On this episode, host Robert J. Marks concludes his conversation with economics professor and author Gary Smith about the hype and limitations of generative AI. Smith is the Fletcher Jones Professor of Economics at Ponoma College and a frequent contributor to Mind Matters News. In this portion of the conversation, Smith and Marks explore the hype around artificial general intelligence (AGI) and explain how current large language models lack true reasoning and creative capabilities, despite regular claims of impending AGI from people like OpenAI’s Sam Altman. Smith provides examples demonstrating how these models make nonsensical or incorrect responses to logical problems and financial questions, highlighting their inability to understand context and perform meaningful reasoning. The “Band-Aid” approach of fixing model failures is not a viable path to achieving AGI. The conversation concludes with a cautionary note about the risks of over-trusting AI systems, especially for important real-world decisions. This is Part 2 of a two-part discussion.

Additional Resources

• The AI bubble is looking worse than the dot-com bubble. Here’s why – Market Watch
• The Promise of Artificial General Intelligence is Evaporating – Mind Matters News
• Listen to Part 1 of this discussion.