Machine Learning, Human Cost

In an era when algorithms increasingly shape who is admitted, hired, and heard, two provocative novellas expose the hidden human cost of machine-mediated judgment.

In Weighted for Bias, a pre-med student discovers that a medical school's admissions algorithm-built to eliminate prejudice-has quietly learned to reproduce it. As institutional values collide with data-driven decisions, the Dean must confront the uncomfortable truth: technology cannot fix a bias we refuse to name.

In The Capture Protocol, a physician-executive becomes the target of an AI-powered recruitment system that extracts more than career insights. What begins as professional curiosity spirals into a battle against an opaque network of data harvesters, engineered incentives, and algorithmic deception. When the physician plants the digital equivalent of a Trojan Horse to expose the system's fraud, the result is a chilling lesson in how confident machines can be-and how wrong.

Together, these stories probe the fragile intersection of artificial intelligence, ethics, and medicine, revealing how even the most sophisticated systems inherit the blind spots of their designers. With the precision of a physician and the insight of a storyteller, Arthur Lazarus challenges us to reconsider trust in the age of automated judgment.

Timely, unsettling, and deeply human, Machine Learning, Human Cost asks a question that grows more urgent by the day: When our systems learn from us, what exactly are we teaching them?