The AI Concepts Every Engineer Should Know
A living guide to the core ideas of AI engineering: how models think, how we talk to them, how agents act, and how we keep it all safe.
The Blog
What I think about AI and the infrastructure underneath it, plus numbered lessons that break down new ideas in AI engineering, step by step.
The Docker Moment for Data Centers: NVIDIA DSX POD vs Docker & Kubernetes Pods
Docker made software deployment predictable. NVIDIA’s DSX POD tries to do the same for the AI data center itself. Where the comparison holds, why it matters, and where it breaks.
Read the articleMore posts
-
-
Why AI Security Matters: Big Banks Just Locked Out AI Agents
In one week, five of America’s biggest banks stopped AI agents from moving customer money. What happened, why it matters, and how to use AI agents safely.
-
Beyond Chatbots: Inside NVIDIA’s GTC Berlin Keynote and the Blueprint for the Agentic AI Revolution
Six takeaways from Jensen Huang’s keynote, from AI factories that manufacture tokens to agent security enforced in hardware.
-
Build an MCP Server: Tools, Resources, and Prompts
A hands-on walkthrough of a small Model Context Protocol server: when to use each of MCP’s three building blocks, and how clients connect over STDIO and HTTP.
-
Unlocking System 1 AI: Jev
How Jev replaces slow text generation with fast, calibrated decisions for high-volume application workflows.
-
Will AI Kill Us? Or Are We Really Afraid of Losing Control?
What a missing watermark can teach us about AI alignment, reliable constraints, and the growing debate over control.
-
Vibe Coding vs. Spec-Driven Development: From Chaos to Clarity
Why production AI engineering needs durable specifications, a Project Constitution, disciplined validation, and agent-independent workflows.
No posts here yet.