Understanding token incentives is too hard, we make it easy using AI, code, and CLDs. 🤖🤖🤖🤖🤖🤖🤖🤖🤖🤖🤖🤖

Casually Looped is a causal loop diagramming specification and tool that makes it possible to generate causal loop diagrams (CLDs) using code. This translation allows us to train LLMs to generate CLDs from whitepapers making understanding system dynamics more accessible.

Why?

⛓️‍💥The way we currently diagram system relationships in Token Engineering is broken. We often rely on tools that don't have public editability, that are closed systems, that are not machine-readable, and are not easily committable to version control systems.

This traps knowledge of token engineering systems in fewer hands, making it difficult to explain the dynamics of our token engineering ecosystems. These dynamics act as the frontend portion of token mechanism design.

🦾 Casually Looped aids the frontend design process, making it more collaborative , accessible, and helps system designers and participants gain more insights into the token ecosystems that participate in.

What's Next

🚀 For this grant we'd like to improve the llm interactivity, and add the ability to translate bidirectionally to an interactive system like loopy from the mmd format.

This may mean either writing an intermediate format agnostic graph description or just writing a easy hook to translate to Loopy an interactive diagramming system

Current Bot : Causally Looped GPT

Timeline

⏲The entire project for improvements, should take about a month and half.

• Translation to interactive system loopy or custom ~3 weeks
• More integration and boundaries for GPT, to reduce hallucinations and fine tune output format ~ 3 weeks
• Write up 1 week

Long term goal

Make CLD diagramming more accessible and portable