DIGITAL PUBLIC GOODS / APACHE 2.0

Three operating systems
Three constituencies

Individuals, small organizations, cities.
They are not products. They are digital public goods — open source, free, permissionless: anyone can use, modify and deploy them.

01 / Individuals · Designing

Sin90 — the individual growth loop

Everyone runs the same loop every day; only the contents differ. Opening your eyes and checking the clock, scrolling WeChat, scrolling Xiaohongshu — even just pulling the curtain to look at the weather — all of it is taking in information. That is your first step.

Information what you take in Action what you do Connection people · tools · world Communication you keep ties in order to talk And you talk in order to — take in information again. That is how the loop closes. Life · learning · social · work — all personal growth happens inside this loop SIN90 · AN INDIVIDUAL OS BUILT AROUND THE GROWTH LOOP · APACHE 2.0

Sin90 has exactly one goal: to stand around this loop and give an ordinary person a lift in work, life, learning — even starting something of their own, so that some of the wellbeing of digital life actually reaches them.

02 / Communities · small orgs · businesses · In progress

Cos72 — the organization collaboration system

Plenty of small organizations have no engineering muscle, no technical staff, and can't even afford a single FDE. So can they be their own FDE? Break the job apart, collaborate to find the innovators and the builders, and build the organization's own AI capability the ANE way. I think they can.

CORE 01

Knowledge Base

知识库

The knowledge an organization has accrued is scattered across databases, the heads of long-serving staff, and the impressions of customers. In the age of AI, it has to be laid down and kept.

CORE 02

Organizational Workflow

组织工作流

Mission, vision, values, business logic — why does this thing exist at all? All that soft matter finally has to land through process. Nonprofits and businesses: same principle.

CORE 03

Organizational Brain

组织大脑

An LLM API (OpenAI / Anthropic / DeepSeek⋯) + a self-trained LoRA: fine-tune on the organization's own data for its own scenarios, until the model knows that organization inside out.

The self-trained LoRA is a conclusion for now. The future may break it — but today, it is what lifts an organization's process efficiency the most.

— On the organizational brain

Connectors / Sensors (agents of all kinds)

An organization's workflow, like an individual's, needs a great many connectors and sensors. At bottom they are all agents:

  • Scraping and analyzing industry data, and judging the news
  • Auto-connecting outside suppliers and pulling quotes
  • Pushing new products to customers on a cadence, and after-sales service

These agents take in information and hold up your organization's workflow.

03 / Cities · Long term

CityOS — the city becomes an agent

Once every organization has a brain that knows its own context well enough, sharing between organizations will make new patterns of collaboration emerge — and resources will pool inside the city.

EMERGE 01

Inter-Org Collaboration

组织间协作

Nobody has to be omniscient. Each side only has to know enough of the other's information and context. New patterns of collaboration will grow on their own.

EMERGE 02

Attract Resources

吸引资源

Collaboration bounded by a city can reach across that boundary to draw in more outside resources.

EMERGE 03

Attract Talent

吸纳人才

The city itself becomes an agent that collaborates with everyone — instead of playing everyone off against each other.

Chiang Mai is a tourist city. How does it draw more visitors than the other tourist cities of the world? I believe that is a question the mayor, the city government, and every resident would like answered. CityOS exists to answer it.

No one has tried this. But only those who go first get the dividend of digital wellbeing.

— On CityOS