Daniel Guetta on the Guts of AI, Agentic AI & Why LLMs Hallucinate | The Real Eisman Playbook Ep 46

The podcast explores the current state and future potential of AI, particularly large language models (LLMs). Daniel Guetta, a Columbia Business School professor, explains the mechanics of LLMs, including embeddings and transformers, and addresses the phenomenon of "hallucinations," noting they arise from the models' reliance on pattern mimicry rather than true understanding. The discussion covers the practical applications of AI, such as content moderation and agentic AI, emphasizing the importance of well-structured data and IT systems for effective implementation. Guetta and host Steve Eisman also consider the potential impact of AI on software companies and management consulting, weighing the balance between hype and real-world value. While acknowledging the limitations and biases inherent in current AI models, Guetta expresses optimism about their potential for positive change across various industries.