A thorough explainer on how quantization makes LLMs 4x smaller and 2x faster while losing only 5-10% accuracy. Covers floating point precision, compression techniques, and how to measure quality loss, with interactive examples throughout. Read more
We’re excited to announce that SymfonyLive Berlin 2026 will take place April 23–24, 2026 at CineStar CUBIX Alexanderplatz, right in the heart of Berlin, directly on Alexanderplatz and easily accessible by public transport. Before the conference,…
There’s so much going on in the AI space, and how to work with AI agents is changing every day. Everyone is overwhelmed and almost numb from so many possibilities, yet you need to find a way to work with AI, not to get left behind, right? You might use AI agents all day long, parallelizing them with AI orchestrators like Agent Teams, Gastown, tmux, git worktree, and AI-based IDEs, but in the end, you just coordinated an AI. You still have to learn what it created, understand it, check for…
We have been tracking North Korea’s Contagious Interview operation since 2024 and maintain a dedicated campaign page that now tracks more than 1,700 malicious packages linked to the activity. In this newly identified cluster, the threat actors operated under GitHub aliases including golangorg and published malicious packages across five open source ecosystems: npm: dev-log-core, logger-base, logkitx PyPI: logutilkit, apachelicense, fluxhttp, and license-utils-kit Go Modules:…
GLM-5.1: Towards Long-Horizon Tasks Chinese AI lab Z.ai's latest model is a giant 754B parameter 1.51TB (on Hugging Face) MIT-licensed monster - the same size as their previous GLM-5 release, and sharing the same paper. It's available via OpenRouter so I asked it to draw me a pelican: llm install llm-openrouter llm -m openrouter/z-ai/glm-5.1 'Generate an SVG of a pelican on a bicycle' And something new happened... unprompted, the model decided to give me an HTML page that included both the SVG…
Anthropic didn't release their latest model, Claude Mythos (system card PDF), today. They have instead made it available to a very restricted set of preview partners under their newly announced Project Glasswing. The model is a general purpose model, similar to Claude Opus 4.6, but Anthropic claim that its cyber-security research abilities are strong enough that they need to give the software industry as a whole time to prepare. Mythos Preview has already found thousands of high-severity…
Microsoft has published its Agent Governance Toolkit, an open source project that brings runtime policy enforcement to autonomous AI agents. The release lands as the industry grapples with a widening gap between how fast AI agents are being deployed and how little infrastructure exists to govern what they do once they're running. The toolkit is available under the MIT license at the Microsoft GitHub organization and supports Python, TypeScript, Rust, Go, and .NET. Agent Governance Is Getting…
Before the main conference kicks off, SymfonyLive Berlin 2026 offers two full days of hands-on workshops to help you sharpen your skills and dive deeper into the Symfony ecosystem. 📅 April 21–22, 2026 📍 Berlin, Germany All workshops will be conducted…
Research: SQLite WAL Mode Across Docker Containers Sharing a Volume Inspired by this conversation on Hacker News about whether two SQLite processes in separate Docker containers that share the same volume might run into problems due to WAL shared memory. The answer is that everything works fine - Docker containers on the same host and filesystem share the same shared memory in a way that allows WAL to collaborate as it should. Tags: docker, sqlite
Steve Schoger shows how he uses Claude Code to design and build UIs, turning natural language prompts into polished interfaces.
Imagine going from a 100-million-row dataset to an interactive analytics app with just a few prompts. What used to take hours or days can now be done in minutes by combining local-first databases and BI tools with an agentic coding workflow. When Rill bet on YAML dashboards and CLI-first workflows in 2022, they weren’t thinking about AI agents. Neither was MotherDuck when they built serverless DuckDB around the thesis that most data fits on a laptop. But it turns out, what is developer-friendly…
Read the full post on https://stitcher.io/blog/more-depedency-considerations
My thoughts on why Agentic Engineering is a better path than Vibe Coding, and the workflow I use to turn AI agents into a structured engineering process. Read more
Google AI Edge Gallery Terrible name, really great app: this is Google's official app for running their Gemma 4 models (the E2B and E4B sizes, plus some members of the Gemma 3 family) directly on your iPhone. It works really well. The E2B model is a 2.54GB download and is both fast and genuinely useful. The app also provides "ask questions about images" and audio transcription (up to 30s) with the two small Gemma 4 models, and has an interesting "skills" demo which demonstrates tool calling…
Release: datasette-ports 0.2 No longer requires Datasette - running uvx datasette-ports now works as well. Installing it as a Datasette plugin continues to provide the datasette ports command. Tags: datasette