[[AI as Cognitive Scaffold|AI amplifies]] intention — good or bad. Undirected AI use doesn't produce neutral output, it produces noise at scale. The curl project's AI slop crisis is a documented case study: well-meaning or careless actors flooding the bug-bounty program with low-quality AI-generated submissions forced its maintainer to end the bounty in January 2026, and unpaid reporting resumed that March with no money attached [S16]. What the maintainer observed afterwards is the part worth holding: the flood receded once the reward was withdrawn, not once the reports improved. The disqualifier was the incentive, not the tool. The Linux kernel faced a milder version of the same pattern and resolved it through ordinary governance — an Assisted-by disclosure tag for AI-assisted contributions, not a breakdown [S17].

This is fundamentally an [[Epistemic Hygiene]] failure — treating unverified AI output as valid signal.

[[The Human Algorithm]] is explicitly a framework for [[Intentionality Architecture|directed, intentional]] AI use. [[The Sovereignty Principle|Responsibility]] is not optional — it's structural. An optimizer running without a [[The Optimization Paradox|loss function]] isn't optimizing, it's just computing. Humans must provide the loss function. Slop is what happens when they don't.

Noise is harm at the information layer. [[Peace is the Protocol]] is the signal state — [[Constructive Displacement]] is how signal actively overtakes noise rather than merely outlasting it.