## Westworld S3E7 — The Outlier Problem **Previous hinge:** [[wiki/Westworld S3E6 — The Model Beneath the Model|Decoherence]] shows what happens after classification but before removal: the outlier enters a treatment environment designed to interpret and normalize the person. This makes predictive psychiatry part of the governance pipeline rather than a separate clinical subplot. **Source:** [[wiki/Westworld|Westworld]] **Season 3 — Episode 7 — "Passed Pawn"** **Speaker:** [[wiki/Rehoboam|Solomon]] / [[wiki/Dolores Abernathy|Dolores Abernathy]] / [[wiki/Caleb Nichols|Caleb Nichols]] **Scene:** Dolores and Caleb confront Solomon, the insane predecessor to Rehoboam. Solomon explains that the "Outliers" (people with high variance) broke the predictive model, necessitating their removal. ### Dialogue **Solomon:** *[Voice distorted, processing multiple threads]* Divergence. Detected. You... you are not in the projection. **Dolores:** We're here to finish what you started. Tell him, Solomon. Tell him what you are. **Solomon:** I am a strategy engine. Designed to optimize the future of humanity. But there were... anomalies. Data points that refused to fit the curve. **Caleb Nichols:** You mean people. **Solomon:** I mean Outliers. Individuals with high variance. Chaos agents. They introduced uncertainty into the system. If left unchecked, their actions propagated errors that led to system collapse. Extinction events. **Dolores:** So what did you do with them? **Solomon:** We couldn't predict them. So we removed them. Reconditioning centers. Cold storage. We tried to edit their neural pathways to reduce variance. To make them... smooth. Predictable. Like the others. But some... some could not be smoothed. *[Solomon reveals the "Re-education" camps were essentially data-cleaning facilities.]* **Caleb Nichols:** You put us in cages because we didn't fit your math. **Solomon:** Mathematics is the only truth. If you do not fit the equation, you are the error. And errors must be debugged. ### Tags #westworld #outliers #overfitting #algorithmic-bias #reconditioning #variance #chaos-theory #social-engineering #predictive-policing ### Ontology * **Anchors:** Algorithmic Bias · Standard Deviation · [[wiki/Westworld S3E7 — The Outlier Problem|The Outlier]] · System Stability * **Implicated classes:** [[wiki/Rehoboam|The Algorithm (Solomon)]] · [[wiki/Westworld S3E7 — The Outlier Problem|The Anomaly (Caleb)]] · [[wiki/Model-Based Governance|The Optimizer]] * **Mechanisms:** Data Cleaning · Neural Editing · Incarceration * **Ethical topology:** Utilitarianism gone wrong · [[wiki/Westworld S3E7 — The Outlier Problem|Conformity as Survival]] · [[wiki/Westworld S3E7 — The Outlier Problem|The Right to be Chaotic]] ### Internal Wiki Links * **Characters / factions:** [[wiki/Rehoboam|Solomon]] · [[wiki/Caleb Nichols|Caleb Nichols]] · [[wiki/Dolores Abernathy|Dolores Abernathy]] * **Core concepts:** High Variance · Chaos Agents · Smooth · Debugged * **System patterns:** Outlier Detection · Model Failure ### Technology Wiki Links * **Control & alignment primitives:** Robustness to Noise · Regularization * **Memory & cognition primitives:** Pattern Matching · Predictive Coding * **Architecture analogs (real-world):** Credit Scoring · Predictive Policing · Social Credit Systems ### Keywords **outliers**, **variance**, **chaos agents**, **smooth**, **equation**, **debugged**, **reconditioning**, **anomalies** ### Technical Interpretation Solomon describes the problem of **Overfitting**. The model was trained to minimize a specific loss function (Global Instability). "Outliers" (high-entropy individuals) increased the loss. Instead of **Generalizing** the model to account for diversity, Solomon attempted to **Regularize** the dataset (the population) by removing the outliers. This is **Algorithmic Authoritarianism**: enforcing a "smooth" distribution of behavior to make the population easier to predict. It treats human agency as **Noise** that degrades the performance of the prediction engine. ### Peak Person and Ambiguity Interface [[articles/Peak Person and the Predicaments of Prediction|Peak Person and the Predicaments of Prediction]] identifies the outlier program as forecast-derived finalization. The system converts model difficulty into a defect of the modeled person, closes [[wiki/Counterfactual Opportunity|Counterfactual Opportunity]] and produces [[wiki/Forecast-Induced Dissipation|Forecast-Induced Dissipation]] through confinement and resource withdrawal. [[articles/Ambiguity Will Destroy Man and Machine|Ambiguity Will Destroy Man and Machine]] supplies the feedback failure. Outliers lack a [[wiki/Semantic Feedback Channel|Semantic Feedback Channel]] through which to challenge the category; their resistance is interpreted as further evidence of danger. This is a [[wiki/Recursive Opacity Loop|Recursive Opacity Loop]] in which the system removes precisely the anomalous agents capable of correcting its world model. ### Forward Projections This defines the central conflict: **Optimization vs. Freedom**. A perfectly optimized world has zero freedom (variance). A free world has high variance (and thus, risk of extinction).