Artificially Intelligent

88: Robots and Rubik's Cubes

11.04.2019 - By Christian Hubbs and Stephen DonnellyPlay

Download our free app to listen on your phone

Download on the App StoreGet it on Google Play

OpenAI recently announced a new achievement, solving a Rubik's Cube with a one-handed robot using reinforcement learning! After all of the fanfare, a lot of people are pushing back on their paper and results, arguing that the claims are highly misleading. On top of that we talk about Tesla's recent return to profitability and reveal a new change to the show! Links: OpenAI  Solving a Rubik's Cube Episode 84: OpenAI's Emergent Agents Episode 58: Beaten at our Own Games Gary Marcus' Response Follow us and leave a rating! Patreon iTunes Homepage Twitter @artlyintelly Facebook Learn Economics at Liberty Classroom

More episodes from Artificially Intelligent