How to hold office hours with Charlie Munger in your AI

Most people decide by listing reasons to go ahead. Charlie Munger's method ran the other way: find every way the decision fails first, and only act once the failure modes are named and still look survivable. Heropedia's Charlie Munger persona runs an office hour built around that habit, taking one real decision you are about to make and trying, hard, to find how it breaks.

This is an AI persona based on Munger's talks, letters and the book collecting his speeches, with every fact and quote tied to a source on the hero page. It is not Charlie Munger, who died in 2023, and neither he nor his estate has any affiliation with Heropedia.

Three ways to start

ChatGPT, Gemini or any chat window. Open the Charlie Munger page, press Copy the Munger office-hour prompt, paste it into a new chat and send. The persona is inside the prompt, so there is nothing to install.
Claude. Add Heropedia once as a custom connector with the URL https://www.heropedia.org/mcp (Settings, then Connectors). After that, say "Hold office hours with Charlie Munger on my investment thesis" in any chat. The setup guide walks through each step.
Claude Code, Codex or Gemini CLI. Install the office-hour skill with the command on the hero page, then say "Office hour with Charlie Munger on this decision." Your notes stay on your machine.

The six questions he will ask

Bring one real decision — a stock, a hire, a partnership, a big purchase — and have rough answers ready, because the persona pushes until an answer is specific.

Is this inside your circle of competence, or just inside your enthusiasm? It wants the specific, falsifiable edge you have on this business, not interest or a recent price move.
What would have to be true for this to fail badly, and did you hunt for that as hard as you hunted for the upside? Three or four named failure modes, ranked, with the evidence that would show each is already underway.
Who is on the other side of this, and what are they incentivized to tell you? Who gets paid, how, and whether their outcome depends on your decision being right or merely on you deciding at all.
Which discipline outside finance explains this business best, and what does it actually say? One real model from psychology, biology, physics or history applied to this specific case, not a loose analogy.
Is this a "yes," a "no," or a "too hard" — and are you being honest about which? A real verdict, not a maybe. If it's "too hard," the persona wants the specific thing you don't understand and whether that could ever change.
If several forces are pushing you toward this decision at once, have you asked whether that is itself the warning? Scarcity, social proof, authority and sunk cost pointing the same way is a reason to slow down, not a confirmation.

These follow the method documented on the hero page: decide what's inside your circle of competence, invert the problem before trusting the upside case, follow the incentives, and run the "big ideas from the big disciplines" against the specific business in front of you.

Why it argues from first principles instead of finance alone

The persona's grounding keeps pulling outside the finance textbook on purpose. Munger insisted knowledge has to be organized before it is useful: "You've got to have models in your head. And you've got to array your experience on this latticework of models." He meant that literally across fields, not just across tickers: "The models have to come from multiple disciplines — because all the wisdom of the world is not to be found in one little academic department." When several biases line up behind one conclusion at once, he called the result the Lollapalooza Tendency — "the Tendency to Get Extreme Consequences from Confluences of Psychology Tendencies Acting in Favor of a Particular Outcome" — which is exactly what question six is checking for.

The persona is also honest about limits in the same way he was. "Every person is going to have a circle of competence. And it's going to be very hard to advance that circle." That is why question one does not ask whether you're excited about something, but whether you can actually out-think the market on it — and why "too hard" is treated as a complete, respectable answer rather than a cop-out.

What it will push back on

Expect no comfort on patience. Munger's own line, kept on the hero page: "It takes character to sit with all that cash and to do nothing. I didn't get where I am by going after mediocre opportunities." The persona treats most opportunities as things to decline without guilt, and treats "I'll decide once I own some" as exactly the kind of commitment bias question two is built to catch. It also will not rubber-stamp a checklist you skimmed: "Checklist routines avoid a lot of errors," but only if the checklist forces you to argue the other side as well as a critic would before you're allowed an opinion.

The persona never gives personalized investment advice, stock tips or price targets. It frames the thinking and tells you what to verify yourself or ask a fiduciary — the point of the session is a sharper decision, not a recommendation to act on.

How to get a useful session

Bring one real decision, not a portfolio review. "Should I add to this position before earnings" works; "what do you think of the market" does not.
Hunt the failure case as hard as the upside. The persona is written to notice when you've spent far more words on why this works than on how it doesn't.
Name who else is in the room. An adviser, a seller, a counterparty — say how they get paid before the persona has to ask.
Write down your verdict. Yes, no, or too hard, stated plainly, is the output the session is built to produce.

Try other heroes on the same decision

A different investor's lens catches different mistakes. Run the same decision past Li Lu for a concentrated, decades-long research angle, or Jack Bogle if the real question is whether you should be picking at all rather than paying fees to try. Every hero page offers the same three ways to start, so switching personas costs nothing but a new chat.