Beyond Vibe Coding: Building and Maintaining an Expense Tracker with GitHub Spec Kit
A practical walkthrough of specifications, clarification, quality gates, implementation, convergence, and structured bug fixing.
Night thoughts
Latest articles
Teaching a Small Language Model to Triage Incidents: A Beginner’s Unsloth and QLoRA Tutorial
Beyond DeepSeek and Qwen: Five Teams Building China’s Open-Source AI Foundations
Teaching a Small Local Model to Think in Differentials
Technical Breakdown of Ornith-1.5-122B: Impressive Data, MoE Architecture, and Licensing Caveats
Onyx: The Open-Source AI Assistant That Keeps Company Data at Home
Most companies have information scattered across Google Drive, Slack, Jira, Confluence, and dozens of other tools. When an employee needs an answer, they spend hours searching. AI can help, but sendin
The Harness Is the Agent: Inside Google’s Harness²
Most conversations about AI capability still start with the model. Bigger weights, more compute, longer context, better benchmarks. The model is the thing; everything else is scaffolding. Google Resea
Ask Your Software How It Works: A Practical Guide to REA
How AI agents are turning reverse engineering from a weeks-long chore into evidence-backed answers about the software you already own
The Memory Wall: How NVIDIA’s RTX Spark Changes the AI Bottleneck
Introduction When people talk about AI hardware, the conversation usually centers on compute. How many teraflops can a chip push? How many operations per second? Faster chips sound like faster AI. But
The Future Is Both
New to the jargon? Start with the glossary → Most AI systems today pick a side. They run entirely local or entirely cloud. That's a design choice, not a law of nature. The choice made sense when loca
The New Shape of AI Models
The New Shape of AI Models New to the jargon? Start with the glossary → Here's a riddle. Two AI models are released on the same day. Both have 70 billion parameters. One runs comfortably on a laptop.