The Limits and Possibilities of Artificial Intelligence
In this panel discussion from the COSM technology conference, the panelists address the limits and possibilities of artificial intelligence (AI). George Montañez, an Assistant Professor of Computer Science, discusses generative AI models and how they work by encoding relations between text and images. He also mentions the phenomenon of model collapse, where AI systems degenerate when trained on their own outputs. Robert J. Marks, a Professor at Baylor University, talks about the history of AI and the limitations of AI in terms of understanding and creativity. He also discusses the concept of the Chinese room experiment and how AI systems do not truly understand what they are doing. Bill Dembski, a mathematician and philosopher, discusses the concept of digital immortality and the limitations of computational power. He also mentions the importance of human creativity and inspiration in contrast to AI systems. The panelists also touch on topics such as prompt engineering, common sense, and the singularity.
