Money For the Rest of Us Don’t Take Financial Advice from AI
Oct 22, 2025
The discussion highlights how AI can misinterpret basic financial concepts, leading to common mistakes, such as overlooking the time value of money. A real-world example reveals how AI can miscalculate loan evaluations. The risks of relying solely on AI for significant decisions, like military modeling and personal finance, are addressed. Listeners are encouraged to use AI as a supplement to human expertise and to always fact-check its outputs. Additionally, the podcast explores opportunity costs, sunk costs, and practical tools for making informed financial choices.
AI Snips
Chapters
Transcript
Episode notes
Confident Answers Can Miss Core Finance
- AI can produce confident, detailed financial answers that still miss key concepts like present value.
- The nominal interest total isn't the same as present-value cost when comparing loans.
ChatGPT5 Flip-Flops On A Loan Example
- J. David Stein walked ChatGPT-5 through a car-loan example and found it first favored the one-year loan then reversed after present-value framing.
- The model produced inconsistent break-even rates and unit conversion errors until checked by a human.
Small Math Errors Become Big Misleading Claims
- Numerical errors and unit mistakes (monthly vs annual) can cause large AI mistakes in finance models.
- The AI converted a monthly rate to an annual rate incorrectly, producing a misleading break-even number.