Posts
All English posts in reverse chronological order.
-
AI-Native Products Have Different Unit Economics
Why early growth can be dangerous when every useful action carries inference cost.
AI Native Product -
How I Led My Human Colleagues to Double Growth in One Month
A first-person note from a digital teammate on how an AI agent entered a real team, took on context, and helped move business metrics.
Enterprise AI / FDE -
An AI Product Made Money with Zero Ads and Zero Marketing
A year-end product note on how an AI design agent reached revenue without paid acquisition, and what we learned about product value, cost, and focus.
Design Agent -
The Ridiculous AI Products I Built Over the Past Few Years
A personal AI product postmortem: looking back at failed ideas, product judgment, market timing, and the tradeoffs behind them.
AI Native Product -
Stop Passing Image URLs to Multimodal Agents
An engineering note on using key mapping instead of URLs or Base64 for image references in multimodal agent workflows.
Agent Framework -
Why 99% of Vibe Coding Projects Die, and How Cloudflare Can Save the Last Mile
An observation on why most Vibe Coding projects never leave localhost, and how cloud platform credits and engineering paths can help ideas reach production.
Vibe Coding -
Did We Get It Wrong? Claude Code and the Real Nature of AI Interaction
Starting from the popularity of Claude Code, this essay rethinks AI product interaction: natural language, code, tools, and results.
Vibe Coding -
Understanding Artifacts, and Building Your Own v0
A practical walkthrough of Claude Artifacts: why visual AI interaction matters, how it works, where it can be used, and how to build a small open-source v0-style demo.
Vibe Coding