Grant Sanderson of 3Blue1Brown on Notation, Infinity and How to Really Learn Math

High-Level Overview

Lex Fridman sits down with Grant Sanderson, the math educator behind the YouTube channel 3Blue1Brown, known for its programmatically animated explanations of linear algebra, calculus and other fields. The conversation starts far from the classroom, with whether an alien civilization would do mathematics the way we do, and uses that question to get at something Sanderson cares about a great deal: how much the way we write and introduce an idea shapes whether anyone can understand it. From there the two range over whether math is discovered or invented, why the laws of physics are simple enough to write down, the simulation argument, and whether infinity is something a person can be comfortable with.

The second half is more personal and practical. Sanderson describes the formula that first made him fall for mathematics, his favorite video to make, how he writes a script, who he imagines he is explaining things to, and his advice for anyone trying to learn math. The tone is thoughtful, playful and often funny, with Fridman happily playing devil's advocate and Sanderson gently teasing him about mortality. The conversation flows from philosophy of mathematics, to the craft of explanation, to a short and warm personal close.

The Discussion Explores:

• Whether intelligent aliens would share our mathematics, and how much notation and mode of thought steer what math becomes
• Why Sanderson thinks the usual way of writing the exponential function hides what Euler's famous formula is really saying
• Math as a cycle of discovery and invention, illustrated with the Pythagorean theorem
• Three kinds of mathematician, and why they disagree about how math relates to physics
• Why reality seems compressible into simple equations, and Sanderson's doubts about the simulation argument
• Infinity as an abstraction, and why examples should come before definitions
• The most beautiful idea Sanderson has met, his favorite video, and how his scripts get written
• How to actually learn math: do problems, program, and teach

Deep Dive into Main Ideas

Would aliens do math our way?

Fridman opens by asking whether an intelligent civilization elsewhere would have different mathematics. Sanderson thinks it would probably be very different, starting with the obvious: the notation. He expects they would count, since the natural numbers and the idea of repetition lead straight to addition and multiplication. But past that point he sees choices. The way we extend to the real numbers is one of them; he mentions the surreal numbers as a distinct system that also captures continuity and could, in principle, sit behind a description of planetary motion just as well.

His larger point is that the choices different human mathematicians have made may only scratch the surface of what a completely different mode of thought, or mode of interacting with the universe, would produce. Notation is the part you would notice first, he says, but the mode of thought matters more. Still, he argues notation carries more weight in how we think than we usually give it credit for.

The trouble with e to the x

Asked for a least favorite piece of notation, Sanderson picks the way we write the exponential function. Writing it as e to the x tells you to think of a particular constant multiplied by itself repeatedly. That is fine for population growth or compound interest, but it becomes strained for the square root of two and badly strained for e to the pi i. He would rather introduce the exponential as a function, defined for instance by the simple differential equation it solves: it describes things whose rate of change depends linearly on their own value.

Seen that way, the complex case is not mysterious. He compares it to a tetherball whose velocity is always perpendicular to its position: the rate of change is still tied to where the thing is, but by a 90 degree rotation instead of a scaling. So e to the pi i says: pick an origin, always move perpendicular to the line from the origin to you, and after walking pi times the radius you are halfway around a circle. He cannot guess how many hours scientists, engineers and students have lost trying to twist repeated multiplication into that picture.

When Fridman calls it the most beautiful equation in mathematics, Sanderson pushes back. "I think the notation makes it mysterious." He finds the underlying idea quite pretty, but not the most beautiful thing in the world. In his telling, e and the circle constant (he prefers tau, the period of the function in the imaginary direction) belong to the same central function for nearly orthogonal reasons, and the famous equation presses them closer together than they naturally sit. He likens it to putting all the children in one bed: "there is a family resemblance, but it's not that close." Later he goes further and calls the equation, as written, one of the most hideous, because it bundles three confusing conventions: e, pi instead of tau, and the word imaginary. He also notes the small irony of using a constant to stand for calculus, the study of change.

Discovered or invented?

Sanderson's answer is that it is a cycle. Discoveries about the world tell you which math will be useful, and the math itself is, in a sense, invented. The Pythagorean theorem is his example. Historically, and in its geometric proofs, it feels like a discovery. But look at how two-dimensional space is formalized as pairs of real numbers with a distance defined on them, and the theorem starts to look like a definition: we defined distance so that it would be true. What actually happened, he says, is that a genuine physical discovery informed which metric to choose. Other metrics give consistent math, but not math that applies to the world we study.

From there the abstract object is studied for its own sake, which brings new inventions and mysteries, which may yield new discoveries. When Fridman suggests that modern mathematics has drifted beyond what the mind can comprehend, with dimensions beyond three or four, Sanderson disagrees. Higher dimensions can be highly applicable to a three-dimensional universe, he says, because of state spaces; a five-dimensional manifold does not need the physical world to be five-dimensional to be useful.

Physics, math, and three kinds of mathematician

On the difference between physics and math, Sanderson describes math as the study of abstractions over patterns and pure patterns in logic, while physics is grounded in wanting to understand the world we live in. He says the answer depends heavily on who you ask, and sketches three veins of motivation. Some mathematicians love pure puzzles and are drawn to things like combinatorics. Some are physically motivated, with chaos theory as an example of purely mathematical statements driven by applications. And some simply love abstraction and generality: the people who invented topology and later got into category theory.

He cites Vladimir Arnold, who would say "math is a branch of physics", against the category theorists, for whom usefulness in describing the world is more of a happy coincidence. That coincidence is the surprising part: that the study of pure puzzles and abstraction also describes the fundamentals of quarks.

Why is the universe compressible, and is it a simulation?

Fridman asks why reality fits into clean, relatively simple equations. Sanderson offers a filter effect: perhaps physicists only find interesting the things simple enough to describe mathematically, and anything too complex gets called biology. He admits he has a hard time imagining what intrinsically complicated fundamental laws would even look like, and wonders whether a world with incompressible laws could host beings able to ask the question, something like an anthropic principle. He is candid that he does not know. Fridman raises the idea that our evolved brains may only perceive the compressible parts; Sanderson answers that we can fly spaceships, which is a reality check that our models are not too far afield.

On the simulation hypothesis, a listener question, Sanderson says it is conceivable but he does not buy the usual argument that nested simulations make it likely we are in one. Simulating a universe like ours takes enormous resources, so any bound on the top-level universe's information processing limits the number of layers. He thinks people give too high a prior to an unbounded stack, and compares it to a Pascal's wager. He adds that he was surprised to learn that in our own universe information is physical: general relativity and quantum theory together limit how many bits can be stored in a given region before it collapses into a black hole.

Making peace with infinity, and examples before definitions

Fridman says infinity does not make sense to him. Sanderson reframes it as an abstraction in the ordinary sense, the same kind of thing the mind does when it unifies ever-changing images on the retina into a single notion of Lex. Infinity, for him, is "the property of always being able to add one more". You never have to hold all the words or numbers in your head at once. He also suggests swapping the word abstract for general, so that the danger Fridman worries about becomes overgeneralization: ideas so general they apply to nothing usefully.

This leads to his teaching philosophy. He admits that he often finds mathematical texts so abstract he cannot get a grip on them, and that the channel channels that frustration. Courses tend to start from the top, with the definition of a vector space and its properties. "I don't think that's how understanding works at all." He prefers to start with concrete vectors, lists of numbers or drawn arrows, and notes that a visual forces the issue because you have to pick one specific example to draw. In his scripts he fights his own instinct to state the general thing first. Show a set of examples, let the brain feel the similarity, and the formula then articulates something already there.

The most beautiful idea, and a favorite video

Sanderson thinks the most beautiful things are those you understand a little but not entirely. His example comes from childhood: a small book, from a series he recalls as Wooden Books, with a formula setting a sum over all the natural numbers equal to a product over all the primes. He later learned it was the Euler product for the zeta function, which essentially encodes the fundamental theorem of arithmetic. He says he is still in over his head on how the zeros of the Riemann zeta function relate to the distribution of primes, and that this remaining mystery is why it stays beautiful. What matters is that nothing in it feels arbitrary or contrived, so another civilization would probably be studying it too. He jokes that he is past the dating stage with Euler's formula and is now dating the zeta function. For readers with some college math he points to the article on analytic number theory in the Princeton Companion to Mathematics.

His favorite video to make is "Who cares about topology?" It begins with the unsolved inscribed square problem (does every closed loop contain four points forming a square?) and then the weaker, solvable rectangle version, whose proof runs through a torus and a Mobius strip. He liked that these shapes arise naturally there and do real work, instead of being play with construction paper. Drawing the surface, which the original argument never shows, was a surprise even to him.

Process, mortality, and how to learn

The hardest part of his process is a topic he wants to do but cannot structure; he has been stuck on one for two or three months and thinks the right move is to set it aside. The judge in his head? "I think it's my past self." That has downsides: his quaternion video answered what he once wanted to know about four dimensions, but he would not recommend it first to a robotics or graphics programmer. A video is done when the script is done, and a script built around an aha moment is done when someone who did not understand now could.

A lighter exchange on mortality follows. Sanderson says he would live forever if he could, and questions whether death is what gives life meaning: "Would your life be four times as meaningful if you died at 25?" He points out that he is rarely thinking about death and is still motivated to make things. Fridman holds his ground, with good humor.

For learners, his main advice is direct: "Try to force yourself to do more problems than you naturally would." Glance at the end-of-chapter exercises before reading, and do not count a chapter as finished until you have worked a few. He calls this hypocritical, since his videos rarely include exercises, and says watching them is only part of learning. He recommends Khan Academy for beginners, suggests programming as a route into math, and repeats a rule of thumb, which he says may be a fictional anecdote, that you retain far more of what you teach than of what you read.

Conclusion

This is a conversation about how understanding happens. Sanderson's view is consistent from the alien thought experiment to the study tips: ideas are easier to grasp, and often prettier, when the notation and the order of presentation match what is actually going on, and people learn from concrete examples and worked problems before general definitions. Along the way he offers a balanced account of math as both discovered and invented, a skeptical take on the simulation argument, and a calm way of thinking about infinity. He ends on a happy memory of playing music on a gondola ride and a tentative promise to bring music back into his work, and Fridman closes with Paul Lockhart's line that "the first thing to understand is that mathematics is an art".

This is a summary of "Grant Sanderson: 3Blue1Brown and the Beauty of Mathematics" from Lex Fridman Podcast, giving you the GIST so you can decide if you want the REST. Listen to the full episode.