model rot

When your LLM prompts stop working as well — the model has been updated, training data has shifted, or the context window is now choked with irrelevant history.

Gradual degradation of agent performance as its conversation history fills with tool outputs, errors, and old instructions. The failure mode that mandates compaction or summarization.

What happens when AI models are trained on the output of earlier AI models — quality degrades generation by generation as hallucinations compound. Increasingly real as the web fills with AI slop.

The active state of being fully in your rotted era — when you've surrendered to the chaotic pull of the internet and stopped fighting the brain rot. Rotted mode isn't a lapse, it's a choice: phone in hand, no plans, consuming content of questionable quality and uncertain origin for hours without guilt or apology. It's distinguished from passive rotting by its intentionality — you're not spiraling, you're deliberately sitting in the pit and enjoying it. Often declared before a weekend with nothing to do and zero ambition to fix that.

The rare and coveted form of brain rot that actually pays off — when your terminally online tendencies lead to something genuinely good, funny, or socially useful. W rotted people are so deep in specific internet corners that they discover trends early, make connections others miss, or produce content that resonates precisely because they've absorbed so much of the culture. It's the validation that sometimes all those hours online were not completely wasted: you were doing research, and now your weird niche knowledge is suddenly relevant.