1 rich Libsyn https://www.libsyn.com 90 600 What happens when an AI system faithfully follows a flawed goal? In this episode, we explore how even well-designed algorithms can produce dangerous outcomes — from amplifying hate speech to mismanaging infrastructure — simply by optimising a reward function which, like all reward functions, fails to encode all that matters. We discuss the hidden risks of reinforcement learning, why over-optimisation can backfire, and how game theory helps us rethink what it means for AI to act &quot;rationally&quot; in complex, real-world environments. &amp;nbsp; Jobst Heitzig is a mathematician at the Potsdam Institute for Climate Impact Research and an expert in AI safety and decision design. Game Changer - the game theory podcast http://tws-gamechanger.libsyn.com/website <iframe title="Libsyn Player" style="border: none" src="//html5-player.libsyn.com/embed/episode/id/40135170/height/90/theme/custom/thumbnail/yes/direction/forward/render-playlist/no/custom-color/88AA3C/" height="90" width="600" scrolling="no" allowfullscreen webkitallowfullscreen mozallowfullscreen oallowfullscreen msallowfullscreen></iframe> https://assets.libsyn.com/secure/content/198644825