I built something unusual and open-sourced it.
The PUT Monolith is a compact, system-neutral, AI-ingestible ruleset for reasoning about public finance in a future where automation, AI, and robotics dominate value creation.
It’s not a policy proposal. Not political. Not an LLM jailbreak.
It’s an architectural core: the invariants, guardrails, and ethical constraints that make consistent reasoning possible across any AI model.
The Monolith is small enough to text to a friend, but strong enough to constrain an AI’s reasoning and force it to honor fairness, system alignment, footprint-based contribution, and non-regression.
Multiple models (GPT, Claude, Grok) have already demonstrated stable reasoning when loaded with it.
The goal is to provide a portable building block for:
- AI alignment & systems governance
- reasoning about incentives
- economic modeling / public finance
- research, critique, and open testing
MIT-licensed.
GitHub repo:
https://github.com/publicusagetax/PUT-Monolith-v2
Feedback, critique, forks,
