Artificial Intelligence (AI)

by Christian Caballero. From the threat of nuclear war to rogue superintelligent AI to future pandemics and climate catastrophes, the world faces risks that are both urgent and deeply uncertain. These risks are where traditional data-driven models fall short—there’s often no historical precedent, no baseline data, and no clear way to simulate a future world. In cases like this, how can we anticipate the future? Forecasting tournaments offer one answer, harnessing the wisdom of crowds to generate probabilistic estimates of uncertain future events. By incentivizing accuracy through structured competition and deliberation, these tournaments have produced aggregate predictions of future events that outperform well-calibrated statistical models and teams of experts. As they continue to develop and expand into more domains, they also raise urgent questions about bias, access, and whose knowledge gets to shape our collective sensemaking of the future.

by Sohail Khan. Netflix knows what you like, but does it need to know your politics too? We often face a frustrating choice: share our data and be tracked, or protect our privacy and lose personalization. But what if there was a third option? This article begins by introducing the concept of the privacy-utility trade-off, then explores the methods behind strategic data distortion, a technique that lets you subtly tweak your data to block sensitive inferences (like political views) while still maintaining useful recommendations. Finally, it looks ahead and advocates for a future where users, not platforms, shape the rules, reclaiming control of their own privacy.