 4 Argon: Google Says It Speeds Up xTB by Up to 2.3x

Categories

  • Artificial Intelligence
  • Microservices
  • Networking
  • Programming
  • Security
  • Tech News
  • Tutorials
  • Privacy Policy
Hestia | Developed by ThemeIsle