Survival of the Best Fit (www.survivalofthebestfit.com)

Survival of the Best Fit is an educational game that teaches how hiring algorithms can reproduce and amplify human bias. Players experience hiring scenarios that reveal how choices about data, features and objective functions lead models to prefer certain candidates, often via seemingly neutral proxies. The game reframes AI not as an unstoppable external force but as a set of design and policy decisions people can understand and influence, making it a hands‑on tool for public education and advocacy around algorithmic fairness. For the AI/ML community, the project is a compact reminder that technical decisions—training data curation, labeling practices, feature engineering, loss functions and evaluation metrics—have social consequences: biased inputs and improperly defined objectives produce inequitable outcomes and feedback loops that entrench inequality. By making these mechanisms tangible, the game encourages practitioners to prioritize transparency, accountability, robust fairness testing and stakeholder engagement. Survival of the Best Fit thus serves both as outreach to non‑technical audiences and as a prompt for researchers and engineers to embed fairness and governance into the development lifecycle of hiring systems and other pervasive software.