The robots who predict the future
The article reviews arguments that predictive algorithms, trained on biased historical data, increasingly determine outcomes in employment, health, and civic life, with profit motives amplifying harms. Proposed remedies include democratic data control such as data trusts, corporate taxation for social harm, and institutional reforms, while critics warn foundational decision-theory assumptions entrench harmful automation in health-risk prediction.
Why it mattersAlgorithms that screen out people likely to develop chronic health problems should force insurers to audit models.