AI Safety Research

Ensuring that Current and Future Intelligent Machines are Well-Behaved


  • December 10, 2019: We will present a Seldonian contextual bandit algorithm at NeurIPS 2019. We applied our algorithm to predicting whether someone will commit a crime in the future, predicting whether someone will repay a loan, and optimizing an online tutoring system to maximize student performance, all while enforcing various fairness guarantees. [link]
  • November 22, 2019: Our paper introducing Seldonian algorithms was published in Science [link]. [Press releases 1, 2]