Archives

2026

• Oct 5 Vibe Code in a Secured Way
• Oct 5 Rio 0.7.1 released
• Oct 1 Agent Codemode Explained
• Sep 29 Where Does a Claude Agent Run?
• Sep 26 The 100 Ways to Extend AI Agents (Well, More Than Zero)
• Sep 24 Rio 0.5.0: One-Shot Agents Without Mid-Turn Steering
• Sep 21 Jev-Like Models
• Sep 20 Open Takes on Jev: SemIf and Laya
• Sep 20 Debugging an LLM Training Run
• Sep 20 C2C Links Models Through Their KV-Caches
• Sep 19 Testing PrismML Bonsai 2
• Sep 19 Reading GuppyLM
• Sep 18 Testing Jev: Three Playground Runs
• Sep 16 Jev: State In, Typed Decisions Out
• Sep 14 How Do You Prove a Coding Skill Works? A Ponytail Case Study
• Sep 13 AI Foundations 7 - Agents and Autonomous Loops
• Sep 13 AI Foundations 6 - Prompting and Evaluation
• Sep 13 AI Foundations 5 - Tool Calling
• Sep 13 AI Foundations 4 - Context
• Sep 13 AI Foundations 3 - Tokens and Pricing
• Sep 13 AI Foundations 2 - Hallucinations and Limitations
• Sep 13 AI Foundations 1 - How AI Models Work
• Sep 11 Don't Trust the LLM Provider
• Sep 11 Can a Fruit Fly Brain Learn Numbers?
• Sep 9 Managing an Agent's Context Window
• Sep 9 A Fruit Fly Circuit as a Speech Emotion Reservoir
• Sep 7 LoopX: A Control Plane for Long-Horizon Agents
• Sep 5 SKILL.state: State Instead of History
• Sep 4 Early Reactions to GPT-6 Astra
• Sep 4 Coding Will Become Common, Engineering Will Stay Scarce
• Sep 3 Pretraining
• Sep 3 Mostik Links Models in Latent Space
• Sep 3 Atlas: World Labs' Omni World Model
• Sep 2 RMSNorm
• Sep 1 Train A BPE Tokenizer
• Sep 1 Agent Concepts vs Linux Concepts
• Aug 30 The Environment Is the Context
• Aug 30 Good Culture is the Biggest Productivity Hack, Not AI
• Aug 29 micromlp: A From-Scratch Neural Net That Predicts Housing Prices
• Aug 28 Vanilla Neural Networks
• Aug 28 Partial Derivatives, Reverse-Mode Autodiff
• Aug 28 Give AI Agents Git Access, Not Git Credentials
• Aug 27 Where MCP Is Headed Next
• Aug 27 Agent Skills
• Aug 26 What Is an Agent Harness?
• Aug 26 Agents Should Be Durable, Not Long-Lived
• Aug 24 Training Qwen to Paint Watercolors with Pure RL
• Aug 24 Marin 535B-A23B Starts Training, in the Open
• Aug 24 Four Ways to Teach an AI to Draw a Cat
• Aug 23 What Is Agentic UI?
• Aug 23 Neural Network, Explained in Plain Language
• Aug 23 Linus Torvalds on Debugging the Kernel with AI
• Aug 23 Evals, Explained in Plain Language
• Aug 22 Next Token Prediction, Explained in Plain Language
• Aug 22 ELI5
• Aug 22 AI Engineering SKill Map
• Aug 21 Pretraining a Mini Kimi K3 for $252