My current summary of the state of AI risk

Abstract

Here’s the current, gloomy, state of AI risk: Scaling AI capabilities has made impressive progress in the past decade, and particularly in the past 3 years. Deep Learning has passed over the thresh…