

Anil Ananthaswamy, "Why Machines Learn: The Elegant Maths Behind Modern AI" (Dutton, 2024)
Jul 30, 2025
Anil Ananthaswamy, an award-winning science writer and author, delves into the intricate mathematics behind modern AI. He reveals how machine learning influences critical decisions in our lives, powered by concepts as old as linear algebra. Ananthaswamy explores the journey from simple algorithms to complex neural networks, diving into Hebbian learning and the evolution of generative AI. He also ponders the connection between human cognition and AI, emphasizing the need to grasp the math that underlies these advancements for safe and effective use.
AI Snips
Chapters
Books
Transcript
Episode notes
Learning By Relearning The Math
- Anil learned machine learning by returning to coursework and coding, which revealed beauty in the underlying math.
- He wrote the book to explain why algorithms work by exposing the math behind them.
Math Explains Model Limitations
- Limitations of AI differ by algorithm and are explained by the math behind them.
- Understanding the math clarifies why models like nearest neighbors or deep networks fail or hallucinate.
Use Repetition To Cement Concepts
- Repeat complex concepts in different guises to reinforce understanding.
- Present ideas multiple times to mimic human learning and increase retention.