Unveiling AI's Illusions: A Deep Dive into Understanding and Misunderstanding with Gary Marcus and Michael Wooldridge Patreon: https://www.patreon.com/mlst Discord: https://discord.gg/ESrGqhf5CB Twitter: https://twitter.com/MLStreetTalk Artificial Intelligence continues to advance at an unprecedented rate, but are we falling victim to the hype and misconceptions surrounding its true capabilities? In this enlightening and thought-provoking discussion, Professors Gary Marcus and Michael Wooldridge take us on a journey to uncover the truth behind AI's capabilities, limitations, and the future of this groundbreaking technology. Join us as they explore the current state of AI models like GPT-4 and how they differ from human intelligence, dissecting the difference between intelligence and understanding, and providing valuable insights into the potential role of AI in shaping future technologies. Along the way, they also address the Swiss Cheese Problem and its implications on AI performance, and the intenSion and extension of understanding in language models. Whether you're an AI enthusiast, a researcher, or simply curious about the future of technology, this conversation will provide you with a better understanding of the intricate world of AI and its implications on our lives. Don't miss out on this opportunity to learn from two leading experts in the field as they separate fact from fiction in the rapidly evolving world of artificial intelligence. Table of Contents: The Tsunami of AI Misinformation [00:00] GPT-4: Capabilities and Limitations [04:12] William Shatner and Gary Marcus [07:56] Gary on WIRED - Human vs. Machine Intelligence [09:57] Intelligence vs. Understanding [10:35] Prof Michael Wooldridge (Oxford University) https://www.cs.ox.ac.uk/people/michael.wooldridge/ Prof. Gary Marcus: https://garymarcus.substack.com/ References: 1. Marcus, G. (2023). GPT-5 and Irrational Exuberance. https://garymarcus.substack.com/p/gpt-5-and-irrational-exuberance 2. Mitchell, M. (2023). The Debate Over Understanding in AI's Large Language Models https://arxiv.org/abs/2210.13966 3. Yasaman Razeghi (2022). Impact of Pretraining Term Frequencies on Few-Shot Reasoning (Yasaman Razeghi et al) https://arxiv.org/abs/2202.07206 4. Wooldridge, M. (2021). The Road to Conscious Machines: The Story of AI https://www.amazon.co.uk/Road-Conscious-Machines-Story-Pelican-ebook/dp/B07XCC7BMQ Check out Gary's new podcast here! https://podcasts.apple.com/us/podcast/humans-vs-machines-with-gary-marcus/id1532110146 Videos clips used from: https://www.youtube.com/watch?v=Yf1o0TQzry8 https://www.youtube.com/watch?v=-fopYsgFdzc https://www.youtube.com/watch?v=F2LeyULGF2Q WIRED interview with Gary: https://www.youtube.com/watch?v=Puo3VkPkNZ4 Our thanks to the Turing Institute who gave us a press pass to film at AIUK https://ai-uk.turing.ac.uk/programme/ Post production note: we (mindlessly) used an old clip of Gary being interviewed by his childhood hero William Shatner from the Russian propaganda station "RT" (filmed before the invasion) - this was a mistake, and we were guilty of what we accused GPT users of i.e. not engaging brain. In case it didn't go without saying, we (and Gary) don't endorse RT and abhor the actions of Russia in Ukraine.