AI Snake Oil

Outline

AI Snake Oil (2024) helps readers tell the difference between artificial intelligence that works and AI that does not. Arvind Narayanan and Sayash Kapoor explain why generative AI can be genuinely useful, while predictive AI used to judge people in hiring, policing, or health care often fails. The book offers clear tools for spotting exaggerated claims and thinking critically about AI in everyday life.

• Not all artificial intelligence is the same, and the book separates AI that genuinely works from AI that does not.
• Generative AI can be genuinely useful, according to the authors.
• Predictive AI used to judge people in hiring, policing, or health care often fails.
• Exaggerated claims, flawed science, and misleading marketing all feed the hype around AI.
• The book gives clear tools for spotting inflated claims and thinking critically about AI in everyday life.

Arvind Narayanan and Sayash Kapoor are computer scientists at Princeton University who study the real-world impact of artificial intelligence. Narayanan is a professor of computer science and director of Princeton’s Center for Information Technology Policy. Kapoor is a researcher there, and together they became known for their widely read newsletter that also carries the title AI Snake Oil.