OhMyGame 开源 10 天拿到 936 个 star 的一个小复盘star 曲线、流量来源Notes, August 2026“Less is more,” the “60% rule,” and “doing something extremely well” are really about the same thing: figure out the few things that truly matter, focus on them, and do them as well as you possibly can. For everything else, 60% is good enough. Lately, I feel like my brain is constantly looking for some kind of stimulation, and because of that, I end up…
Hard Work != High OutputWhy long hours and individual effort do not guarantee high output, and how efficiency, leverage, direction, and fast execution multiply results.
Think in 10x2x thinking is linear: it asks how to do more of what already works. 10x thinking forces you to question the current path entirely. Because we cannot simply multiply our time and effort, a 10x goal pushes us toward fundamentally better strategies, higher standards, and nonlinear solutions. 80% of the results often come from 20% of the effort. 10x growth…
Maybe It's Time to Rebuild the 3D StackAI-native 3D Studio: Use AI to create 3D worlds you can render and play.
My Thoughts on AI Interactive DramaAI interactive drama can be seen as an extension of AI-generated short-form drama: video and storytelling remain at the core, but viewers can make choices at key moments and influence what happens next. Interactive video is not a new concept. Around 2020, the success of The Invisible Guardian led major Chinese internet companies to view it as a promising…
Traffic to BrandA private note.
Introducing cut-as-codeAn open-source experiment in turning video editing into a coding problem for AI agents.
Embodied AI Models I’m Following1/6 A quick map of the embodied AI model teams I’m recently following. I’m mainly interested on the "robo brain", including model, data, and training. Three broad groups: VLA/action policies, world-action models, and alternative architectures. 2/6 VLA / action-policy models: Physical Intelligence — π0.7 Figure — Helix 02 AgiBot — GO-1 / ViLLA NVIDIA — GR00T…
CEO's Job (1)What a CEO should focus on beyond building the product.
The Math Behind SharingA simple model for how shared content drives product sign-ups.
Make the Core Work FirstWhy adding more features will not fix a weak core product.
Designing with Code vs. Figma MCPWhat I learned from designing with both code and Figma.
Vofy homepage: landing page or dashboardA product note on why Vofy's homepage feels unclear when it tries to be both a landing page and a dashboard.
Building for everyone is making Vofy confusingI realized Vofy is trying to serve too many users at once, and that is making the product harder to understand.
How I Map AI Creative ToolsA small product map for comparing AI creative tools by what users actually buy.
TikTok's MathAssume TikTok acquires a cohort of N users, with an average customer acquisition cost of CAC per user.
Hard Work != High OutputWhy long hours and individual effort do not guarantee high output, and how efficiency, leverage, direction, and fast execution multiply results.Think in 10x2x thinking is linear: it asks how to do more of what already works. 10x thinking forces you to question the current path entirely. Because we cannot simply multiply our time and effort, a 10x goal pushes us toward fundamentally better strategies, higher standards, and nonlinear solutions. 80% of the results often come from 20% of the effort. 10x growth…Maybe It's Time to Rebuild the 3D StackAI-native 3D Studio: Use AI to create 3D worlds you can render and play.My Thoughts on AI Interactive DramaAI interactive drama can be seen as an extension of AI-generated short-form drama: video and storytelling remain at the core, but viewers can make choices at key moments and influence what happens next. Interactive video is not a new concept. Around 2020, the success of The Invisible Guardian led major Chinese internet companies to view it as a promising…Traffic to BrandA private note.Introducing cut-as-codeAn open-source experiment in turning video editing into a coding problem for AI agents.Embodied AI Models I’m Following1/6 A quick map of the embodied AI model teams I’m recently following. I’m mainly interested on the "robo brain", including model, data, and training. Three broad groups: VLA/action policies, world-action models, and alternative architectures. 2/6 VLA / action-policy models: Physical Intelligence — π0.7 Figure — Helix 02 AgiBot — GO-1 / ViLLA NVIDIA — GR00T…CEO's Job (1)What a CEO should focus on beyond building the product.The Math Behind SharingA simple model for how shared content drives product sign-ups.Make the Core Work FirstWhy adding more features will not fix a weak core product.Designing with Code vs. Figma MCPWhat I learned from designing with both code and Figma.Vofy homepage: landing page or dashboardA product note on why Vofy's homepage feels unclear when it tries to be both a landing page and a dashboard.Building for everyone is making Vofy confusingI realized Vofy is trying to serve too many users at once, and that is making the product harder to understand.How I Map AI Creative ToolsA small product map for comparing AI creative tools by what users actually buy.TikTok's MathAssume TikTok acquires a cohort of N users, with an average customer acquisition cost of CAC per user.