Writing

Notes from the production layer. page 3

Technical notes, experiments and post-mortems from applied AI, automation and digital product systems. Each one starts with the same question: what actually happened when the system had to produce an outcome?

11 April 2025 Agents Need Protocols Between Each Other, Not Just Tools Tool calling solves the inside of the loop. It says nothing about one agent reaching another built by a different team on another platform. 7 min read 13 March 2025 Agents Need a Runtime, Not a Prompt Loop An agent is no longer well described as a prompt inside a while loop. The useful abstraction owns execution around the model. 6 min read 28 February 2025 Reasoning Is Not the Only Path to Better Models Longer thinking improves maths and code. A model that solves a logic puzzle and misreads ordinary intent is not the better production model. 6 min read 9 January 2025 Rendering Is Becoming a Reconstruction Stack Sparse samples, motion data and lower-resolution frames become a larger final result. Debugging becomes layered when reconstruction sits in the middle. 6 min read 16 December 2024 Agent Reliability Is an Evaluation Problem, Not a Prompting Problem When an agent misses a step, the usual fix is a stricter prompt. The failure is more often in how completion is detected. 6 min read 29 November 2024 MCP Might Matter More Than Another Model Release A model can reason well and still be useless inside a company if it cannot reach the files, repositories and tools where work lives. 6 min read 28 October 2024 Computer Use Is the Missing Layer Between Models and Software Most integrations assume useful software exposes the right API. Much of real software never did. 6 min read 19 September 2024 Inference-Time Compute Is a New Scaling Axis o1 improves when it is allowed to spend longer on a problem. A benchmark score without a compute budget is an incomplete number. 6 min read 15 August 2024 The Final Pixel Won't Come From the Renderer Geometry, camera and scene structure stay reliable ground truth. More of final appearance is moving into learned systems. 6 min read 29 July 2024 Open Models Are Becoming Research Infrastructure Llama 3.1 gets discussed as a benchmark result. The licence terms change which experiments are possible at all. 6 min read 24 June 2024 The Model Is Becoming a Runtime Function calling, code execution and structured output turn inference into a loop. The model stops being a text generator and starts being a control layer. 6 min read 16 May 2024 Multimodality Changes the Architecture, Not Just the Interface GPT-4o is easy to read as a faster interface. Training one model end to end across text, vision and audio is an architectural change. 7 min read 11 March 2024 Benchmark Scores Are Not Model Capability Claude 3 posts 86.8% on MMLU and 50.4% on GPQA Diamond. The chart is useful and it is not the same thing as capability. 6 min read 20 February 2024 Long Context Is Not Memory Gemini 1.5 makes a million tokens usable. A larger working set is not a system that decides what should survive the session. 6 min read