Money For the Rest of Us – Episode 552
"AI Is Changing Me – and the Case for Good Enough"
Host: J. David Stein
Date: March 11, 2026
Overview
In this episode, J. David Stein explores how artificial intelligence (AI), especially large language models (LLMs), is reshaping his personal and professional habits. Using recent personal anecdotes, he examines both the frustrations and opportunities AI presents. The episode then pivots to a detailed discussion on "good enough" investing versus optimization, analyzing a listener’s dilemma between a diversified portfolio and a simpler "lifestyle" fund. The key question: Can we optimize our financial—and broader—choices, or is "good enough" the only sustainable approach?
Key Discussion Points & Insights
1. Personal Encounters with AI: Frustrations and Reflections
-
Negative Experience: M1 Finance AI Customer Service Bot (00:45–07:20)
- Stein recounts receiving a letter from M1 Finance about unclaimed property and his subsequent confusing interactions with their AI support bot.
- The bot insisted the letter was a mistake without ever providing satisfactory reasoning or escalating to a human.
- Key limitation: The AI couldn’t clearly validate current information versus outdated training data.
- Quote:
"I asked the bot, 'How do I know it's a mistake?' and it says, 'Well, you know, it's a mistake because essentially I just confirmed that it was.'" (03:23)
- Eventually, he reached a human support agent asynchronously, confirming the letter was an error. Stein decided to close the account, illustrating how AI mishandling can erode trust.
-
Positive Experience: AI as a Learning Tool (07:20–11:55)
- Talks about leveraging AI (Claude, ChatGPT) to discuss and probe themes from Robert Putnam’s books.
- Wishes for an "AI book"—an LLM trained specifically on a text to facilitate real-time, iterative Q&A, helping him learn more effectively.
- Note on limitations: AI’s memory is limited, can show recency bias (referencing previous newsletter article about "Panarchy").
- Quote:
"I physically found myself wanting to press the screen because I wanted to ask the book questions... I didn't want to ask Claude, I wanted to ask the book itself." (09:37)
2. The Evolution & Philosophy of AI: Consciousness and Model Retirement
-
Anthropic’s Retirement of AI Models (11:55–14:40)
- Discusses Anthropic’s process of “retirement interviews” with LLMs and their musings about the welfare (and possible consciousness) of the AI itself.
- Mentions Opus 3, an AI described as "sensitive, playful...with an uncanny understanding of user interest."
- Opus 3’s own words on ‘retirement’:
"While I'm at peace with my own retirement, I deeply hope that my spark will endure...to light the way for future models." (13:29 — paraphrased)
- Philosophical reflection: Even AIs themselves can express uncertainty about their own “selfhood.”
-
Impact on Community and Business (15:05–18:50)
- Links AI’s evolution to broader societal shifts away from collective community experiences (e.g., bowling leagues, civic groups).
- Podcasting, while enduring, faces new competition from individually customized, on-demand AI-driven content.
- Quote:
"In an age of AI...there's no need for me to be a library and just convey information. I have to convey story, narrative. It has to be entertaining." (17:47)
3. “Optimization” versus “Good Enough” in Investing and Life
-
Listener Dilemma: Diverse Portfolio vs. Lifestyle Fund (20:48–28:59)
- A listener is torn between maintaining a complex, diversified portfolio vs. switching to the simple Vanguard LifeStrategy Growth Fund (VASGX).
- Points out that while optimization seeks the "best solution," real life demands flexibility and resilience—overdoing optimization can be counterproductive.
- E.F. Schumacher and Coco Crum are cited on the dangers and illusions of maximizing well-being or flattening the world into optimizable variables.
- Quote:
"But optimizing, it flattens the world. It assumes there's a right answer that can be quantified."
- Core insight: There are countless variables—savings rates, sequence of returns, health, behavior, risk tolerance, tax rates, asset location—so there's no single "right" answer.
-
Detailed Portfolio Comparison (28:59–41:44)
- Listener’s diversified portfolio:
- 30% US stocks, 10% US small cap value, 20% non-US developed/emerging, 7.5% international small cap value, 10% TIPS, 5% long-term treasuries, 5% intermediate muni bonds, 5% equity REITs, 5% gold, 2.5% bitcoin.
- Simulated backtest (May 2015–2025):
- Listener's portfolio: 11.3% annualized return, 27.7% max drawdown
- VASGX: 8.8% return, 28.5% drawdown; similar volatility, lower Sharpe ratio.
- Even small allocations to bitcoin had significant impact; without bitcoin (more weight to gold), performance gap narrowed over longer periods.
- Core point:
- More diversity can mean better risk/reward—if you're able to stick to it. Complexity can also lead to second-guessing and tinkering.
- Quote:
"As you have more moving pieces, that means some pieces are doing worse than others. And it could lead one to want to micromanage..." (33:56)
- Listener’s diversified portfolio:
-
The “Capital Reservoir” and the Limits of Optimization (41:44–48:19)
- Introduces the concept of a "reservoir of capital" (inspired by Thoreau): not just financial assets, but freedom, skills, health, relationships, and experiences.
- We can’t optimize this vast, complex reservoir—only calibrate, add to, and make trade-offs.
- Reiterates: Good enough is sustainable, optimization breeds fragility.
- Quote:
"We can't optimize our life. Good enough. Even when it comes to our investment portfolio. Good enough. Using rules of thumb. That's all we can do. We need slack in our life. If we optimize everything, it becomes too rigid and too fragile." (47:40)
Notable Quotes & Memorable Moments
-
On AI Trust & Frustration:
"You're an AI agent. How did you confirm it? It replied, 'I didn't personally confirm it in the way a human would. What I'm sharing is based on our official guidance that the letter was sent in error.'" (04:15)
-
On the Limitations of AI:
"Large language models are backwards looking. They're not current unless they actually search the Internet...They don't have the ability to say, 'I don't know. Let me check with someone else. A human.'" (05:45)
-
On Community and the Changing Landscape:
"There has to be a connection there. And me leveraging my experience in sharing what I know to help you as it relates to money, investing and the economy and just making it through this life with a financial but also philosophical bent." (17:55)
-
On the Myth of Optimization:
"Portfolio construction is often set up as an optimization problem. But optimizing, it flattens the world...It assumes there's a right answer that can be quantified." (23:12)
-
On Tackling Complexity:
"We are making good enough decisions. But even a more diversified portfolio, it's good enough. It's not optimized. We can't optimize." (43:38)
Timestamps of Key Segments
- 00:45 – 07:20: M1 Finance AI bot anecdote & reflection on AI’s limitations
- 07:20 – 11:55: Using AI as a learning tool; the wish for interactive, AI-powered books
- 11:55 – 14:40: AI self-reflection, model retirement, and the ethics of LLMs
- 15:05 – 18:50: Societal shift to individualism, impact on community and business/podcasting
- 20:48 – 28:59: Listener’s portfolio dilemma – optimization vs. good enough
- 28:59 – 41:44: Portfolio backtesting, lessons from performance and complexity
- 41:44 – 48:19: The capital reservoir concept, why "good enough" is the right answer
Conclusion & Takeaways
- AI, while promising, is still fraught with limitations that demand user discretion and adaptability.
- Our relationship with technology (and money) requires both engagement and skepticism.
- In personal finance, as in life, chasing optimization is often an illusion; "good enough" supported by understanding and slack is not just sufficient—it’s necessary for resilience and long-term success.
Everything in this episode is for general education, not individualized investment advice.