An early warning system for novel AI risks May 25, 2023 by Deepmind New research proposes a framework for evaluating general-purpose models against novel threatsRead More Previous Post “Building a model that can save as many lives as possible” Next Post Democratic Inputs to AI
