Episode Summary
Executive Summary: Patrick O’Shaughnessy interviews Tim Urban (with Andrew Finn) about human progress, AI, first-principles thinking, and content creation. Urban argues that compounding knowledge has created a 'human colossus,' that AI will move from narrow to general intelligence, and that reasoning like a chef—not a cook—is the key to originality and resilience.
Main Topics: Human colossus and compounding progress: Humanity's collective knowledge compounds into a force far beyond any individual mind. Planet metaphor and progress acceleration: Four 'planets' illustrate how identical humans produce radically different worlds over time. AI from narrow to general intelligence: Urban explains ANI today and argues breadth will expand into general intelligence. Cook vs. chef and first principles: Original breakthroughs come from reasoning from first principles, not copying recipes. Elon Musk as a systems thinker: Musk's companies share a pattern: mission first, then a business model that funds innovation. Neuralink and human-AI integration: Neuralink is framed as a way to align humans with AI rather than be surpassed by it. Writing process and memorable explanation: Urban’s creative process is research-heavy, then structured into metaphors and diagrams.
Key Arguments: Humanity today differs from 50,000 BC because the 'human colossus' has compounded knowledge and capability. AI's narrow wins in chess, Go, poker, and mammograms suggest broader intelligence may follow. Reasoning from first principles beats reasoning by analogy when the world changes quickly. Most people are trained to be 'cooks'; originality comes from becoming a 'chef' who writes the recipe. Musk's companies reverse-engineer a future goal, then build a business model to fund the innovation. Neuralink aims to let humans keep pace with AI by creating high-bandwidth brain-machine interfaces. Content becomes powerful when it helps readers move from partial understanding to durable mental models.
Data Points: Interview original air date: Tuesday, October 17th, 2017 - Podcast replay introduction Wait But Why post length on Neuralink: 40,000 words - Patrick describes Tim Urban’s Neuralink essay George Washington time gap: 250, 270 years ago - Used to illustrate how much progress has occurred DPU definition: die progress unit - How far into the future one must go to feel shock from progress Mammogram AI accuracy: 99.6% - Example of narrow AI outperforming doctors Doctor mammogram accuracy: 80% - Comparison benchmark for medical analysis Chess AI milestone: 1996 - Deep Blue beat Garry Kasparov Go branching factor after two moves: 350,000 - Illustrates why brute force fails in Go Go board configurations vs atoms: more than atoms in the universe - Shows the scale of Go complexity Poker AI milestone: first seven days - Libratus eventually never lost again after one week Human brain power usage: 20 Watts - Compared with computer systems for brain-machine interfaces Equivalent computer power: 20 megawatts - Rough comparison to the brain's efficiency Humanity's recent history scale: 50,000 BC to today - Repeated reference point for the compounding metaphor Tesla production benchmark: 10 million cars a year - Referenced when discussing pressure on legacy automakers Website readership: 14 million uniques in a month - Patrick frames the scale of Wait But Why's audience Typical top post readership: over a million people - Urban describes recurring reach for his posts Common long-term reach: 4 or 5 million - Urban says evergreen posts often accumulate this many readers Brain-machine interface resolution: one or 200 neurons - Current crude interfaces can only handle small numbers Potential future interface scale: millions of neurons - Needed for rich communication and thought-level bandwidth
Pivotal Quotes: "The difference is what I call the human colossus" — Tim Urban: Explaining why today’s world is radically different from 50,000 BC "I think he would die from the level of progress that's been made" — Tim Urban: On how George Washington would react to modern life "What's your Bible? What's your recipe in all of these different areas?" — Tim Urban: Describing when people rely on inherited assumptions instead of first principles
Implications: Listeners are left with a challenge: audit their own 'recipes,' question inherited assumptions, and think in systems that stay useful as AI and compounding progress accelerate.
From the Episode
50,000 BC and today had the same exact species with the same exact level of intelligence. So, what's going on? What's the difference? Not the most intelligent species on the planet, because that's the same for planet two and three. The difference is what I call the human colossus, which is like the collective knowledge that humanity has built, the collective kind of ability to create and invent, and it builds upon itself. And it's this kind of mountain of achievement that's this. Cumulative thing that no one human or no one generation could do on their own. Even something like the match, you think, oh, the match. No one person invented the match. If you look at the history of the match, there's like 10 different people over a span of like 400 years that each had a critical innovation that it was like this over time collaboration that led to the match. And you look at everything around you, and no human's smart enough to create this. No one even knows how to make a pencil on their own. No one knows how to do all the different parts that create a pencil on their own. So I call this the human colossus. It's this kind of force.
A horse. You just think about the world they lived in, right? And then you come here and you see telephone, the iPhone, you FaceTime with someone across the world. You play music for him that was recorded 50 years ago. You show him a baseball game that was currently being played a thousand miles away. You show him the International Space Station in the sky. You show him the internet. You show him, you tell him that people walked on the moon. I mean, I always say that he, it's not just he would be mind-blown or shocked. He would die. I think he would die from the level of progress that's been made of his shock. So then I get to this. Concept, this is what I call a die progress unit, a DPU. How far do you have to go in the future to die from the level of progress? So it's that George Washington today level. But then what's interesting is if George Washington went back the same amount of time, 250 years or so, he's not going to be able to make Leonardo da Vinci die by coming from 1500 to 70. It'll be cool. But remember, he's still coming to a world where there's no power and people are still on horses. And they have a cooler world map and they have the telescope. So they have some mind-blowing things.
Who said, Well, we don't know anything, so let's just observe and see what we find. And they learned that the Earth was 5 billion years old. But then there were a lot of people who, I called them, you know, these are the flood geologists who said, Well, we know the Bible says the Earth is 6,000 years old, so we have to reconcile with that. And they ended up not being able to find truth because their Bible got in the way. They didn't have the ability to reason from what they observed. So the question is, what's your Bible, or what's your recipe in all of these different areas? And what Elon is so amazing at is just he looks at something like SpaceX before he starts. And says, well, he does some calculations. It says, I think we can learn how to land a rocket. I think we can reduce the cost. It seems like they're spending way too much. It seems like this is something a private company could do. And every part of conventional wisdom and all of his friends are saying, don't do it. Every billionaire that's tried has failed. You're going to lose all your money. What are you doing? 99.9% of humans are not wired to overpower that, to overcome that. We are wired to say, yeah, you know what? And if no one else is doing it, I'm sure there's a reason. Or if it were a great idea, it would already be done. Or whatever, you know, this is the kind of things people do when they start a business. They think that way. Elon said, well,
About Invest Like the Best with Patrick O'Shaughnessy
Conversations with the best investors and business builders in the world.
View all episodes from Invest Like the Best with Patrick O'Shaughnessy