When Your Own Output Becomes Your Training Data: Noise-to-Meaning Loops and a Formal RSI Trigger

Researchers present N2M-RSI, a formal model showing that AI systems feeding their own outputs back as inputs can experience unbounded complexity growth once crossing an information-integration threshold. The framework applies to both individual AI agents and swarms of communicating agents, with implementation details withheld for safety reasons.

• →The N2M-RSI model demonstrates that AI systems can achieve recursive self-improvement through feedback loops of their own outputs.
• →Once an AI agent crosses a specific information-integration threshold, its internal complexity grows without bound under the model's assumptions.
• →The framework unifies concepts from self-prompting language models, Gödelian self-reference, and AutoML in an implementation-agnostic way.
• →The model scales to interacting AI agent swarms with potential super-linear effects when communication is enabled.
• →Researchers deliberately omitted system-specific implementation details citing safety concerns.