There's a substantial community of true believers -- many of them technology entrepreneurs in their 60s, many with beards -- who contend that sometime within the next few decades, death will cease to be an inevitability. Perhaps this comes from having read a great deal of science fiction. Perhaps it is a natural response of those faced with mortality. Or perhaps, having seen the dramatic changes of the past few decades -- remember, these are people who helped invent the personal computer, genomics, and the Internet -- they have concluded that a cure for death might reasonably follow.

In the gospel of this near religion, salvation is called "the singularity," and its promise lies in the development of advanced robotics and artificial intelligence. In The Singularity Is Near, the entrepreneur Ray Kurzweil argues that, by 2030 or so, we will be able to merge our brains with artificial intelligence and live forever. Kurzweil and a bevy of like-minded proponents -- Google is a co-sponsor, with NASA, of a new Singularity University -- have banded together to study these developments.

Nolan Bushnell is an aging entrepreneur with a white beard -- and, as one might expect, he is sympathetic to the dreams of singularitarians. He believes that robots, of one kind or another, will be central to our lives within two decades. Ask him to imagine the future, and he will tell you about cell phones implanted in our ears and MP3 players controlled with buttons under our tongues. And, of course, cars will drive themselves. "There are so many benefits," Bushnell says, as if to marvel at the fact that it hasn't happened yet. "Accident rates go to zero. You can get home when you're drunk. Kids can get to school without parents. Gas mileage goes up." He could go on and on.

But, unlike committed believers in the singularity, Bushnell also has deep-seated doubts about the future. He worries that regulations will derail the next generation of inventors, and that venture capitalists -- or "lemming capitalists," as he calls them -- will systematically ignore important ideas in pursuit of short-term returns. Much as he'd like to have his Prius run on autopilot, he doesn't know if it will happen in his lifetime. "It all depends on how big the government will get, how pervasive regulation will be," he says. "Right now, the pace of technological growth is just not fast enough."

It's easy to miss this pessimistic strain when you first meet Bushnell, a big man with a baritone voice and a wry smile that oozes self-confidence. He founded Atari in 1972, when kids in their 20s didn't start big, ambitious companies, and sold it four years later for $28 million. It was Bushnell who gave Steve Jobs his first real job, as an Atari engineer, helping pave the way for Apple Computer and a generation of Silicon Valley companies. Bushnell would go on to start 20 businesses of his own -- including the improbably successful arcade and pizza parlor Chuck E. Cheese's Pizza Time Theatre -- and he's still at it today, running a high-tech restaurant chain called uWink, whose patrons order food and play video games using touchscreen kiosks at their tables. Last year, Paramount announced that it would produce a movie based on his life. Leonardo DiCaprio is set to star.

But Bushnell has seen enough to know how bad things can get, and in his office he keeps a beige android called Topo, a monument to his worst failure. Even in its decrepit state, the 3-foot-tall robot is a remarkable piece of machinery. Its squat, angular body sits on two splayed wheels that look like legs. The robot is surrounded by a clutch of other toys, tools, and memorabilia. There's the talking teddy bear Bushnell took to market in 1985 and a poster for the video game Pong, which he created in 1972. There's a picture of him sailing around the Hawaiian island of Oahu in the Transpacific Yacht Race and a black-and-white reproduction of a 1979 cover of Inc., which features Bushnell and a human-size rat in a bowler hat. (That would be the original Chuck E. Cheese's mascot.) There's a soldering iron and a voltmeter, nods to many days and nights spent tinkering with televisions.

But the centerpiece is Topo. The robot no longer works. In fact, it never really worked. "If a computer crashes, it doesn't break anything," Bushnell observes. "But when one of these went haywire, it was not a pretty thing." Topo cost Bushnell more than $20 million and wiped out most of a massive personal fortune. He went from living in a $6 million mansion in Woodside, California, to a modest house in Los Angeles; from his own Learjet to Southwest Airlines tickets. Worst of all -- because Bushnell never cared that much for money -- it led to a legal battle that caused him to miss the entirety of the dot-com boom. In short, it ruined him.

But Bushnell is OK with that. He keeps Topo with him to remember that business is a struggle and that the struggle is part of what makes it fun. "There's some shitty parts to running a business," he says. "But at the same time, creating stuff is neat."

For as long as he can remember -- for even longer than he's been in love with robots -- Nolan Bushnell has considered himself an entrepreneur. He was born in 1943 in Clearfield, Utah, a small town on the Great Salt Lake. His mother was a schoolteacher, and his father owned a 12-employee cement contracting business. He got his first lesson in business at age 10, when he figured out how to replace a broken tube on his parents' television set. Impressed, they let him go door-to-door showing off his TV-repair skills to the neighbors. At the time, a TV set cost more than a month's salary for most families and wasn't the kind of thing that was typically entrusted to a child. So Nolan set his price for a service call absurdly low -- just 50 cents to open up a TV, compared with the regular price of $5 -- and, like any mechanic worth his salt, made his margin on parts. "I would mark the hell out of the tubes," he says. "I could make 15 bucks in an hour."

Bushnell talks about these childhood exploits excitedly and without irony, as if hawking overpriced cathode-ray tubes before high school and, in his late teens, running games at an amusement park outside Salt Lake City were his life's work. He can still rattle off the economics of the TV-repair business or tell you the midway's revenue per customer. (Under his management, it was $1.80, which, he says, was the highest in the country at the time.) In college, he guessed weights and ran knock-down-the-milk-bottles games, even as he was learning how to program million-dollar mainframes at the University of Utah. "I could have stayed in the amusement park business after I graduated college," Bushnell says, "but I knew I could go back anytime."

He never really left. In 1970, after a brief stint at a Silicon Valley electronics company, Bushnell struck out on his own with a plan to make a better arcade. "The idea for Atari came to me while I was at college, playing math games on the University of Utah's huge general-purpose computer," he wrote in the 1979 Inc. cover story. "If I could just put a coin slot on this machine, I thought, I'd have a business. …But at $4 million apiece, the computers were too costly." When he started seeing advertisements for minicomputers that cost just a few thousand dollars each, he quit his job and got to work.

This story makes sense in retrospect, but at the time Bushnell's success hardly seemed certain. Video games were popular among computer geeks on college campuses and in corporate research parks, but they were unknown to most everyone else. "Nolan started when there was nothing to suggest that television-based arcade games could amount to an entire industry," says Steve Wozniak, the co-founder, with Steve Jobs, of Apple Computer. Bushnell's first arcade game was based on an MIT spaceship simulation that allowed two players to shoot missiles at each other. He named it Computer Space and licensed it to Nutting Associates, an arcade company in Mountain View, California. It was the first commercial video game.

Computer Space sold reasonably well -- Bushnell was paid $150,000 in royalties in 1971 -- but not as well as he had hoped. "It should have been three or four times that," Bushnell says. From this, he concluded two things. The first was that video games needed to be simpler. Controlling a spaceship's thrust and rotation proved too complicated for a kid at a bowling alley or a half-drunk guy at a pool hall. The second lesson was that he was perfectly qualified to start a company himself. "I think that one of the reasons Silicon Valley creates so many entrepreneurs is, you work next to someone who has started a company and you say to yourself, 'That guy has been successful and he's a dumb shit,' " Bushnell says. "I saw these guys at Nutting who were not smart folks, and yet they had this successful company. So I said to myself, 'I couldn't possibly screw up more than them.' "

He launched Atari -- the company name referred to a move in the Japanese board game Go -- with a friend and his Computer Space earnings. He distributed pinball machines to pay for the development costs of his first game, Pong. He designed Pong not the way a computer scientist would have built a game but like a hustler on the midway. It was simple -- just a few lines on a dark background and a single-sentence rule book -- and it was instantly understandable to anyone over the age of 3. Bushnell sold the game to arcade distributors and in a home edition at Sears stores.

By 1976, Atari was booking $40 million a year in sales, and its consumer business was exploding. Bushnell wanted to take the company public to finance the production of a new home gaming system, but then, worried that investors might not know what to make of a company like Atari, he accepted a $28 million acquisition offer from Warner Communications. Though an IPO surely would have made him richer -- and though Bushnell has spent no small amount of time questioning his decision since -- the $14 million he took home for four years' work seemed like plenty. Video games had arrived, and the father of video games had become a hero to a burgeoning group of ambitious tech entrepreneurs.

As with many people who have lived their lives in the public eye, it is hard to separate Nolan Bushnell from the story he has fashioned for himself, the one about the rebellious geek who makes good. This is especially difficult in Bushnell's case, because his personal mythology has become something of a cliché. We are no longer surprised when we meet wildly successful entrepreneurs who are young, technically savvy, and cocky. Yet it wasn't always that way.

In 1976, the very idea of a fast-growing start-up -- forget the notion of a 33-year-old, pipe-smoking, whip-smart playboy of a founder -- was somewhat novel. "The thinking was that a growth company shouldn't be trying to grow by any more than 20 percent a year," says Trip Hawkins, who founded the video game giant Electronic Arts and who now runs the game company Digital Chocolate. When he read about Bushnell's success while in Stanford's M.B.A. program, Hawkins realized his own entrepreneurial dream was possible. "Here was this young guy who'd had an amazing success and had a big-ticket sale to a big, legitimate public company," Hawkins recalls. "Nolan's the guy who blazed the trail for me, proving that you could be successful and create a lot of value quickly. He was my John the Baptist."

What admirers like Hawkins couldn't see was Bushnell's trepidation as he built Atari. "I felt absolutely alone," Bushnell recalls. "Everybody I talked to would say, 'You've got to replace yourself, you've got to find somebody with experience." He hired and fired three successive presidents. "I said to myself, 'I may be fucking things up, but not as badly as these guys are." To compensate for his lack of experience, he projected boundless self-confidence and spent money wildly. A 1977 story in Time titled "The Hot New Rich" described Bushnell's yacht -- named Pong -- his Mercedes, his 15-acre estate, his ski cabin, and his sexual gusto. "I have phone numbers in a lot of cities," he bragged to the reporter.

All this made for a beguiling character study. Bushnell was a young man who swore like a carny -- hell, he was a carny -- and yet, with his trademark pipe and blazer, he looked straight out of a Princeton reunion. (Bushnell picked up that jaunty affectation in college and never got tired of the joke.) He acted as if he didn't care much about business, and yet you couldn't deny his success. "There's the old Hollywood cliché about suits and ponytails," says Hawkins. "It was suddenly OK to have a ponytail in charge of a company." Wozniak, who at the time was helping Steve Jobs start Apple, says much the same: "Nolan touched the spirit of so many who dreamed of the connections between technical skills and entertainment. He inspired much of what followed."

Indeed, it's tempting to draw a psychic line to the great ponytail entrepreneurs of the next 30 years: to Netscape's Marc Andreessen in 1996, barefoot and bethroned on the cover of Time, or to Facebook's Mark Zuckerberg, who in 2005 was handing out business cards that said, "I'm CEO…Bitch." At some point in the late 1970s, entrepreneurship went from being a means of wealth accumulation to being -- at least in a few special cases -- a countercultural art form.

Of course, Bushnell never asserted any of this. What he did say -- over and over again to anyone who would listen -- was that business is a game. In fact, almost any activity for Bushnell is a game, a fact I learned when we nearly ran out of gas on Los Angeles's 101 freeway. We were driving between one of his restaurants and his office, and Bushnell, over the objections of his wife, Nancy, who sat in the back seat, had been trying to see if he could get 500 miles on a full tank in his Prius. About five miles from our destination, the car lurched; Bushnell pulled into the right lane as we crawled on battery power to a gas station. We made it -- barely -- and Bushnell was pleased with himself. "How many miles does your car get, Nancy?" he asked his wife, who responded with a tired smile, as if she had heard the line too many times.

When he's not pushing the limits of his Prius, Bushnell plays chess. At any given time, he is playing a dozen or so games on his iPhone. He plays against his friends, his employees, his kids, and his board members. Every 15 minutes or so throughout the day, he takes out his phone, stares at the screen, fiddles for a second, mutters an obscenity or two, and finally puts it away and picks up wherever he left off. "I used to be good, but I'm really sucking right now," he confides as he analyzes a game against his 26-year-old son, Gavin, who works as an assistant manager at the uWink restaurant in the Los Angeles suburb of Woodland Hills.

Despite this obvious love for games, Bushnell is no hard-core gamer -- at least not in the way the term would be understood by contemporary geeks in Atari throwback T-shirts. In the years since he launched the industry, video games have moved from the midway to the home, from the kind of thing you might do on a date to the kind of thing you do in your underwear because you can't get one. The 1990s witnessed the rise of increasingly realistic shooting games like Grand Theft Auto, which encourages players to sleep with prostitutes, assassinate people, and peddle drugs. Bushnell abhors this development and has suggested that Grand Theft Auto is borderline pornographic. "Nolan takes a social approach, as opposed to a techno-fetish approach," says Will Wright, founder of the game company Maxis and designer of the games SimCity and Spore. "I think what he did with Pong was ask what kind of social experience would a video game lead to."

In that light, it's easy to understand how Bushnell could go from something as hip as Atari to something as wacky as Chuck E. Cheese's. Bushnell figured that the games he had been marketing to bars, bowling alleys, and pool halls needed a venue that was cleaner, safer, and more fun. In Chuck E. Cheese's, he imagined a video game company that would control the means of distribution -- in this case, pizza parlors with arcades inside -- just as the old-style movie studios had owned the theaters. Pizza was only the pretext. As far as Bushnell was concerned, he was starting the new Disney. 

Of course, Bushnell's vision, which began with a single San Jose pizza parlor opened in 1977 as a subsidiary of Atari, was decidedly different from the goodhearted earnestness of Disney. The company's mechanical mascot, who entertained diners with a sort of cabaret routine, was the eponymous rat, Chuck E. Cheese. He was a large, surly, cigar-smoking smart aleck -- a robotic incarnation of Bushnell himself. What Chuck E. Cheese's offered was a massive arcade where parents could deposit their children and then enjoy some adult time -- that is, if you ignored the robotic dog doing Elvis impersonations and focused instead on the fact that the place served wine and beer.

After taking a look at the odd experiment, Warner did the only sensible thing: It ordered Bushnell to shut it down. "They thought it was the stupidest idea ever," he recalls. Bushnell protested, and Warner sold him control of Chuck E. Cheese's for $500,000. "Everybody believes in innovation until they see it," Bushnell says. "Then they think, 'Oh, no; that'll never work. It's too different."

From 1979 through 1984, Chuck E. Cheese's grew explosively -- if not always profitably. The company landed twice on the Inc. 100 list of the fastest-growing public companies. It opened several hundred locations across the country. In 1983, Bushnell began preparing for a confrontation with Atari: He launched a video game division called Sente -- the only move in Go that beats an atari -- and promised that he would once again be the king of games. He launched a start-up incubator called Catalyst Technologies. He bought himself an even bigger sailboat -- he piloted it to a fourth-place finish in the Transpacific Yacht Race in 1983 -- and jetted among his homes in Woodside, Aspen, and Paris. He was on top of the world.

From such heights, a fall of some kind must have been inevitable. But it's still hard not to be amazed when you visit Nolan Bushnell today and discover this 66-year-old entrepreneurial legend in such modest circumstances, working in a small office in a poor neighborhood in Los Angeles as he tries to turn around his struggling restaurant chain.

Until I visited, Bushnell had never taken a reporter to his company's headquarters, probably with good reason. UWink occupies two little suites in a dun-colored cinder-block edifice surrounded by check-cashing stores and fast-food spots. There's no receptionist or even a lobby to greet visitors -- just a cluttered workshop full of computer equipment. Because Bushnell typically meets with visitors at one of the company's three restaurants, he's never bothered to keep things presentable. "It's about as low class as you can get," Bushnell says. "But who cares? It's cheap, and my ego doesn't need it. I don't think we'd sell one more hamburger if we were on the fifth floor of some office tower."

Indeed, Bushnell seems to take a certain pleasure in his comparative poverty, bragging about his company's reliance on family labor -- four of his eight children and his wife work at uWink. "I'm a big believer in nepotism," he says, before introducing his own blue-collar definition of the concept: "You tell your kids they have to come work, you treat them like shit, and you don't pay them very much because it's good for the family." As to the trappings of wealth, he doesn't miss them either. "There are a lot of things about having money that are perceived to be cool but that aren't," he says. "Maybe if you're a CEO jerk who likes going coast to coast by himself in a G4, then that's fine. But that's not me. And it never will be."

Of course, Bushnell could have protected his money and his reputation. He could have written books, or become a VC, or hit the lecture circuit. He could have focused exclusively on managing Chuck E. Cheese's and become, like so many great founders who start their careers as builders and end them as administrators, a perfectly adequate CEO. In other words, he could have stopped being an entrepreneur.

Instead, he doubled down, launching a spate of new companies, each of which blended inventiveness with a kind of freethinking that at times seemed close to insanity. There was Etak, which made the first navigation system for cars, a precursor to modern GPS navigation; it was sold to Rupert Murdoch for $50 million in 1989. There was Axlon, which created a talking plush toy called A.G. Bear and sold $15 million worth of them in 1985. Not that these businesses didn't produce clunkers, too. There was, for example, Tech Force, a line of toy robots that were to be controlled by sound waves coming from a television show.

But the craziest idea by far was the quixotic plan to bring robots into people's homes. Bushnell founded Androbot in 1982 with the promise that for a few thousand dollars, he would be able to sell you a robotic companion. It would talk to you, play hide-and-seek with your kids, watch for burglars, maybe grab you a beer out of a specially designed Androbot cooler. He financed the company out of his own pocket, launched it at the 1983 Consumer Electronics Show in Las Vegas, and told the world that an IPO was imminent.

Just six months later, the market for technology stocks crashed, and Bushnell's underwriter, Merrill Lynch, canceled the Androbot offering. That left Bushnell on the hook for roughly $5 million in loans, taken out against his Chuck E. Cheese's stock. Meanwhile, the pizza chain, which had been losing money in its fervor to expand, found itself unable to raise additional capital. Bushnell resigned as CEO in the fall, and within a year, Chuck E. Cheese's would be bankrupt and Bushnell's stock would be worthless. Chuck E. Cheese's emerged intact -- today it is an $800 million company with 500 locations -- but Bushnell's finances did not. He had to keep borrowing to keep Androbot alive, and by the time it was all over in 1985, he owed Merrill Lynch $23 million, which he would spend the next decade trying to pay back.

Bushnell's earnings from Etak and Axlon got him most of the way out of debt, but the legal wrangling continued. Merrill sued Bushnell in 1995 over a $500,000 promissory note, just as he was attempting another company, a futuristic arcade called E2000. He countersued, arguing that Merrill had been partly to blame for the Androbot debacle, but the E2000 investors pulled their money out anyway. Suddenly, the man who had been starting companies his whole life couldn't get one off the ground. "I totally missed the dot-com bubble," he says. "Do you really think that I wouldn't have had a few dot-com companies?" He breaks into a fierce grin -- the only time I have seen Bushnell look genuinely bitter.

In 1998, Bushnell gave up. He sold the home where he and Nancy had raised their kids and used the proceeds to settle with Merrill. "We had to make a decision about whether to continue litigation or just end it," says Nancy. "We could have kept fighting, but life's too short." Nolan was ready to start companies again.

Even after all that pain, Bushnell still can't help loving robots. "Twenty years from now, I just can't believe the world isn't going to have mobile robots running around doing stuff," he says. "I think it should have happened 10 years ago, but that's another story." He likes the idea of being able to sit on his couch, watch TV, and order a perfectly made cup of tea without bothering his wife, his kids, or anyone else. And he thinks robots could dramatically increase our society's standard of living, making workers more productive and removing the need for menial labor.

For Bushnell, robots also represent the ultimate entrepreneurial challenge. "The world is stupid in the short term and smart in the long term," Bushnell often says, and he seems to understand entrepreneurs as the mediators of this disconnect, the people who must drag the world from stupid to smart. Whether he wants to invent a new kind of restaurant or integrate robots into human society, his first step is always to close his eyes and imagine the world as it might be. "I try to get a vision of the future, and then I try to figure out where the discontinuities are," he says. Once he has found a place where his vision of the future differs from the prevailing wisdom, he will start a company.

Of course, Bushnell has been wrong about the future more times than he has been right. UWink is in its third iteration after two brushes with death. After 9/11, Bushnell lost $6 million worth of business and had to lay off his entire staff, as all the orders for the company's video game kiosks disappeared. The company had another near failure at the end of last year, when its stock price fell almost to zero and its three restaurants began losing money. When I visited the uWink restaurant in Woodland Hills at lunchtime on a Wednesday in early January, the place was eerily quiet, with only a handful of diners tapping away at their little screens.

And yet, when Bushnell shows up, wearing blue jeans and puffing his pipe, he is downright jolly -- collapsing economy be damned. In between fragrant rips of smoke and bites of a pepperoni pizza, he explains that, although the lunch crowds have abandoned uWink, his restaurants are still packed evenings and weekends. He's taking the company private -- he says it will be profitable this year -- and he's spinning off a second company that has already sold dozens of kiosks to a Chili's restaurant in the Fort Lauderdale-Hollywood International Airport and to a senior center in Lancaster, California. In short, he's still happily scrambling.

Bushnell becomes even more animated as he talks about a start-up called Gamewager. Two kids from Texas are developing a system that allows video game players to compete remotely with one another for prizes. It's a virtual arcade. They cold-called him, he took the call, and he eventually agreed to provide them with a small amount of seed capital, since the company fits nicely into his vision of game play as a social experience. "They're so excited," he says of his newest protégés. "They're full of hope and optimism, and they're happy to sleep on my son's couch."

Two weeks later, I sit down with the Bushnells in Nolan's office. Nancy suggests that we not talk too much more about the dark years -- the bankruptcy of Chuck E. Cheese's, the failed IPO of Androbot, the loss of their home -- but it's hard to avoid Androbot when you're sitting at eye level with a 3-foot-tall android. The conversation inevitably makes its way to robots.

"I love them, but I'm not allowed to build them," Bushnell says, looking at his wife. "Nancy put me in a robot 12-step program."

I ask how many robot companies he's started.

"Two," he says, definitively.

"Only two?" I ask.

He revises, "Well, three."

"What about the talking teddy bear?"

"Oh. That's perilously close, too." (In fact, the number is either four or five, depending on, among other things, whether or not an animatronic rat counts as a robot.)

At this point, Nancy excuses herself, looking queasy. I start to ask another question, but Bushnell interrupts me. "I just want the future to happen faster," he says breathlessly. Coming as it does in Nancy's absence, this sounds almost like a confession. "I can't imagine the future without robots," he adds, before quoting the pioneering computer scientist Alan Kay, who once said, "The best way to predict the future is to invent it."

Bushnell goes on: "I actually think the technology is here. If I really -- " He drops his voice and cranes his neck toward the hallway to see if he's in the clear. "She's out of the room now, but I'd love to take another run at it," he says. "I really understand the technology, and in some ways, I've already proven the market." Bushnell is serious, but he's smiling, too -- still the hustler on the midway, now offering himself another go.

In the spirit of marital harmony, I don't mention Nolan's plans when Nancy returns. Instead, I ask him if the game is still fun after all these years.

"Absolutely," he says. "Always."

Max Chafkin, Inc.'s senior writer, wrote about the dating website Plenty of Fish for the January/February issue.