In the AI Risk Reward podcast, our host, Alec Crawford (@alec06830), Founder and CEO of Artificial Intelligence Risk, Inc. aicrisk.com , interviews guests about balancing the risk and reward of Artificial Intelligence for you, your business, and society as a whole. Podcast production and sound engineering by Troutman Street Audio. You can find them on LinkedIn.

In this episode, Alec welcomes Dr. Michael Erlihson, Math PhD, AI influencer, and Head of AI at DriveNets, for an insightful conversation on the evolving risks and opportunities in artificial intelligence. Dr. Erlihson shares his journey from a science-focused family in Russia to leading AI initiatives in Israel, emphasizing the foundational role of mathematics in modern AI. The discussion explores the theme "AI is making us stupid," drawing parallels to historical debates about technology’s impact on cognition, and offering strategies to ensure ongoing learning and critical thinking in an AI-driven world. Dr. Erlihson discusses his approach to reviewing scientific literature without AI tools, the importance of connecting historical math papers to today’s AI, and his work optimizing LLM inference costs. The episode closes with a practical lightning round covering AI’s impact on education, employment, data privacy, and the democratization of AI knowledge.

Summary:

AI’s Cognitive Impact: Dr. Erlihson argues that while AI will make most people less knowledgeable, it can make a select few even smarter if used to augment ongoing learning.
Mathematics in AI: Emphasizes the enduring importance of math in AI development, connecting historical mathematical insights to contemporary machine learning advances.
Optimizing AI Infrastructure: Details DriveNet’s focus on reducing LLM inference costs to ensure the economic sustainability of AI deployment.
Education & Employment: Raises critical concerns about the future of traditional education and white-collar employment as AI accelerates automation and self-learning.
Data Privacy Risks: Highlights the underestimated risks of personalizing AI with private data and advocates for stronger safeguards and user control.

Referenced in this episode:

Companies/Organizations:

DriveNets
Artificial Intelligence Risk, Inc.
RCOM
NVIDIA
AMD
Google
AWS
Intel
DarwinAI
Apple

Podcasts:

Movies:

Snow White and the Seven Dwarfs
Terminator