Planck Knowledge vs. Chauffeur Knowledge
Who is in danger of being replaced by AI? An old Charlie Munger story might provide a hint.
This story, aside from being very amusing, may also identify who might be replaced by AI:
I frequently tell the apocryphal story about how Max Planck, after he won the Nobel Prize, went around Germany giving the same standard lecture on the new quantum mechanics. Over time, his chauffeur memorized the lecture and said, “Would you mind, Professor Planck, because it’s so boring to stay in our routine, if I gave the lecture in Munich and you just sat in front wearing my chauffeur’s hat?” Planck said, “Why not?” And the chauffeur got up and gave this long lecture on quantum mechanics, after which a physics professor stood up and asked a perfectly ghastly question. The speaker said, “Well, I’m surprised that in an advanced city like Munich I get such an elementary question. I’m going to ask my chauffeur to reply.”
- Charlie Munger, 2007 USC Law Commencement Address
Charlie’s takeaway from this story is that there are two kinds of knowledge in the world: Planck Knowledge, held by people who really know. “They’ve paid the dues, they have the aptitude.” Then, there’s Chauffeur Knowledge. Those people know how to talk about the subject in a way that sounds intelligent – they may seem impressive, they may have charisma – but in the end, what they’ve got is chauffeur knowledge masquerading as real knowledge.
This is such a useful distinction!
Planck Knowledge is actually understanding something from first principles. Being able to dissect it, develop it, apply it, build upon it. Chauffeur Knowledge is simply memorizing. Repeating. Imitating. Just like the chauffeur did in Munich. He could very well give the lecture, but it required Max Planck himself to answer the follow-up question.
This story describes something I’ve been encountering throughout my career; there was a certain type of person with whom I disliked collaborating. I would cringe at the realization that a manager or a peer of mine is just repeating things that they’ve heard, about agile development, or NoSQL databases, or payment systems. Despite being central to their job, they have never bothered learning how these things worked, and yet - would be confident enough to express their (false) assertions in meetings / emails / documents. Instead of reducing chaos and driving the project forward, they would effectively introduce more confusion and friction. The worst part for me was that they would often get recognized and promoted. Because they do sound smart! But – and now I have the words to describe it – they were relying on Chauffeur Knowledge.
This is exactly the kind of behavior most threatened by AI.
The Planck vs. Chauffeur distinction is also a useful way to describe my contention with AI chatbots. Yes, it’s totally incredible that they can instantly spit out impressive answers for any topic; this is incredibly helpful when I don’t know the first thing about a subject, and don't have the time or willingness to do the work and acquire Chauffeur Knowledge.
That’s the thing though.
AI is absolutely excellent at Chauffeur Knowledge.
When I try to acquire Planck Knowledge on a topic, however, I find AI tools to be a terrible way to do it. Pretty quickly, I start noticing inaccuracies and hallucinations, which make me question everything else. I always end up reverting to the pre-AI ways of reading complete textbooks, going through the full data, listening to all the interviews, taking notes, and drawing my own conclusions. Sure, AI may help in speeding up parts of the Planck Knowledge accumulation process, but it certainly can’t perform it on its own. And it might never could.
This leads to the inevitable conclusion that, folks whose career is relying on Chauffeur Knowledge, should start worrying. Being able to sound smart without understanding what one is talking about, might no longer cut it. AI would likely outperform there.
But this may be terrific news for those of us who appreciate true Planck Knowledge. This could become more wildly differentiated, in a world flooded with Chauffeur Knowledge and AI hallucinations.
What a wonderful outcome it would be if, just like AI drove Lee Sedol to push the boundaries of his creativity on the game of Go, we would all be encouraged to focus on genuine Planck Knowledge.


Note-takers will fade into irrelevance; note-makers will rise to dominance.
Great piece, Matan.
Nice one, Matan. What a cool way to connect Munger's story to AI and its (in)capabilities.