You Better Already Have Everything You Ever Wanted

Lex Fridman opens with Johnny Cash's "All I Do Is Drive," then asks Steve Viscelli the same question Cash put to an old driver: what is the life actually like. Viscelli is a former trucker and an economic sociologist at the University of Pennsylvania. His first book, *The Big Rig: Trucking and the Decline of the American Dream*, grew out of six months on the job; a second, on autonomous trucks, was underway when they spoke. He answers as an ethnographer. He lives with people, works with them, and tries to thickly describe what makes a subculture tick, then, following C. Wright Mills, set that biography inside history and power.

If he has a talent, he says, it is that he can sit down with anyone. His wife teases him for it. He learned it from his father, who ran a factory gate that truckers passed through all day, always had a joke, and remembered names. The rule the son still recites is simple: everyone has something to teach you. He is a ruminator, not a performer. He sat with the trucking material for nearly ten years. The ideal interview, he jokes, would be "tell me about your life as a trucker" and then silence while the person talked about exactly the right things. It never works that way, so the work is preparation.

Long-haul drivers have to stop for fuel, showers, and food, which gave him a natural census point. He picked a truck stop at the junction of major interstates, stood in the lounge with a clipboard, and asked whoever walked through whether they were on break. He figures he asked 103 or 104 people to get the first 100 interviews. Almost no one refused. They sat with a stranger for an hour or more, which told him something: truckers have a lot to say, they are alone a lot, and once you learn to turn the spigot they will talk. He later sends journalists to places like Effingham, Illinois, a knot of huge truck stops in corn country, and tells them to try the lounge.

Becoming a driver is, in his book, a story employers construct. The truck is 70 feet long and can weigh 80,000 pounds. It does not stop or turn like a car. When he started, the first skill was shifting a nonsynchronized transmission: double-clutch, drop the engine an exact amount, then find the next gear. Miss the timing in an intersection and the box grinds while you coast to a dead stop under the light. Get it in gear with the engine lugging and the torque can throw you out of the seat. He calls it riding a bull. The school's trucks were old, smoking, and already destroyed by classes of students before him.

Companies ship recruits three states from home, send a bus ticket, and put them up for two weeks. Classroom in the morning, a giant parking lot in the afternoon. For almost everyone there the stakes are extreme. Trucking has become a job of last resort: the plant closed, construction wrecked the knees, the work was offshored. Recruits have left families and often another job. They may be charged about $8,000 that they must repay if they are not hired. Viscelli watched grown men break down crying because they could not go home and tell a son they had failed. A trainer later told him the boot camp is a two-week job interview. The company is watching who stays calm when the right move is the slow one, because a driver on the road is unsupervised.

The psychological pressure of traffic staring at you is not a side issue. Fridman admits trucks stress him out; Viscelli has seen people freeze in an intersection with horns blaring and gears grinding, then shut down and quit. After six months he was a pretty good rookie and still a rookie, especially at backing and tight corners. The gulf between that and two or three years of experience was huge; twenty-year drivers were another step beyond. Trucking is not classified as a skilled occupation. He thinks that classification is wrong.

City work and highway work are different jobs. Local driving is geometry and patience: tight turns, pedestrians, crowded docks, and a trailer that answers a quarter-turn of the wheel a few seconds late. Some people have a gift for spatial angles. Everyone still has to repeat the move until it is in the body. A lot of that time is unpaid. Drivers wait for a warehouse to free a door, for a forklift to throw fourteen pallets, for someone to call them on the CB. An hour or two can vanish while the meter of pay is off.

Long-haul driving, as he came to understand it, is a set of habits. Men especially, he says, treat driving skill as the ability to maneuver out of tight spots at speed. A really good truck driver understands traffic well enough never to need those heroics. The mantra is leave yourself an out: a place to put 80,000 pounds if the four-wheeler ahead, the one texting, loses control. The best drivers do not wait for that moment. They see the pattern early, slow down, and stay out of the situation entirely.

That is why hiring managers used to want two years without an accident. A road test can show you can snake through cones and slow for a school zone. What proves you are safe is day after day, mile after mile, keeping yourself out of crashes before they form. Drivers with a million or two million safe miles are valuable to a company like Walmart for a blunt reason: they do not run over minivans. The company Viscelli trained at, he was told, killed about twenty people a year. Someone went to the funeral with a large check. "Don't be that person," the pitch went, and then the cost was written into the business model.

Viscelli started at 25 cents a mile and was bumped to 26. A new minimum of $500 a week covered weeks when the miles were not there. Pay arrived when he turned in the bills, the paperwork that lets a company invoice the customer, so checks bunched: some weeks the floor, some weeks $1,200 or $1,300. He was a poor graduate student, and it felt like real money, which gave him a natural incentive to run. Starting pay in similar jobs had moved into the 30-to-35-cents range by the time of the interview. The cents-per-mile figure still hides the clock.

Mileage pay was invented when firms could not surveil drivers and wanted them to have skin in the game. Paid by the hour, you might pull into a diner for breakfast. Over time, more of a driver's week became non-driving work. Railroads took a lot of the long freight. Congestion grew. Drivers are cheap and usually lowest on the totem pole, so their time is used badly. Viscelli would arrive at an upstate New York brewery with a load of beer for Michigan and be told to sit until the dock called on the CB. Two hours unpaid, then he started earning, then the same wait at the other end. Rail, traffic, and cheap labor all push unpaid time onto the person with the least bargaining power.

Federal rules cap long-haul driving at 60 hours a week, so drivers paid by the mile under-report time to cram more miles into the legal window. Companies then say the log shows 45 hours and that is all the minimum-wage clock they owe. Count waiting, paperwork, inspections, fuel, and dock queues the way most people would, and a typical week is 80 to 90 hours. Those drivers often do not make minimum wage. Department of Labor rules, Viscelli notes, say a driver dispatched away from home for more than a day should get minimum wage around the clock, which pencils out near $55,000 to $60,000. That is little money for two jobs stacked so you do not sleep beside your spouse or see your kids at night. It would still, he says, end the shortage if it were actually the floor.

California had about 145,000 jobs requiring a Class A commercial license and about 435,000 people holding one. Getting that credential is not a walk-in DMV test. People quit other work, sat unemployed, paid for school, left home for months, and then said they could not do the job. The shortage is not of licensed bodies. Satellite computers now tell a dispatcher that a truck sat at a GPS pin for four and a half hours, which kills the old excuse that mileage pay is needed because you cannot watch the driver. What remains is a way to dump a customer's dock delay onto the worker. If you are Walmart and one carrier threatens detention fees, you ask the other eleven. From an economist's chair, Viscelli says, the market is working perfectly. Training is often publicly subsidized through trade-adjustment or GI benefits. Diesel particulates and extra congestion are not on the invoice. That is why this is the cheapest, and therefore the winning, way to move freight.

Fridman steals a question from a *Penn Gazette* interview: was there a golden age for long-haul truckers, and what ended it. By the late 1970s, Viscelli says, a typical driver was making well over $100,000 in today's dollars and getting home every night, more than union auto and steel workers by 10 or 20 percent. Jimmy Hoffa had, by the mid-1960s, unified most of the nation's trucking labor under the National Master Freight Agreement. You were covered by that contract or your employer paid close attention to it. The Teamsters were, in that moment, the most powerful union in the country. That was the golden age of job quality.

A second golden-age story ran beside it: the mythic long-haul outlaw. Agriculture was left out of New Deal trucking regulation so food prices would not rise on urban workers, and that unregulated fringe produced the asphalt-cowboy image. Late-1970s movies, TV, and a CB craze filled the culture with salty jargon, "bears" for cops, diner parties, pills, and California turnarounds. The masculine pitch still has force. You have 40,000 pounds of something you care about; I will take it from New York to California; do not worry about it. For people who do not want office politics, that tangibility is the draw.

Deregulation in 1980 let the market set rates and broke the national-contract system. Viscelli's working hypothesis was that drivers who had lived under the union would like it more than those who had not. That was true, but with a twist that surprised him. Even the owner-operators portrayed in the deregulation debate as small-business cowboys shut out by monopolies and a corrupt union told him they made a lot more money when the Teamsters were in power. Trucking tends toward excessive competition. You cannot warehouse a trucking service until rates recover; economists call that derived demand. You keep the wheels moving to pay the notes, even at a loss. An organization that limited entry and held rates up protected pay for people inside the contract and, it turned out, for people outside it too.

Hoffa's reputation is complicated, and Viscelli does not dispute that he broke laws. He also notes that Hoffa invited a husband-and-wife team of Harvard economists to follow him and opened the books so they could see how he negotiated. The older Teamster insight, going back to the 1800s, was that workers do better when employers do better, and employers do better when the market is controlled. Collusion was often initiated by employers who knew they could not limit competition without labor. Private fleets, hauling their own freight, pay more than for-hire carriers because they are not bidding against a dozen rivals and cannot afford disruption. Truckload is point A to B with a full trailer. Less-than-truckload is a terminal system of local pickup, sort, line-haul, and local delivery, closer to how UPS moves parcels. The remaining good union jobs live in that second world.

Did it get lonely. Viscelli did not have children then; he cannot imagine the job now that he does. He had been married five years. His wife hated it, and so did he. The scene that went into the book is a truck stop in the middle of nowhere in Ohio, sitting on a concrete barrier, eating Chinese food, watching fireworks in the distance on the Fourth of July. His wife called from the family barbecue. Their anniversary is July 8. She asked if he would be home. He said he did not know. A cousin's husband, who had driven truck, had warned him that his advantage was knowing the work was temporary.

The last question he learned to ask, once he knew how to open the spigot, was what advice a driver would give a nephew. Grizzled, tough men broke down answering it. You better already have everything you ever wanted, one said, because he had a car ten years old with 7,000 miles on it, a boat that had not seen water in five years, and kids he did not raise. Two weeks out, home dead tired, a punishment list from his wife, gone again Sunday night. The best parts, Viscelli still names: pride in piloting something dangerous, and a behind-the-scenes tour of the economy, the cavernous Bud Light plant that seemed to supply a giant swath of the country. The novelty wears off. After a few months the days blur into white lines, yellow lines, and another dock.

That same lean machine showed its fragility in 2021. About 60 ships sat off Los Angeles and Long Beach, each holding thousands of containers retailers wanted on holiday shelves. The port was stacked with boxes it could not clear because truckers were not picking up, and truckers were not picking up because warehouses were full. Viscelli bristles at "panic buying" as the main story. There was some of it, toilet paper included. The larger shock was a massive shift in behavior: restaurants closed, Sysco was stuck with 50-pound sacks of flour and cases of lettuce, households bought differently, and forecasts still used last year's shelf. Suppliers who had kept two months of bottled water wanted three. Warehouse operators run at 90 percent-plus occupancy. The system had no spare bay for a 50 percent jump in a category. With luck, he said then, the mess might clear by the following Christmas.

Self-driving trucks became the cartoon of robots taking jobs: a truck that drives itself, a decent blue-collar wage, a man without many other ways to earn it, a robot that takes the seat. Around 2016 that anxiety was everywhere, which is how Viscelli got pulled back in. He had just finished the first book when someone at Uber, which had bought Otto, asked him to talk to engineers about drivers' lives. Some of those engineers imagined an owner-operator sleeping in the bunk while the asset he owned kept earning. Public reports were cruder. They took BLS categories of about 3.5 million sales-and-delivery workers, or about 1.8 million heavy-truck drivers, picked a start year and a slope, and assumed the jobs vanished.

Annette Bernhardt at UC Berkeley's Labor Center wanted people planted in real industries to complicate those models and called Viscelli. The industry is too diverse for a straight wipeout, he said, and he had been ready to leave trucking after ten years. He stayed. Direct impacts, in his estimate, run around 300,000 jobs, far below the headlines and still brutal for the communities involved. The number he wants heard is the caveat: 300,000 is a static picture of the existing industry. Autonomy is transformative. It will not swap a self-driving truck for a human one while everything else stays put. It will reshape supply chains, landscapes, and the fight against climate change.

Predicting second-order jobs, Fridman suggests, may be hopeless, as it would have been to foresee Wikipedia from the internet. Viscelli agrees the prediction is the wrong question. The question is what future we want, and whether we will shape it. The laissez-faire story, that technology just happens and we fix the mess on the back side, is not even historically true of this technology. Congress in 2000 wrote a priority that a third of military ground vehicles should be unmanned, before aerial drones had proved themselves, and funded the DARPA challenges that followed. Without that public spark, he says, they would not be having the conversation. Drivers still ask if it is real. His first answer was the Pentagon, the top robotics labs, and Silicon Valley billions. The later version is heavier: if self-driving vehicles do not happen, it will be the biggest technology-failure story in human history. Garage inventors have failed beautifully for a lifetime. Whole industrial giants, GM among them, have not piled on like this and then missed.

Big jumps in trucking, Viscelli argues, come from constellations, not single inventions. Daimler put a motorized truck on the road in 1896 and at first only replaced horses. World War I pushed larger trucks and pneumatic tires; then came paved roads and the internal combustion engine as a set. Forty years after that engine, intercity trucking became essential in the 1930s because shippers wanted out from under the railroads. Self-driving trucks, he thinks, will be a piece of a similar mix. Sam Walton's central insight was a trucking insight: build a distribution center, ring it with stores, and never site a store farther than a human-driven truck can run out and back in a day. Competitors made trucking fit the stores. Walton made the map fit the truck. Logistics became Walmart's edge. Amazon, having mastered the digital half, still has to dominate the physical move.

The dangerous mistake is to treat autonomy as a labor-cost-per-mile substitution. A meat packer today has a driver arrive the night before and wait out a legally required ten-hour break so he can run a full eleven hours to the farthest grocery DC, keeping product out of an overnight refrigerated trailer. The network is designed around that constraint. If the truck can go two or three times as far, you do not need the warehouses where they are. You can centralize inventory and substitute cheap transportation for storage. Landscapes get remade the way big-box supply chains remade them. E-commerce wants the opposite of that old model: huge variety, small quantities, almost everywhere, as fast as possible. The pent-up demand looks, to Viscelli, like intercity trucking under railroad monopolies.

Then the climate arithmetic turns ugly. About five million containers move from Los Angeles and Long Beach to Chicago by rail. Rail is at least three or four times more efficient on greenhouse gases and maybe 20 to 25 percent cheaper on that lane, at the cost of a couple of days. Cut highway cost per mile by 30 percent and the truck is cheaper than rail and five days faster. How many millions of boxes leave the ports on autonomous trucks? If the industry chooses platooning, the highway may look like a train laid on asphalt. Trucks going twice as far and hauling three times as much also blow up the battery story. From a cab you can already see cars move in packs, bunching behind a slow vehicle and then breaking free. The future may not be neat two-truck platoons. It may be hives: twenty trucks in some coordinated cluster, with a pod of self-driving cars hanging back because people do not like to be near them.

Viscelli went to everyone he could find who was building a self-driving truck and asked a lay question: what are you trying to build, and what do you envision it doing. For a lot of them the use case was an afterthought. They knew the technical capabilities, and those were the problems they were solving. They were engineers, computer scientists, roboticists. Society, policy, and business would come later. "That's not how life works, friends," he says, not unkindly. Highway driving is simpler than cities, so the shared plan is to shrink the problem to interstates. From the interviews he distilled five scenarios developers were batting around, plus a sixth he liked that the U.S. military had sketched for fuel convoys.

The first is platooning: trucks following close with cooperative adaptive cruise control so the lead vehicle punches a hole in the air. Peloton was an early name. On a shared panel the founders asked what it felt like to tailgate another truck. Viscelli said drivers would hate it; following distance is the cardinal rule you do not violate. Drivers who actually ran Peloton platoons split: some loved the screen view from the lead cab and the voice link, others said it was too close. The second is teleoperation. Starsky Robotics drove a truck in Florida from a San Francisco office. It looks like a video game; latency is the hard constraint, and NDAs stopped him from saying more than the company's public challenges. He still thinks remote piloting is more relevant than rivals admit, including as a way to limp a sensor-faulted robot to the next Penske instead of stranding freight on the shoulder.

The third is facility to facility. Many logistics yards sit on wide commercial roads next to interstates, without bikes or parked cars. Less-than-truckload line-haul between terminals, the runs UPS and FedEx already keep inside a day, is the obvious first Waymo-style driverless deployment. The fourth, the one most people are betting on, is exit to exit: a human brings a trailer to a drop lot outside town, an autonomous tractor takes the interstate, another human brings it in. That is his older "urban truck port" idea in modern clothes. The fifth he calls autopilot, no kin to Tesla's branding: the driver goes out with the truck, then off duty in the bunk while it drives itself. DOT rules say off duty means completely relieved of responsibility. Construction, weather, or a problem twenty miles ahead that pings you back into the seat means you are on duty, and you cannot get uninterrupted sleep.

Fifteen years of ordinary truck technology already did a version of this story. When Viscelli learned, you shifted by hand, read laminated maps so you would not destroy a bridge or "can-open" a trailer under a low overpass, and planned Buffalo by 6:30 so you missed Cleveland's rush and still made Chicago. Today the truck shifts, the box routes around congestion, and hours of service mostly report themselves. Satellite terminals ended the era when a driver could tell a dispatcher he needed two more hours of sleep. All of that was de-skilling: downward pressure on wages, more speed, more monitoring. The burden, he tells technologists who say autonomy will make the job better, is on them to specify how. Thirty or forty years of history say that technology entering a market where workers are weak and cheap wins, and it does not raise those workers' pay. It lowers the skill bar.

Forward collision avoidance, he notes, can wipe out most of the minivan crashes that made a conscientious $100,000 driver worth his benefits. Capitalists look for expensive labor to cheapen; that is the last two centuries of productivity. Will workers get screwed. "Most likely," he says. "That's what we do." Economists in the comments will remind him that technology creates higher-skill jobs. He knows. The live questions are how many, whether they go to the same people, and whether they sit in the same places: rural workers, often in red states, and the remaining blue-state union line-haul that will try to slow autonomy. A truck whose wheels roll only seven or eight hours a day while the driver lives in it is a spectacularly inefficient asset. Good jobs already raise productivity by putting two or three trailers behind expensive labor. That is the analogy he wants for robots.

The sixth scenario, the one he thought was actually good for labor, leaves the human brain in the lead truck and hangs a drone unit behind it, using the rear of the trailer as a lane reference and cooperative cruise for spacing. One driver, twice the freight. At weak bridges the pair can open to 50 or 100 feet, then close up, which also fixes a problem in his old double-trailer truck-port scheme. The follower's decisions get more important. That is upskilling. In 2017 Senators Jack Reed and Susan Collins asked for research on self-driving trucks. Viscelli fed his six scenarios into a GAO landscape report meant to tee up a DOT and Labor stakeholder process. At the workshop a friend held up the discussion list: they looked familiar, except there were five. A footnote said the sixth was omitted because no developers were working on it. That had been his research method, he notes, dryly: ask developers what they want to build, then add one idea nobody was talking about.

Computer scientists told him the follower still had to be a full self-driving truck in case of a cut-in. He thinks the edge cases of platooning are orders of magnitude fewer than solo robotaxi driving, and that treating them as equal is ridiculous. Who has authority to propose a future is the deeper question. DOT and Labor might have said they wanted fewer deaths, less carbon, less congestion. Instead, narrow safety mandates mean the Federal Motor Carrier Safety Administration can write rules that decide which scenario is profitable without ever being allowed to write a rule because it is good for drivers. About six months after the workshop, a company called Locomation fielded essentially that design with a twist: humans in both trucks, the follower off duty, then a "slingshot" when the lead driver runs out of hours. Had they existed a little earlier, the sixth scenario might have been in the room.

Fridman asks what to put to Waymo's trucking lead. Viscelli's on-the-record question is which first-adopter scenario they will actually run, because the expensive remaining Teamsters jobs sit in LTL line-haul. He also sends listeners to *Men and Machines*, a 1960s photo-journal by Harry Bridges, the socialist leader of West Coast longshoremen. Nearly a hundred pages of how they used to stow olive-oil barrels by hand, pride in a tight stow, then cranes and clamshells. Bridges was persuading members to sign an agreement that let employers automate: jobs would fall, wages and hiring-hall control would rise, and the last person on the dock would be very well paid and might love the work. That postwar contract, wages tied to productivity, is what trucking no longer has. Silicon Valley, Viscelli adds, has made demo videos culturally acceptable. He gives Waymo and Tesla this much: they put vehicles on real roads, which is the only way a policy debate gets data.

On the Tesla Semi, Viscelli's judgment is blunt: electrification and autonomy run opposite ways in the near term. Safety systems and driver-support features can be excellent. Pulling the driver out for long stretches, the way a true driverless tractor would, runs into batteries that are not ready for heavy long haul. Fridman notes that Tesla's city-driving, last-mile, vision-only bet might still fit the human legs of an exit-to-exit network. Viscelli wants to see those electric trucks running out to truck ports, a person in the lead, a drone behind, hydrogen or not. He has no horse in the sensor race. For all he knows, the winning kit will be a biomechanical bat-stem-cell echolocator. What he does admire is first-principles thinking. Legacy car meetings, in his experience, are about making this bracket cheaper. Almost no one asks why the plant has been laid out this way for decades, or whether a facility built from the dirt up for unmanned docking could be ten or a hundred times cheaper.

Fridman had expected pure vision plus a public beta to produce a safety and PR nightmare. It has not, at least so far; cars are not constantly running off the road, and there may be self-selection among Autopilot users. Culture, not just sensors, is the gap. Toyota, GM, and Ford are among the greatest high-volume manufacturers on earth and, he says, horrible at hiring software people. Markets are good at some things. Transportation is a public roadway system everyone already pays for. If we cannot have a policy conversation about how technology remakes that, we are leaving too much to private firms. He does not want companies to stop; he wants them at a table that is not just a lobbyist slipping a sentence into a bill. Elon's species-level goals, Mars included, impress him. Waiting for a hero to fix climate, labor, and roads does not.

Fridman offers a theory: heroes make leaps, markets sand down the inefficiencies afterward, and autonomy needs a Henry Ford. Viscelli will not give up the heroes. He watched SpaceX boosters land nearly at once while his kids were home from school and thought science fiction had been made real. He wants that energy aimed at the clock that is already past time. In 2010 he got a grant focused entirely on climate adaptation rather than prevention, the moment he felt the country was admitting catastrophic impacts were baked in. The project asked what happens to trucking when the breadbasket moves from California to Illinois. Henry Ford and Daimler's first truck matter. So do the public stages: the Erie Canal, which people thought was crazy; railroads that would not have been built without Lincoln's land and funds; Eisenhower's interstates, whose payoff no private entity would have financed; electrification, dams, water. Rapid at-home COVID tests, cheap enough to make by the hundred million, were to both of them an obvious bipartisan public good that did not happen.

Fridman asks whether, going to bed, Viscelli sees a beautiful world of autonomous trucks for the people he has become close to. Viscelli wants to be a glass-half-full person and flinches at the framing. Harnessing the technology for labor and climate will take the best of policy and a rebuilding of institutions whose erosion he heard for years in cabs. He grew up in the Rust Belt. All his ancestors worked at the same factory, still hanging on. Democrats told blue-collar workers not to worry about free trade. Economists will explain in the comments how trade lifts everyone; he suggests they find a drywaller in South Philadelphia and say immigration has not hurt him. Overall gains can be real. A system that lets some people pay very high costs, with no channel to say so, is also real. Trucking is a last resort after the plant or the knees are gone. Go back through fifteen years of his interviews and the same voicelessness is there: no difference between the parties, both will screw us. You can ignore it while the factories leave.

He used to teach a course on rich and poor in Philadelphia, lining up Section 8 housing vouchers next to his own mortgage-interest deduction. In his neighborhood the two subsidies were about the same, but his social identity was not "welfare recipient." That distorted map of who benefits, he thinks, is the emotional base of a politics in which everyone feels cheated and is sure someone undeserving is cheating them. Distrust of media and higher education follows. He finds universities almost magical, vaccine research among the proofs, and then hears people doubt whether higher education is good for the country. Rebuilding trust, he says, starts with honest conversation, or people will not buy it. Technology is not the same thing as capitalism. Workers in Bridges's era could demand a share. The postwar industrial contract tied wages to productivity and kept demand up. Trucking no longer has that contract, which is why a robot entering a weak labor market will not raise pay unless someone makes it.

Does he have to interview the two-hundredth truck driver. Yes, because he will learn something, and people with more power may talk to none, or to five, and decide they have it. He is pulled into expert rooms as a sounding board and feels the weight of that, but he is not those workers, and they may listen to the podcast and say he is full of it. A man who recognized Fridman from a podcast once told him that he and Joe Rogan only talk to Nobel types, never to "regular folk." The category is silly, Viscelli says; prizewinners are often more boring. The rule he kept is to talk to people anyway. Advice to the young is Steve Jobs's line that you cannot connect the dots forward. Viscelli went to graduate school over climate, read management studies about autonomy, then a popular piece on in-cab satellite computers, and wondered what monitoring does to a sense of control. He did not plan fifteen years in trucking. People told him he could not spend a year driving, or years on one big book. Looking back, those were the dots.

The fire to take leaps he traces to Tolkien and a teenage sense that you had to go out into the world. Some of those expeditions were stupid. Working on a forestry campaign, he drove a kayak to remote British Columbia with lake-and-Class-I experience, mistimed a tidal standing wave where two fjords met a cross channel, and later, on real whitewater with people who knew what they were doing, understood how unfit he had been. Trucking made him nervous; class comfort helped. A working-class kid at Colgate, obsessed with philosophy and Buddhism and a fish out of water among lacrosse teammates, learned to learn people. The meaning of life, when Fridman finally asks, is love, and you can find it in work, though not everyone can, and many jobs are a paycheck. For people with the luxury, love the thing you are trying to achieve, the impact that expresses love, whether that is a counseling center or sending someone to Mars. Fridman adds craft: mastery in a task others call menial, including driving a truck better than anyone. Viscelli had just had hundred-year-old floors refinished. The man who sanded them was an artist. That, he says, is craft, and that is love.