Because AI is statistically displacing more women from the workforce, a wave of "disparate impact" lawsuits and regulations is likely. Leveraging legal precedents like Title VII, these actions won't need to prove discriminatory intent—only that a pattern of harm exists—potentially slowing AI adoption.

The Instagram study where 33% of young women felt worse highlights a key flaw in utilitarian product thinking. Even if the other 67% felt better or neutral, the severe negative impact on a large minority cannot be ignored. This challenges product leaders to address specific harms rather than hiding behind aggregate positive data.

Harms like contacting the wrong person arise not from malicious individuals but from automated, error-prone systems designed for scale and low cost. No single person makes the mistake; rather, the system is architected to generate these incorrect outcomes by default, with no accountability.

Risk assessment tools used in courts are often trained on old data and fail to account for societal shifts in crime and policing, creating "cohort bias." This leads to massive overpredictions of an individual's likelihood to commit a crime, resulting in harsher, unjust sentences.

When AI systems are trained on historical data, such as past hiring or policing records, they learn and perpetuate existing societal biases. This creates a dangerous illusion of objectivity, where discriminatory outcomes are presented as neutral, data-driven "predictions" by an algorithm.

While AI can inherit biases from training data, those datasets can be audited, benchmarked, and corrected. In contrast, uncovering and remedying the complex cognitive biases of a human judge is far more difficult and less systematic, making algorithmic fairness a potentially more solvable problem.

When a technology reaches billions of users, negative events will inevitably occur among its user base. The crucial analysis isn't just counting incidents, but determining if the technology increases the *rate* of these events compared to the general population's base rate, thus separating correlation from causation.

The real danger of algorithms isn't their ability to personalize offers based on taste. The harm occurs when they identify and exploit consumers' lack of information or cognitive biases, leading to manipulative sales of subpar products. This is a modern, scalable form of deception.

Dr. Cathy O’Neil argues the true danger of algorithms lies not in their technical sophistication but in their opaque nature, lack of oversight, and the fact that individuals cannot opt out. Even simple logistic regressions can be terrifying under these conditions.