Charlie Munger lives on in LLMs

I never got to meet Charlie Munger in person, but I’ve sought to channel his wisdom for decades.

Munger was content to remain in the shadows compared with his more famous partner, Warren Buffett. Together they ran Berkshire Hathaway and made countless investors wealthy.

I was one of the people who benefited, having owned shares since 2008. 

In 2010, I tried to attend the annual Berkshire shareholder meeting in Omaha but ran into difficulties getting the tourist visa I needed to travel from London.

The folks in the American embassy found it impossible to believe anyone would travel to Omaha, Nebraska on holiday. They started a long background check and I never made it.

Unable to see Munger in person, I devoured his writings. I bought two autographed copies of Poor Charlie’s Almanack, a compilation of Charlie’s speeches compiled by Peter Kaufman. I gave one to a friend who later founded a hedge fund in Hong Kong based on the philosophy of value investing. The other copy still sits in my library.

First published in 2005, the book remains as relevant today as ever. Munger’s multidisciplinary thinking and intellectual rigour haven’t aged a day. The book isn’t just about investing. It’s about learning how to think, how to reason, and how to understand the world.

Most of Munger’s famous quotes – even when they are about making money – transcend investing and seem just as applicable to other pursuits in life. For example: “The big money is not in the buying and selling, but in the waiting." or my favourite quote “The best way to get what you want is to deserve what you want.”

Today we are wrestling with generative AI, a transformational technology. Since Munger passed away in November 2023 at the age of 99 we will never know his thoughts about it.

He was famously skeptical and cautious about new technologies, often stressing that “strange things can happen” when change occurs so quickly.

But I tend to think he would have found large language models useful. LLMs have already surpassed humans in many benchmarks. They are, by design, multidisciplinary, trained on vast amounts of knowledge in multiple languages. Munger would have appreciated the breadth.

He often talked about the psychology of human misjudgment. His list of 25 standard causes of human errors was clever and counterintuitive. He knew the consequence of combining these biases to produce what he called a “lollapalooza” effect.

In a way, LLMs sidestep these pitfalls. They have no hidden motives, no self-interest, no job worries. They don’t get offended. They don’t scheme. They are patient, objective critics. With proper guidance, a LLM can help us identify our blind spots, much like Munger did.

LLMs may provide us with the kind of straight-thinking advice Munger would have given.

For someone often labeled an anti-tech investor, Munger’s spirit endures in LLMs. 

I think he would appreciate the irony.