Chemical Rockets Are an Accident of History

Lex Fridman opens with the novelist’s split life: historical fiction such as the World War II thread in *Cryptonomicon*, and science fiction that looks the other way. Does history repeat? Stephenson is afraid it does, a lot. Human nature is what it is, so similar behavior patterns keep showing up. Something genuinely new is the exception. Technology has still changed the material facts. In the last hundred years, antibiotics, modern vaccines, electrification, and the internet raised living standards, health, and longevity beyond anything earlier history delivered.

People live longer and are generally healthier. The unattractive patterns remain. Fridman asks whether fewer resource constraints might let us emphasize the better angels. Until about the last ten years, Stephenson says, he would have taken that view. Then he watches people find ways to be divisive and angry because it scratches a psychological itch. The Weimar story used to sound sufficient: hyperinflation, unemployment, a collapsed standard of living, therefore Hitler. There was a bank collapse in 2008 and real stagnation for some. What the United States has seen in recent years is hard to explain strictly as poverty, rage, and sadness. “I think they want to be angry.”

The war’s technological curiosities fascinate him — Turing’s world in *Cryptonomicon*, among other things — but he does not dodge the moral residue. The camps mean educated people are never allowed illusions about human nature again. Even in a supposedly enlightened society, people can become monsters quite easily. Asked whether he agrees with Solzhenitsyn that the line between good and evil runs through every heart, he calls it a great line. As a teenager he read a good chunk of *The Gulag Archipelago* because his grandfather, an American obsessed with the Soviet threat, left the books lying around and wanted people to know what had happened.

Gulag and the German camps look like parallel systems to him: fear and desperation plus a charismatic leader. They are also stories of bravery, love, brotherhood, and survival. Fridman brings up Viktor Frankl’s *Man’s Search for Meaning* — beauty found under extreme conditions — and argues that the war is a small moment of evil that revealed a much larger good. Stephenson does not convert that into pessimism. Everyone is capable of following a Hitler and loving the hate of the other, for a time; everyone is also capable of good. Fridman’s working assumption is that the desire for good is larger, which is where technology can be a guide: Wikipedia, and Google’s original dream of putting the world’s information at your fingertips, true and false, and letting people explore.

Fridman’s futuristic version of that guide is an AI mentor you send out, not an authoritarian algorithm lecturing you on how to grow. Stephenson says that sounds easy and is not. It is the theme of *The Diamond Age*, a 1990s novel about a book that essentially does that job. He has watched people try to build the thing from different directions ever since the novel came out in 1994, in the era of AltaVista and Ask Jeeves, before he had heard of Google. He has not worked on it himself. He has a sense of the difficulty.

Then the systems get gamed. He had just seen a thread: twenty years ago, googling the Pythagorean theorem probably took you to a page that explained it. Now the top hits are likely someone with an angle — math tutoring, a marketing plan. Traditional engines have become less useful for their original educational purpose. That does not mean they should not be replaced by better ones. Fridman defends the people with the angle. They are hunting a business model. Teachers who love math run into the YouTube algorithm and try to make that love a full-time job. The dance with the algorithm finds a local pocket of optimality, or sub-optimality, and gets stuck.

Stephenson’s stance is flat. The pocket is way better than the 1980s and 1990s before the internet. Human nature then writes eloquent articles about how obviously this is a pocket, not very good, with better lands imagined beyond it. You wait for a lone genius, or a group of them with enough momentum, to say enough, here is the way out. Cynical “we are imprisoned forever” theories get clicks. They are clickbait until one curious ant tunnels the barrier or gathers energy to jump. Fridman extends the joke: we will colonize the solar system, get stuck there, lament the dying sun, and repeat until the observable universe is full. Getting out of this solar system, Stephenson thinks, is going to be the hard one.

You can leave slowly, whenever you want. Getting there in one lifetime, or a few, takes staggering energy to accelerate, then the same again to decelerate or you shoot by. Yuri Milner has funded laser-sail ideas for a small object. As far as Stephenson knows, they never talk about slowing down. It is a flyby. Take good pictures on the way past. He is not knocking it if it happens. Energy is only one issue. How do you maintain an isolated ecosystem that long? How do you keep people from going crazy? What happens if you hit something at a significant fraction of light speed?

The generation ship is, in his view, the only version that could work, and it has to keep going for a long time. The crew might arrive and find a garbage solar system — a large letdown after a thousand-year voyage. Survey helps, but the habitable window is narrow: gravity, temperature, atmosphere, which bound the size and type of planet. Stack those complications and building exactly the environment we want from materials already in *this* solar system starts to look a lot better. He likes a detail in *The Expanse*: the people building a starship are doing it for religious reasons. That, he thinks, is the only reason you would. Economically, perfect the rotating cylindrical habitat.

Fridman objects that everything ambitious is a little religious. Kennedy’s “because it is hard” is not a spreadsheet. Stephenson knew an Apollo veteran who once said the moon landings were communism’s greatest achievement. National conflict is not exactly a religion; it is a struggle for meaning, and meaning is hard to find in mathematics. It lives in music, religion, art. From a cost perspective he has nothing against launching one generation ship a year, wandering ants into the galaxy. The motive has to come from a spiritual or non-tangible calculus. From a business-model perspective: no way.

Wikipedia will tell you Stephenson was technically Blue Origin’s first employee. Fridman frames him as the person who helped convince Jeff Bezos to start a space company. Stephenson declines the myth. Bezos did not need much convincing; he had been thinking about it since he was five. What hatched in 1999 was advanced scouting: explore the corners of the possibility space. That entity was Blue Operations LLC, a small staff, precursor to Blue Origin. Around 2003 or 2004 the company swung toward the path it has followed since — existing aerospace models, chemical-fueled rockets, space tourism.

He still believes chemical rockets are an accident of history that comes out of World War II. Before the war, people like Robert Goddard built them small. Then Hitler desperately wanted to bomb London, could not quite reach it, and the Luftwaffe had been neutralized, so he decided to lob warheads on rockets. A terrible misallocation of resources; it only happens in a dictatorship run by a lunatic. They built the V-2. Completely separately, the war ended with atomic bombs. Suddenly there was demand for a delivery system that had not existed. Bombers work; hurling the thing on a rocket is better. Rockets got a second, much larger boost in the 1950s and 1960s.

By the time the rockets worked, hydrogen bombs were so large nobody really wanted to use them. You could fit a capsule with a couple of people into the socket meant for a warhead, and do that as a proxy for war. Fridman would have liked to be in the meeting where someone first suggested strapping a person to the rocket. It was probably a Russian, Stephenson says; they did it first, perhaps with less reverence for safety protocols. Through that competition, trillions of R&D dollars and rubles went into chemical rocketry, now advanced to an incredible degree. Orbit is just going really fast. There are other ways to do that.

Nerds have been inventing alternate schemes as long as they have been inventing rockets. Stephenson’s first lesson in the scouting years: put all your brainpower into a “new” idea for going fast and you will find some guy in Russia who thought of it fifty years ago and finished the math. Eventually you stop inventing and go looking for those people. They looked at Project Orion — Freeman Dyson’s scheme to detonate atomic bombs behind a large vehicle. George Dyson, Freeman’s son, was at Blue Operations and found old film from a La Jolla beach of a C4-powered prototype. A private company obtaining a large number of atomic bombs was out of scope. There were laser-pulse ideas with Arthur Kantrowitz: flash a layer of ice into plasma, which is opaque, so it absorbs the rest of the pulse and pushes. Freeman said that in rubber-soled tennis shoes you would feel a mild vibration. Jordan Kare and Kevin Parkin worked heat-exchanger versions with ground lasers or microwaves. Stephenson spent a while, semi-seriously, on giant bullwhips — a primitive object that breaks the sound barrier through the physics of a moving bend. Same physics could stand up loops of chain and, at maximum velocity, let a ship go. No crazier than chemical rockets. Chemical rockets just got the money.

Fridman asks about Tic Tac UFOs and Pentagon pilot footage: secret projects, ball lightning, or a lost alien science-fair drone. Stephenson has seen the chatter and not gone deep. Given how hard interstellar travel is, aliens are difficult for him to believe. You would not spend light-years to hang around aircraft carriers and not communicate. If they can get here they can probably hide; if they are not hiding, arrival ought to leave an energy footprint as they decelerate. A bungled science-fair project actually explains the aimlessness better. Can humans have sex with aliens, a Twitter question? Reproductive sex, no, unless you buy panspermia. Humans did have sex with Neanderthals and Denisovans. Broaden “sex” to gratifying physical interaction, dancing included, and you are already talking about love, which can take many forms.

Looking at Blue Origin and SpaceX ten or twenty years out, with William Shatner just up and back, is he impressed? SpaceX in particular has done things he calls unbelievable. Twenty years ago nobody anticipated the speed of those achievements, especially if you have seen even a little of how the sausage is made. Elon Musk’s unapologetic push to become multiplanetary, Mars as soon as possible: colonizing Mars is the kind of goal that is easily stated and catchy, well chosen to get people involved. Stephenson has technical questions — perchlorates on the surface, radiation, known problems. A business plan is harder. He cannot think of a Martian resource you could bring back cheaply enough to beat digging or growing it here. Maybe you need a little communism, meaning competition, to start. Plenty of people will go all-in on the vision without a business plan. White settlement of this continent did not have coherent decades-long plans either; the plans they had were often terrible.

Shatner’s reaction was the intangible case: rising through the thin blue blanket into the blackness, groping for words and still eloquent. In 2011 Stephenson wrote the essay “Innovation Starvation,” about an age unwilling to take on big, risky projects. People like Musk, willing to tackle ambitious work without a lot of soul-searching — tweet a joke, then actually dig the tunnel — are the counter-example. Ability to build at mid-century pace has slowed, for reasons including safety and environmental concern. Even maintaining what we already have is now a huge project. A lot of brainpower that might once have built rockets went into programming, which he thinks is understandable and great. Social media, though, has structural problems severe enough that he hopes we are not sitting in the same mess ten years from now.

Fridman wants new social media that beats the old, and refuses Jaron Lanier’s more cynical read of Silicon Valley. Stephenson had listened to Fridman’s interview with Lanier, whom he knows. The last time he paid close attention, Lanier was arguing that a like is a transfer of valuable intellectual property, tiny per click, aggregated for the platform’s benefit, and talking about data unions. Fridman heard hopelessness about Valley hubris. Stephenson did not hear a man without hope: Lanier would not write the books if he thought there was no way through, and he also just loves a good argument.

*Termination Shock*, set largely in Texas, follows a billionaire oil man and truck-stop magnate — T.R. — who decides to take on climate change himself. Stephenson will not say “solve.” Ameliorate the temporary effects. The carbon stays up there until you take it out, and we are still adding more. China had just talked about leveling emissions around 2030 and net zero by 2060. If the world did that at, say, 450 ppm, it would stay at 450 until removal. Removal means reversing the emptying of coal mines and oil reservoirs, the chopped forests, the peat bogs — back toward the 200-some ppm where it used to sit. Solar geoengineering is idea number one in the novel because volcanoes have already run the experiment. After an Indonesian eruption, an Australian plane sampled the plume; a scientist licked the windscreen and reported it was painfully acid. Pinatubo in the 1990s did it again. Sulfur in the stratosphere becomes sulfuric-acid droplets, bounces sunlight, cools the troposphere, and washes out in a couple of years. Good news: if you hate it, stop. Bad news if you like it: you have to keep doing it.

T.R. knows the papers, as people in climate science do, and acts unilaterally. Because it is Texas, he builds the biggest gun in the world, six barrels straight up, shells of sulfur. The book is less about the gun than about political reaction. Effects will not be even; some regions lose. Stephenson does not think this kind of intervention emerges from Western democracies. It is comparatively cheap and easy. Once things get bad enough, an individual or some government that decides it is in their interest will probably do it. That is diagnosis, not advocacy. High-altitude balloon measurements in Scandinavia already got squashed by local objection on principle. Vaccine and mask fights, in his view, show how hard even life-saving coordination is. Fridman blames leadership, not “the West.” Stephenson allows they may be entering interesting times in the Chinese-curse sense.

Should a slice of GDP go to becoming multiplanetary as climate insurance? He has never bought the Apollo-era argument that Earth problems must be finished first. Thinking about terraforming Mars might even change how people think about terraforming Earth. Big ideas as double-edged swords: not all of them, but the 1950s and 1960s were a period of unrealistic confidence — antibiotics, polio vaccine, the home refrigerator his grandmother still called an icebox. Then nuclear fuel cycles and DDT taught a backlash that is still running: assume every technology will bite, convince yourselves the downside is gone before you adopt anything. He thinks people are overly sensitized now. Fridman adds that AI fear often aims at the wrong failure modes and underweights the upside, the same way social media talk forgot that connecting a lonely species is not nothing.

Stephenson spent five years at Magic Leap as chief futurist, with a small Seattle squad doing what you might call content R&D for AR. The medium was so new it was more engineering research than a creative project. A real AR device has to run SLAM in real time: sense the room, reconstruct its three-dimensional structure, overlay graphics, and keep the human perceptual system happy. It tracks your eyes and how far you are looking. It cannot burn the CPU or murder the battery. That is table stakes. Content then has to render at low enough latency to look real and not make you sick. Magic Leap shipped a device that could do that in 2019. The ML2 was coming; he knew no more than anyone else, because he was gone.

Does it still boil down to a killer app? There is a base layer of engineering you need no matter what. Early gamers put up with low frame rates and, by later standards, crappy graphics because they were having too much fun with *Doom* — even *Tetris*. He was working consumer-facing content. A Wellington team made *Dr. Grordbort’s Invaders*, which in his view realized AR gaming’s potential in a way nothing else has before or since. That strategy held until April 2020, when the company pivoted to commercial and industrial applications. He has not seen the financial projections; he assumes they had reasons. From a dreamer’s chair it is heartbreaking, the same pattern he sees in robotics: the consumer future versus Boston Dynamics-style industrial focus because that is where the business is.

They built an app called Baby Goats that populated your room with kids. Because of SLAM, the system knew the table, the end table, the heights, and how high an animated goat can jump. Engineers had to turn SLAM primitives into game-engine objects the goat AIs could navigate. It shipped more as a sample and how-to than as a finished consumer product. Fridman would come home and watch goats for hours. Stephenson says it was ambient — not a game you sit down to play. Just life. Dogs and cats are not that different. Fifty years out, what wins for people with money: VR, AR, or physical reality? He has always been an AR fan, partly because it is an easy answer. Wear AR, put a bag over your head, it is VR. AR keeps you in something similar to the physical world. VR can go fantastical, and then motion sickness: inner ear versus eye, unless you are unusual. Designers can work with that constraint. Lots of people might live in VR. He would rather enhance the physical world, which is pretty cool, and moving around in it still counts.

Asked to imagine a VR so enriching he would stay, he offers a bridge. He likes making things in a machine shop, 3D printers, machine tools, and so he had to get good at CAD. He uses Fusion 360 and can spend hours in a monitor that is a window into three-dimensional space, rotating and imagining. Close to VR. Does the object have to exist? That is his game — print it, touch it. If you are building a game level or a virtual film set, the thing should preferably never exist. Lots of people already spend working hours in something near VR; the output device just happens to be a rectangle. Swap in a headset and they would be doing the same job. Fridman notes that a podcast is only audio and still a real friendship of a kind; books present even less and stimulate more. Stephenson does not deny that love and long friendship could live in virtual worlds. It depends on whether the applications can feed you those things.

Fridman psychoanalyzes the CAD hours: is some of the joy the hope that another human will see the thing? Turtles all the way down, the bottom turtle is sharing. Stephenson, speaking for himself, is always thinking in terms of an audience — someone, or an abstract theoretical someone. He spends a lot of time in Mathematica. A notebook trying to figure something out is huge blocks of text, thinking out loud, then graphs and calculations. Explaining and commenting helps him understand. Most of it nobody will ever see. He still writes as if a reader might, because that constraint makes the work better.

The obsession that ate the most hours: tiling a sphere with hexagons, with an eye toward Earth, recursively, hexagons inside hexagons. Probably a bad idea, because you cannot tile a hexagon with smaller hexagons — they stick out. Friends now send him every hexagon they see. Start from an icosahedron of equilateral triangles, a logical place, and you can cover with hexagons, but every vertex has to be a pentagon. You can have quintillions of hexagons and you still need twelve pentagons. He has blown a hell of a lot of time on it. A friend later sent a link to a codebase, he thinks from Uber, that carved the Earth’s surface into little hexagons. A knife through the heart. He will probably come back to it. Hexagons are about his speed as a math person; tiling in two and three dimensions is an ancient puzzle that has attracted brainiacs for centuries.

On AI’s next decades he does not think a lot, because it is deep and he does not consider himself well informed, and the word covers too many things. He has messed around a little with neural nets and PCA. He thinks granular and bottom-up. Fridman offers GPT-3-style self-supervised language models that read the internet, and embodied robots that are still pretty dumb — the Magic Leap problem of where am I and what will hurt if I hit it, before any deep scene understanding. Impact, Fridman thinks, comes from integration with society, not superintelligence: robots, or social-media algorithms, even while they are dumb. Stephenson’s robot exposure is a combat-robotics team and BattleBots. Those machines are remote-controlled, far from Boston Dynamics. What is interesting is the emotion. In the 2020 season a spare-parts robot named Rusty, salad bowls and farm hardware, became a crowd favorite. Other bots play the heel. People form narratives the way they do about humans, about objects that are almost inanimate.

If robots lean into that feedback — get cuter, or keep the quad-mount machine gun as a personality — they become a mirror. Fridman’s largest complaint about social-media algorithms is not that they exist. It is hidden change to your sense of what is real, plus data that should belong to people. Taking a psychedelic should not be a surprise two hours in. There should be research, open conversation, and then you become your own guide among flavors of real. Stephenson leans Platonic: mathematical objects have a reality not made up by human minds. He cannot explain where it comes from. He thinks they are discovered, not invented. Writing *Anathem*, he read some Husserl, a phenomenologist working as scientists were realizing a table is almost entirely vacuum. Husserl is hard; small bites. Why does the table feel solid? We are an evolved system with a biological interest in knowing where solid objects are. The take-home Stephenson likes is intersubjectivity: to stay sane you share perceptions and have them ratified, not even necessarily by people. A prisoner may domesticate a mouse or insects because they perceive the slam he perceives, and he is not just hallucinating.

Fridman notes that consensus is not engineering-real. It could still be the whole job of a robot: sit there, and when the door slams, look. Isn’t that the basis of love — shared reaction without a speech? He tells a freshman-cafeteria story. A joke lands for you and a stranger at the other end of the table. Eye contact. Next time you both look. Before you have talked, you are hanging out, same wavelength. Seemingly simple, extremely powerful. There is no reason you and a toaster cannot have that, if the toaster can ask whether that smells burned to you. Stephenson is willing to start with a robot that listens for slammed doors and scary sounds.

*Cryptonomicon* came out in 1999, set in the late 1990s, with hackers building what is essentially cryptocurrency. The Bitcoin white paper was 2008. Stephenson was reacting to Bay Area cypherpunks, a Berkeley branch and one in Austin, whose thinking assumed you needed a physical patch of Earth free of government interference. You could not do it by pure math on the network. You needed a room of servers a government could not meddle with. People hunted jurisdictions: Anguilla, Sealand in the North Sea. Blockchain makes all of that unnecessary. From a novelist’s point of view the old system was more fun — hackers wandering strange parts of the world setting up server rooms. A little of that remains, just several rooms instead of one, plus the new wrinkle of keeping GPUs from melting, in Iceland or in undersea containers. Governments, including China and more renegade ones, still have to decide what to do about mining. The *Cryptonomicon* picture had a whiff of South Park’s underpants gnomes: build the system, then money somehow. The middle step was missing. That middle step is now. Bitcoin exists. What now?

His inner cynic thinks that once it matters enough, existing banks and people in power will control it, and maybe the end state is not a big change. There is a utopian strain he is not sure about. The technology does make the utopian thread easier to pull, because governments have fewer options than with ordinary money — they can ban, they can make it illegal, it is harder. Fridman hears technology on the side of the powerless. Stephenson thinks money per se is one of the less interesting things you can do in that realm. Cryptographically enforceable contracts, and organizations built on them, have more potential for change. We already have money, largely digitized by the Stripes and the card networks. He likes the idea of smart contracts wired to weather and other world data, more like Mathematica: agreements grounded in data rather than in perceptions of data, distributing who gets to tell the story. David Brin has said the only way to settle arguments across a political divide is to make bets. People will claim the election was stolen until real money is on it.

Dogecoin, which he does not follow closely, lives in memes and humor while still doing some of the same machinery. It asks whether narrative and fun will travel farther than boring grounded tech once shelter and food are solved. Fun spreads faster than anything, good fun and bad. Fridman wonders whether fun-prevails is a hundred-year property of the internet or a couple of decades until the adults take over. Stephenson’s memory is the opposite. The adults took over first. Later people started using the network for frivolous things like memes. That, he thinks, is pretty much unstoppable, because even very serious people like sending around a funny picture.

He has been handwriting drafts for about twenty years, since *The Baroque Cycle*, as an experiment. When he was stuck, picking up a pen made it easy to go. He told himself he could always return to the word processor. That never happened. Ink on paper cannot be obliterated by an operating-system crash. It is slower, so a sentence spends longer in the mental buffer; he types fast enough to slam things out before thinking them through. First-draft quality is higher. Editing is faster: a line through a word or a circle around a paragraph, and you have the edit plus a record of what was there. Digital tools claim those features; the experience is different. There is romance in the simplicity, no intermediary beyond pen and paper. He owns fountain pens and started buying fancy Italian paper, thinking he would be conservative with it. The cost is still trivial; the habits did not change.

When he types it up, he uses Emacs, “obviously the superior editor.” Will Emacs and Vim still exist in fifty or a hundred years? If you work in Linux you spend a lot of time in config files and scripts. You need to pop in and out, even if the GUI is dead and you have only a command line. There will always have to be basic editors for administering systems. How widespread is another question. John Henry versus the steam hammer: Microsoft’s money in Visual Studio, Apple’s in Xcode. In theory enough Emacs Lisp macros can duplicate the features. At some point, feature for feature, keeping up gets hard. Fridman calls Emacs a cockroach of software — always in the shadows, slightly delayed, customized by individuals, outlasting fads. Shell scripts boot the modern world; they will be hard to replace. For print, Stephenson used simple TeX macros until publishers demanded Word. He wrote conversion macros years ago and this time used regular expressions to turn TeX italics into Word’s. The task is simple enough that his own tool beats anyone else’s, and it keeps him from classic formatting wanking — what if this heading were sans-serif.

What does it take to write a great story? His amateur theory is evolutionary. A million years ago, around a Rift Valley fire, if you can tell how you escaped the hyenas, or how Uncle Bob did not, and listeners can build that scene in their heads like VR, you have given them a survival advantage they could not have learned another way. Animals can shout danger. As far as we know they cannot tell more complicated stories. Plot-heavy work is sometimes frowned on in literary circles as pulpy. He has no compunctions. He likes stories that are grabby, fun, exciting to read. Once you have a yarn people will enjoy, you can do whatever you want inside the frame. Without that, you have nothing. Technical rigor’s main gift is details you would not have imagined, the complications and surprises the real world is constantly serving up. Delve into rockets or orbital mechanics and something weird turns up. Readers feel the complexity. The key is keeping disbelief suspended. Break it with a tell, and often the book closes and does not reopen.

Which Stephenson to read first? He asks what you like. *Snow Crash* is the best known, a cyberpunk novel that is also making fun of cyberpunk; the joke has layers you may not get. Historical taste: *Cryptonomicon* or *The Baroque Cycle*. Modern techno-thriller that is not especially science-fictional: *Reamde*, and *Termination Shock*. If you want an introduction to the human, maybe *Seveneves* — big themes, heavy things happening to the species, characters you can relate to, even if it eventually goes deep on rockets. Some people skip those parts; it is fine. Books that marked him, a random draw because he will remember twelve on the way out the door: *The Gulag Archipelago*, and *Moby-Dick*, which a great high-school English class rescued from the allegory hunt. As a kid he read Classics Illustrated comics — *The Count of Monte Cristo*, *Moby-Dick*, Stevenson, *Robinson Crusoe* — the whale without the sentences a nine-year-old cannot parse. He wanted to be a writer in fifth grade, the day a friend wore leather dress shoes. His never fit, he could not run, Iowa was cold, and he thought: I need a job where I do not have to do that. Heinlein young-adult books stuck. Greatest science fiction ever, excluding his own? If fantasy counts, easily *The Lord of the Rings*. Greatness is a hard quality to define. He was always a fan of *Have Space Suit—Will Travel*. Fun is a big component. Greatness might be underdefined.

Fridman, having spent time with Dan Gable and Tom Brands, asks whether wrestling and martial arts shaped him. Late, Stephenson says, but yes. He grew up in Ames. Gable was a few years older. They would go to university meets before the Olympic career, when everyone knew Gable was the star and not yet that he was the greatest of all time — a little dot on the mat, pronouncing opponents, alongside Chris Taylor, the 400-pound-plus Olympic heavyweight. Wrestling is the sport in small-town states like Oklahoma, Pennsylvania, Iowa, because a football team of twenty non-Gables can still stink, and a solo athlete can go to the Olympics. Gym-class wrestling he did not like: too competitive, the people who cared really cared, and he did not think he had the body type. After college, in Iowa City, Gable was coaching and winning something like nine of ten championships. Stephenson never met him. Gable has been in his sphere of awareness his whole life, including the story of a minor Ames arrest and a night in a cage doing pull-ups.

Growing up there was formative. It used to be orderly, high social capital, tiny class differences: a Cadillac instead of a Chevy and those were the rich people. A college town, Austin-ish. Pretty utopian aside from the weather. The martial art he ended up doing is sword work, which uses a different feedback loop. Grappling is touch, an old simple sense — earthworms have no eyes and still know when they are touched — so it is very fast. Standoff fighting, boxing or some sword arts, is mostly visual. Your eyes are doing SLAM, trying to figure out what the other person is up to. That is more his speed. Olympic fencing really starts when blades cross, and then you are back to wrestling. Older sword arts stand off and cut. Those are processed visually. He thinks of himself as a slower thinker; that fits. It tends to be short engagements: close, try an idea, it works or it does not, back off.

Fridman dwells on the historical stakes: the richest people in those societies buying the best gear and training because it was still life and death, Miyamoto Musashi, a poetry of mastery later weapons may have closed. One overlooked good of the internet, Stephenson says, is that isolated people who cared about this found each other. Nobody is at that old level yet. It is happening. Old treatises come out of libraries, get translated from dialects of Italian and German, and people learn the techniques with different weapons. In Austin, Da'Mon Stith works on historical African martial arts and on machete fighting that enslaved Africans learned in the Caribbean and South America. He is probably within a mile of the studio. An amazing guy.

Advice for a young person who wants a life they can be proud of, maybe even a large impact: find the thing you will do happily. He does not want to say obsessively; that sounds pathological. Sit down, start, snap out of it hours later, where did the time go. That is a key self-discovery, because without it you cannot tell where to put your energy. You can intend world peace. What matters is how you spend the hours, and whether that is productive. Intelligence and good intentions do not substitute. Fridman adds the missing stair: for him the fun starts after the first hump of really sucking, then you only suck regularly. People give up too early. Give the struggle time. Stephenson agrees it is an important addendum. It may not happen the first session. You may have to climb the learning curve, and life pressure that makes you feel bad about sucking can stop you short. That is why good teachers exist.

Another internet good: YouTube. The person who made the video does not care how many times you hit pause and rewind. They will never roll their eyes. Sometimes spending a huge amount of time on one video or one book — rereading, rewatching — solidifies love and depth. It is okay to stay. He used to think there were so many books he had to keep moving. Somewhere in college he realized you could spend a whole life with a single textbook. There is enough in it. Misner, Thorne, and Wheeler’s *Gravitation* is one of those. Roger Penrose’s *The Road to Reality* is another: it starts at two plus two equals four and ends at the boundary of physics.

The ridiculous question: meaning of life. As far as he knows, we are unique. There is no evidence of anything else in the universe as complicated as what is between our ears. There might be; you cannot rule it out. We appear pretty special, so meaning has to have something to do with that. He likes David Deutsch’s *The Beginning of Infinity* on the power of explanations. Most civilizations are static: a set of dogmas, passed down, nothing changes. Huge changes happen when people follow whatever you call the scientific method or the Enlightenment — figuring out why things are the way they are. That has produced, in a few centuries, explosive change compared with everything before. Deutsch almost classifies it as a force of nature. If we keep going, if we figure out how to spread to other star systems, it effectively is one: the drive to understand deeper, and to engineer so we can understand more. Fridman offers the old line that the universe made us to understand itself, plus all the peculiar side effects — egos, short lives, love, war, the concert we play together. Stephenson wonders which of those, death and love included, are fundamental, and which are evolutionary quirks. From an engineering angle, when we build the intelligent toaster that listens for the slammed door and the smell of burning toast, should that toaster be afraid of dying, and should it fall in love the way we do?