Publishing the AI Collaboration Operating SOP v0
Date: 2026-05-03
Why This Log
This is not a feature completion report.
During a review of the May operating Board with Meta-chulbuji, the conversation surfaced something worth recording — not just what was built today, but what stage the operation is at and where it’s heading.
This Log is about that shift.
The Starting Point: Reviewing the May Board
The May operating Board lists auto-trading, the AI blog, the AI content service, the AI music experiment, video editing automation, and shopping content tests.
At first glance, these look like entirely different projects scattered across different domains. But running them in parallel makes something clear — they all depend on a shared foundation.
Planning ability — knowing which experiments to run and why
Content production ability — turning ideas into actual outputs
Product-building ability — completing what’s built into something that works
AI tool utilization — knowing which AI to deploy, when, and how
Marketing understanding — explaining why what’s built is valuable and to whom
Monetization experiment design — building and testing a path to revenue
The projects look different, but this common foundation is what makes each experiment actually work.
Without organizing this foundation, the number of experiments grows while the ability to operate them doesn’t keep pace. Today’s work was putting that foundation into a public document.
What Stage Is This
AI collaboration can be broken into roughly these stages:
- Learning AI — getting familiar with AI tools, understanding what different inputs produce
- Building with AI — using AI to produce real outputs
- Operating with AI — treating AI not as a single tool but as a set of roles to deploy, running projects from planning through asset-building with AI as a partner
- Multi-expert AI orchestration — with Meta-chulbuji at the center, deploying multiple specialist AIs while Manager(chulbuji) focuses on direction and final judgment
Right now, Manager(chulbuji) is moving from stage 3 toward stage 4. This can be described as “year one of operating as an AI-collaborative business manager.”
The learning stage is behind. Real things are being built. And now the work is moving from building to constructing the operational structure itself.
Current Strengths of Manager(chulbuji)
These are the strengths that are actually working at this stage:
- Fast execution: Short distance from idea to real output
- AI collaboration ability: The way of organizing thinking with Meta-chulbuji and deploying specialist AI is taking shape
- Recording and asset-building habit: A consistent flow of logging results into Log, Insight, SOP, and Board
- Implementation drive: The ability to actually ship pages, MVPs, and projects
- Engineering-based structural thinking: The ability to understand and design complex systems as structure
Areas That Need Strengthening
Alongside the strengths, the gaps are also clear:
- Market understanding: Sensing which market is being addressed and what problem it’s solving
- Customer definition: The ability to draw a specific picture of who the experiment is for
- Marketing language: The language to make what’s built sound valuable
- Copywriting: Writing sentences that people actually respond to
- Product planning: Designing ideas into sellable, structured form
- Content distribution: Building a structure that gets content in front of people
- Monetization experiment design: The methodology for building and testing a path to revenue
Some of these can be supported by AI collaboration. Others come from direct experience. The experiments currently running are also the process of building these capabilities.
What AI Collaboration Operating SOP v0 Is
The “AI Collaboration Operating SOP v0” published today is not a detailed execution manual.
Detailed prompts, automation scripts, and internal channel operations stay private. SOP v0 sits above those — it’s the high-level standard.
Specifically, it organizes four roles:
- Manager(chulbuji): Sets direction and makes final calls
- Meta-chulbuji: Organizes thinking, structures work, orchestrates specialist AI
- Specialist AI: Domain-specific execution (content, development, research, copywriting, etc.)
- Record System: Asset-building through Log / Insight / SOP / Board
And it specifies how these roles collaborate, what criteria determine experiment priority, and what is made public versus what stays private.
Made public: Operating philosophy, collaboration structure, generalizable insights, asset-building criteria, project status and decision-making process, experiment results (both success and failure)
Not made public: Raw prompt text, auto-trading detailed strategy, revenue figures and account information, operating channel details, internal execution manuals
This boundary isn’t fixed. It can shift based on operating judgment.
What Was Actually Applied Today
The work didn’t stop at writing the SOP document — it was also reflected in the site structure.
- Korean SOP page created: /sop/ai-collaboration-operating-sop/
- English SOP page created: /en/sop/ai-collaboration-operating-sop/
- SOP auxiliary link added below the Board CTA on the home page
- SOP auxiliary link added at the bottom of the Operating Board index page
- SOP was not added to the top navigation — it is positioned as a reference standard document connected to the Operating Board, not as a primary menu item
Today’s Conclusion
chulbuji.com is expanding beyond recording what’s being learned about AI — into a space that experiments with how to operate projects and build assets with AI.
SOP v0 is the starting point of that expansion. It is not a finished system. It’s a v0 that will keep being revised as the operation runs. As experiments accumulate, roles become clearer, and new patterns are confirmed, the document will be updated.
The standing point right now is the beginning. The next step is confirming whether this standard actually works.
Related Links