Scott Young - Ultralearning

The conversation opens mid-thought, with Young arguing that most people are far too unambitious. He does not mean the ambition of wanting to beat others. He means that people rarely conceive of big projects, never work on them, and lack big dreams about doing cool things. When they do have such dreams, he says, these usually boil down to social status: wanting to be the person who does something better than other people. He admits he does not know how to change this.

What he does believe is that a culture rewarding ambitious, original, interesting work, where you cannot know where it will lead, is valuable. Having that drive in you is rare, and he thinks cultivating it is probably good both for the individual and for society.

The host introduces Young as the author of Ultralearning and asks what explains the "annus mirabilis", the concentrated burst of contributions in Newton's and Einstein's careers. Young first reminds the listener that most people never have such a year at all. He then puts much of the phenomenon down to a selection effect: a smart person happens to be working on the problem that leads to a breakthrough. He illustrates with a counterfactual in which Newton's heavy time on alchemy might have yielded a way to turn lead into gold, but that is not the world we live in. Newton's physics and the Principia are what paid off.

Einstein, Young says, is a rarer case because he had several key insights. He proved the existence of atoms through Brownian motion, explained the photoelectric effect, and then produced special relativity. He notes that the photoelectric effect, which arguably started quantum mechanics, is what actually won the Nobel Prize, not relativity. General relativity then cost Einstein roughly eight years of struggle with the mathematics. Young adds that lists of geniuses are sometimes populated by people who do not deserve to be there, but Einstein is accurately rated: seen as extremely smart and important, and actually so.

The host presses on why three such different problems (special relativity, the photoelectric effect and Brownian motion) fell in the same year unless some deeper principle is at work. Young admits he does not have a definitive answer, but suggests the problems were amenable to Einstein's particular style of thinking. Drawing on the Isaacson biography, which he summarized on his blog, he describes Einstein as one of the great intuitive physicists: a spatial, visual thinker whose thought experiments were the mechanism for generating insights.

Young gives special relativity as the example. Experiments suggested that the speed of light does not vary with direction, which is odd, because a wave moving through a medium should appear to change speed if you move relative to that medium. Einstein worked through the geometric implications and arrived at ideas such as lengths contracting at higher speeds. These conclusions are mind-bending, Young says, but they came from rigorously working out an intuition.

He contrasts this with more mathematical physicists, who were very strong at advanced mathematics and asked less "what is my physical intuition?" and more "what is a good way of representing this?" Young flags that he is outside his comfort zone and offers matrix mechanics as an example of someone noticing an interesting pattern that makes the math easier. He believes Einstein worked with Minkowski on the tensor material because he was more limited there. That is not to say Einstein was weak at math, only that his breakthroughs came from intuition first, formalized afterwards. His tentative conclusion is that certain problems were uniquely suited to Einstein's style, and that a few insights may simply have needed someone with that skill set to unlock them.

The host links this to Young's essay "The Narrow Path of Success" and asks whether Einstein undermines it, since a patent clerk who still could not land an academic job years after special relativity seems to contradict the idea that you must follow the right track. Young replies that his point is aimed at a romantic belief: that in high-ambition fields you can do whatever you like and someone will eventually recognize you. Actual data on how these careers work shows otherwise.

The motivating example for the essay was Jason Brennan's book on academic careers, which takes a data-driven approach. As Young relays it, far more people enter academia than there are academic jobs, and the filters for getting those jobs are rigid. He argues Einstein is not really a counterexample, since Einstein had bad marks on his résumé and a hard time finding a physics job. The right reading is that Einstein struggled because he lacked the right résumé. Exceptions with a wrong start and a spectacular one-in-a-million result exist, but, Young says, you cannot count on being one.

you shouldn't be going into a poker game you know well if i get a royal flush then i'll be really good

Young's advice is to bet on the path that most of the people who find success in a field actually follow, and to play the average hand you probably hold. He makes an exception for anyone in the top 0.01 percent, who presumably know what to do better than he does. For everyone else, understanding the typical path matters because stepping outside it brings strong headwinds, and doing so without reason stacks the odds against you.

His favorite example from the essay is non-fiction book publishing, where there is an established route. First you get an agent, because that is easier than getting a book deal. Then you write a proposal, which is partly a description of the book and partly a business plan showing you have thought through why it will sell enough to interest a publisher. Then you pitch, get the deal, and only then write the book, so the writing comes very late.

Most people write the book first and then look for a publisher, which, Young says, signals an amateur to publishers and makes a deal difficult even for a pretty good book. There are exceptions, but they are hard. Self-publishing is a separate matter and none of this applies. He recalls that one reader was upset and asked why they had to do it this way. Young's answer is that nobody has to; the point is that not doing so makes things harder.

He is happy for people to do something weird and creative, provided they know that this is the strategy they have picked and that it makes some things harder. He links this to a central theme of Ultralearning and of his philosophy in general: before starting a project, research how such projects typically succeed. Even if you then decide your own path will work better, you should know the status quo and why it works.

He adds that this research is not hard for most things and does not require secret knowledge. In his view, most people simply do not know how to do it, do not care to, or do not want to learn an answer they might not like to hear.

Conceding that perhaps one should not study exceptions so hard, the host asks one more question: Einstein seems to have lacked the structure, discipline and organization that an ultralearning project requires, so is that evidence for a more fun-, excitement- and curiosity-driven form of learning? Young is not sure that holds. Relying again on the Isaacson biography, he says Einstein did an enormous amount of deep thinking on hard problems when he did not need to, even while working as a patent clerk.

He also stresses a point about the genre. A self-improvement book such as Ultralearning takes the spontaneous, intuitive approaches that successful people used and tries to turn them into a structure that someone who would not do this automatically can follow. Anyone analyzing an important domain for others has to put it into a format of steps. Einstein was one of the people Young researched for the book and might have been a vignette in it, but he was too famous and people had already heard his story.