Garbage collection day
person concept tool org talk claim — click a node to jump to its page; hover an arrow for the relation
A weekly dedicated practice used by Ryan Lopopolo's team at OpenAI in which the entire team's focus shifts to reviewing and systematically eliminating recurring AI output quality problems — specifically, every instance of "slop" (low-quality or erroneous agent output) observed during the preceding week.
Practice Description
Lopopolo describes the ritual as a Friday cadence: "one day a week, Fridays, we called it garbage collection day, where our entire job was to take every bit of slop we had observed over the course of the week". The goal is not merely to fix individual instances of bad output, but to "figure out ways to categorically eliminate it from ever happening in the first place" — a root-cause elimination approach rather than a patch-and-move-on workflow.
Role in Harness Engineering
Garbage collection day is positioned as a core component of harness engineering — the discipline of building the scaffolding, constraints, and feedback loops that allow humans to steer and agents to execute reliably. By institutionalizing a recurring slot for quality debt reduction, the practice operationalizes the idea that agent output degradation should be treated as a systematic engineering problem, not an ad hoc one. Lopopolo advocates for this approach explicitly, recommending it as a structural practice for teams building on top of AI agents. Advocacy citation
Relationship to Evals and Quality Feedback Loops
The practice implicitly functions as a manual evals-adjacent loop: observed failures during the week accumulate, are reviewed collectively, and are then translated into categorical fixes — whether to prompts, tool constraints, guardrails, or workflow design — that prevent whole classes of errors from recurring.