Collision & Resonance

Reactions to a three-hour interview with Zhang Yueguang (Miaoducamera creator, ~$40M raised for 'AI friends'). On why AI companions must evolve, why the feeling of being needed matters more than satisfaction, and what 'Dao rises, skill fades' means for products and teams. Not a systematic analysis — a few ideas from the podcast that happened to align closely with my own experience building AI companions.

My blog had 83 Chinese posts — all of them read like translated English. Stiff, formal, academic. So I studied China's best WeChat writers, distilled their craft into a 435-line writing system, and fed it to Claude as a system prompt. Then I rewrote every single post. This is the story of that guide.

MrBeast wrote 36 pages. Netflix distilled theirs to 5. Duolingo designed a 64-page illustrated book. All three generate over $1M per employee. All three are solving the same problem: how to transfer taste, judgment, and operating principles to another entity so it can execute autonomously. These handbooks are CLAUDE.md files for humans — and they reveal three distinct archetypes for how organizations might scale in the agentic AI era.

One style guide fixed translation-ese, but a tech review shouldn't sound like social commentary. Different content needs different voices. I developed a systematic method for 'stealing' any author's writing style — collecting articles, dissecting patterns, and extracting a structured Style Profile. Here's the full process, demonstrated with a real example.

If employee handbooks are CLAUDE.md files for humans, what should the handbook of an AI-native company look like? What kind of people does it hire — and how does it screen for taste instead of skill? This piece designs the AI-native employee handbook from scratch: the talent profile, four screening methods, and the five sections every AI-native handbook needs — from mission to judgment frameworks to agent collaboration protocols.

A style library, no matter how large, can't cover everyone's taste. Some users want to write like an author who isn't in the library. Some want the voice of one specific article, not a generalized style. The solution: real-time style extraction. Give the AI any reference article, and it clones the writing DNA on the fly.

Reactions to a podcast on AI-first org architecture. Use AI as a tool and you're racing the clock — 10x at best. Use it as a system and you rebuild the whole workflow, retreating to just two jobs: give intent, review output. But the review gate is drowning, and that's the part nobody wants to say out loud.