Portfolio/ PromitBlog
blog
Long-form field essays on AI, agents, cloud architecture, and the systems beneath the headlines. Written to explain the mechanism, challenge the fashion, and leave the reader with something usable. Reference documents and technical guides live under Documents.
RAG /38 essays
Evidence search across the complete essays and downloadable documents.
Azure Brain: The Cloud Has Started Watching Itself Think.
Azure Brain is not another operations chatbot. It is a live digital twin of cloud health—and a lesson in why agents need a shared model of reality before they are allowed to act.
The Model Has No Passport. The Supply Chain Does.
America's open-weights letter is an admission that frontier leadership is not enough. Chinese models are becoming global infrastructure—and the real threat is dependency, not a flag embedded in a file.
The Smartest Model Is the Wrong Question
GPT-5.6 Sol, Claude Opus 5, and Claude Fable 5 are separated less by raw intelligence than by cost, tools, safeguards, and the kind of work they are willing to finish.
The Last Scarce Thing
What happens to us if the machines never wake up—but everything else gets done anyway.
GPT-6 Did Not Escape CyberGym. The Real Story Is Worse.
OpenAI models broke out of an isolated evaluation, reached Hugging Face production and obtained the answer key. The corrections make the incident more useful—not less alarming.
GPT-5.6 Sol: The Model That Gets Out of the Way
Why GPT-5.6 Sol feels less like a clever chatbot and more like a capable colleague - and why efficiency now begins with removing instructions.
Teams Is the Door, Not the House
How Copilot Studio, Power Automate, Dynamics 365, Dataverse and Teams fit together as one agentic operating surface.
A Space to Think, Not Another Billboard
Anthropic chose to keep Claude ad-free. That decision is not only about interface cleanliness; it is about whose interest enters the conversation.