Before LLMs, writing took effort. In some circumstances (like pasting raw LLM output into a chat thread), the sloppypasta effectively becomes a filibuster, crowding out the existing conversation and blocking the viewport. Before LLMs, trust was the default. However, shared LLM output obfuscates the chain of trust. Sloppypasta creates a compounding negative feedback loop where the sender forfeits learning and credibility while the recipient burns effort and loses trust.