You can now borrow the decision-making power of a billionaire investor to solve your daily problems without reading a library of books.
Charlie Munger, the late vice chairman of Berkshire Hathaway, was famous for his latticework of mental models. He believed that to solve complex problems, you need to look at them through the lenses of multiple disciplines: physics, biology, psychology, and economics, simultaneously. I just found an incredible guide by an observant Reddit user who realized that AI is the perfect tool to simulate this exact thought process. This savvy professional converted Munger’s most famous mental models into specific prompts that force ChatGPT to think like a multidisciplinary investor.
The “Latticework” Logic
The core concept here is that most of us try to solve problems using only the tools we are most comfortable with. If you are a hammer, everything looks like a nail. Munger avoided this by collecting big ideas from every field of study. The creator of this framework points out that while it took Munger decades to learn these models, AI has already processed all of them. You simply need to ask the AI to access that specific data.
By using this approach, you aren’t just asking for advice; you are running a stress test. You are asking the AI to pull from hard sciences and soft skills at the same time to validate your thinking. The original poster explains that when you ask, What mental models apply here? regarding a subscription product, the AI instantly considers churn physics (math), sunk cost fallacy (psychology), and network effects (economics). It creates a much richer, 360-degree view of your situation than a standard question ever could.
💡 Inversion and Competence Checks
The first major insight from this framework is the power of Inversion. Humans are naturally optimistic when planning projects, which often blinds us to obvious risks. The expert highlights Munger’s favorite math trick: Invert, always invert. instead of asking how to achieve success, you ask the AI to describe exactly how to fail. If you ask, What would guarantee my marketing agency fails? the AI maps out every landmine you need to avoid.
Another critical element is the Circle of Competence. In the information age, it is easy to feel like an expert after reading a few headlines. This framework uses a specific prompt to fight that arrogance. By asking, What’s my circle of competence here and where does it end? you force the AI to draw a sharp line between what you actually know and what you are merely pretending to understand. This is particularly useful for investment decisions or entering new industries where a lack of deep knowledge can be expensive.
💡 Hunting for Bias and Hidden Assumptions
The second key takeaway is how this method exposes the invisible forces that shape our bad decisions. The Reddit user suggests using Munger’s focus on incentive-caused bias. We often take advice from people without fully analyzing their motivations. By prompting the AI with Where’s the incentive-caused bias? regarding a financial advisor’s recommendation, you get a clear picture of whose interests are actually being served.
Furthermore, every plan relies on things going right, but we rarely list those dependencies explicitly. The prompt What would need to be true for this to work? acts as a truth serum for your strategy. The author uses the example of betting on AI replacing lawyers. When you run this prompt, the AI lists out fifteen different legal, cultural, and technological dependencies you probably hadn’t considered. It moves you from magical thinking to logistical planning instantly.
💡 The Reality Check of Base Rates
The final and perhaps most powerful insight is using AI to check Base Rates. This is a statistical concept Munger loved. When we have a new business idea, like opening a restaurant, we tend to focus on our unique passion and the delicious food we plan to serve. We ignore the general statistics of the industry.
The innovator behind this post suggests asking, What is the base rate here? This commands the AI to look at the historical success and failure rates for your specific endeavor. It gives you the cold, hard survival statistics before you invest your life savings. It anchors your dreams in mathematical reality. The author also notes a smart safety measure: AI can sometimes hallucinate or force a model where it doesn’t fit. They recommend always following up with a check: Does that mental model actually fit, or are you forcing it? ensuring the advice remains high-quality.
📌 Prompt of the Day: The Munger Stack
Here are the specific prompts the original creator developed. You can use them individually or chain them together for a deep analysis of any major decision.
1. The Multidisciplinary Scan
“I am [describe situation/problem]. What mental models apply here?”
2. The Inversion Technique
“I want to [goal]. Invert, always invert. What would guarantee failure?”
3. The Knowledge Check
“I am considering [action]. What is my circle of competence here and where does it end?”
4. The Bias Hunter
“I am analyzing [situation]. Where is the incentive-caused bias?”
5. The Reality Anchor
“I want to start [project]. What is the base rate here?”
6. The Premortem
“What would need to be true for this to work?”
If you want to dive deeper into the full list of mental models and see more examples from the source, check out the link below!
💡 FAQ & Troubleshooting
How can I ensure the AI isn’t just making up connections between models?
AI occasionally misapplies mental models from the wrong domain to fit a request. If a connection feels weak, explicitly push back by asking, “Does that mental model actually fit, or are you forcing it?” This prompt forces the AI to re-evaluate its logic and keep the reasoning honest.
Can I combine these prompts for a deeper analysis?
Yes, you should chain the models to create a “compound effect.” Start by asking which models apply, then ask the AI to “invert” the problem, identify “incentive-caused bias,” and finally list “what would need to be true” for success. Each layer often reveals blind spots that the previous prompt missed.
How do I use this framework to spot my own blind spots?
Use the specific prompt: “What am I not seeing because of [bias]?” You can replace the bracket with terms like confirmation bias, availability bias, or authority bias to get an instant reality check on ideas you might be over-optimistic about.
What is the best way to get a realistic probability of success for a new idea?
Utilize Munger’s concept of statistical thinking by asking, “What’s the base rate here?” This prompts the AI to provide historical data and survival statistics (e.g., the actual failure rate of restaurants) rather than just encouraging your aspirations.
I converted Charlie Munger’s mental models into AI prompts and now I think like a multidisciplinary investor
byu/EQ4C in