$Where Michael Thinks

blog.michaelzuo.vip

Michael thinks in public.

Notes on AI systems, software engineering, and the operating taste required to ship real work with agents.

$ load latest-thinking

ok 13 posts indexed

ok 8 topic filters online

new Feedback loop is all you need

May 8, 2026 · 6 min read

Feedback loop is all you need

Agentic coding works when the loop has ground truth. Prompts help, but verification is what makes the system improve.

ai-engineering
Apr 25, 2026 · 5 min read

The Success You Cannot Reproduce Is an Illusion

Correlation explains why AI coding can feel magical and unreliable: the system, not the model alone, turns probability into control.

cognition
Apr 22, 2026 · 6 min read

Don't Build Around Today's Best Model

Capability rankings keep changing. If switching the leading model is hard, the bug is your architecture, not your prompt.

ai-engineering
Apr 18, 2026 · 8 min read

Can Six Sigma Rein In Your LLM?

Prompt engineering tunes the mean. Narrowing the distribution takes a different toolkit — one that came out of factories a hundred years ago.

ai-engineering
Apr 16, 2026 · 6 min read

Your LLM is always telling you a story

The model always hands you a systematic-sounding story. It usually isn't the truth. What FBI cognitive interviewing teaches you to do next.

ai-engineering
Apr 15, 2026 · 4 min read

How to make your LLM reliable

Four rules from Karpathy's skills repo — and why they're the same rules you'd teach a junior engineer.

ai-engineering
Apr 4, 2026 · 6 min read

When Everyone Has AI, What's Your Edge?

AI creates more jobs than it destroys — history proves it. Here are the 5 skills that matter for what's coming.

career
Mar 29, 2026 · 3 min read

Always Hungry: A Self-Evolution Mode for the AI Era

What if your project could watch the open-source community, learn from it, and upgrade itself — without you doing anything?

agents
Mar 22, 2026 · 7 min read

Becoming an AI Psychosis

I gave my AI agent a design language and 6 core pages. It built 20+ production pages on its own. I didn't touch a thing.

agents
Mar 18, 2026 · 8 min read

What Figma Code Connect Changes: Revising the AI Coding Pipeline Strategy

Figma Code Connect + custom strategies working in layers — combining the best tools to maximize AI coding pipeline output.

design-systems
Mar 17, 2026 · 9 min read

Catching the AI Boat: A Software Engineer's Transformation Guide

AI doesn't replace engineers — it replaces the version of engineering we grew up doing. Here's how to make the shift.

career
Mar 15, 2026 · 10 min read

Reducing AI's Dependency on Figma Standardization Through Engineering

Don't ask designers to change how they work. Absorb Figma's messiness in the engineering layer instead.

design-systems
Mar 13, 2026 · 5 min read

One Metric Is All You Need: Measuring AI Engineering Productivity

Velocity measured effort. FPI measures trust. Here's why it's the only metric that matters when AI writes your code.

ai-engineering