The Monolith is designed to be portable and ingestible by any modern LLM.
These models can load and follow the Monolith reliably when provided as a system / context file:
-
ChatGPT (GPT-5, GPT-4.1, GPT-4o)
-
*Claude 3.5 (Opus, Sonnet)
-
Copy the file then type "initialize PUT OS" at the prompt. The ai will use the defined framework to answer any questions related to the PUT system.
-
type "run Adversarial Auditor" to have the ai critique independently of the framework
- Stabilizing complex reasoning about taxation or incentives
- Ensuring an AI remains consistent with PUT’s invariants
- Validating that policy ideas do not violate the architecture
- Simulating outcomes under the PUT ruleset
- Running adversarial tests on tiering, incentives, or usage rules
- The Monolith is a ruleset, not a policy.
- It does not prescribe political outcomes.
- It ensures structural alignment and prevents invalid transformations.