A center applying epistemic best practices to predicting & evaluating AI progress
Artificial Intelligence and Epistemic Institutions
Forecasting and evaluating AI progress is difficult and important. Current work in this area is distributed across multiple organizations or individual researchers, not all of whom possess (a) the technical expertise, (b) knowledge & skill in applying epistemic best practices, and (c) institutional legitimacy (or otherwise suffer from cultural constraints). Activities of the center could include providing services to AI groups (e.g. offering superforecasting training or prediction services), producing bottom-line reports on “How capable is AI system X?”, hosting adversarial collaborations, pointing out deficiencies in academic AI evaluations, and generally pioneering “analytic tradecraft” for AI progress.
A center applying epistemic best practices to predicting & evaluating AI progress
Artificial Intelligence and Epistemic Institutions
Forecasting and evaluating AI progress is difficult and important. Current work in this area is distributed across multiple organizations or individual researchers, not all of whom possess (a) the technical expertise, (b) knowledge & skill in applying epistemic best practices, and (c) institutional legitimacy (or otherwise suffer from cultural constraints). Activities of the center could include providing services to AI groups (e.g. offering superforecasting training or prediction services), producing bottom-line reports on “How capable is AI system X?”, hosting adversarial collaborations, pointing out deficiencies in academic AI evaluations, and generally pioneering “analytic tradecraft” for AI progress.