Down the street and up a small alley from the Sony Corporation's HQ in Tokyo's North Shinagawa industrial district is the company's other headquarters: its Computer Science Laboratory. The lab's set up in a modest building that sits across from traditional Japanese houses topped by wavelike roofs peaked with Lioli tiles. Like any number of Sony's facilities, it doesn't look too impressive from the outside - just another nondescript glass-and-metal box. But step inside.

Where most Sony building interiors are quiet and furrowed with rows of matching desks, here a cacophony emanates from congested rooms stuffed with humming electronic equipment, tottering towers of papers, and unclassifiable thingamajigs. At Sony's Shinagawa high-rise, most Sony Men wear navy suits, white shirts, and predictable neckties; Sony Women are generally turned out in drab ensembles. Yet CSL is in Technicolor: an engineer in paisley here, another in shocking yellow there, even a guy lurking the hallways in a purple trench coat. And while the artwork on Sony office walls - if there is any - tends toward cold abstractions, here, senior researcher Jun Rekimoto's lab sports Miles Davis posters.

Rekimoto works amid a spectacular composition of high wires, pulleys, and crisscrossed beams of projected light. Imagine Q's lab from a James Bond movie - if Q were forced to share office space with a tweedy professor. In the center of the room sits a table upon which two-dimensional iconic representations of dinnerware and a flowery centerpiece replace the real thing. It's a computer on legs, a 4- by 5-foot interface that can turn any flat surface into a PC desktop.

Using a laser pointer like a pair of chopsticks, Rekimoto reaches across the table, "picks up" a file folder, and places it on the wall opposite the Miles portraits. What he has actually done is transfer an MPEG movie file from a computer represented by the tabletop to another computer whose interface is a "wall" consisting of a silver screen held taut by an Erector Set-like partition. One day, Rekimoto says, what he calls ubiquitous computers will be embedded throughout our environment - controlled not by trackballs, mouseclicks, or URLs but by everyday gestures.

Opposite Rekimoto's lab, there's another overstuffed, seemingly impenetrable office. But someone is at work in there, a man with tortoiseshell Buddy Holly glasses, wearing a thick twill suit and a necktie strewn with Monet's flowers. It's Hiroaki Kitano, and he emerges from the disorder with a collection of photographs of robots under his arm - automatons that he tells me will one day compete in a soccer match with humans. When a visitor responds that soccer-playing robots are a quixotic fantasy, Kitano disagrees sharply. He insists that a team of robots will play the World Cup champions by 2050. And the date isn't arbitrary; Kitano has calculated a timeline of advances in robotics that predicts the year of the match. The outcome? "We'll play flawlessly," he says. "The robots will win." First Deep Blue, now this.

Down a parallel corridor, in a closet-sized space spilling over with oscillators, galvanic skin-response sensors, and heart and breathing monitors, Canadian-born Kim Binstead tries on her "hypermask." The mask, cadaverous and chalk-white, is equipped with sensors that allow a computer to monitor and track its location. A projected image of a face remains "attached" to the surface of the mask regardless of its position and angle, changing - front view, profile - as the wearer moves. With this technology, says Binstead, "one can wear someone else's face."

Throughout the Tokyo CSL, a sister lab in Paris, and a separate Distributed Systems Lab in Silicon Valley, similar experiments are under way. Many of them are merely curious; others conjure Asimov and Gibson. Some of them would be less surprising in a basement at Carnegie Mellon or MIT, but here they beg a question: What the hell does any of this have to do with Sony, world-renowned maker of TVs, camcorders, and Walkmans? It's a $60 billion question, and the flip answer is: everything.

True, Sony remains, at heart, a consumer-electronics business - a hardware company. But it won't be one for very much longer. In just four years, president and CEO Nobuyuki Idei has begun a sweeping transformation of Sony, so it can be as dominant a player in the digital universe as it is in the analog one. He's engaged a massive reorganization. He's backed the PlayStation 2, a platform so robust it may be all the computer some gamers and their families will ever need. He's recommitted Sony to computing - focusing, significantly, not on a single device but on an array of them that all talk to one another in a no-PC-required home network. And he's articulated a vision of how, once digitized, Sony media properties will finally be integrated with the firm's traditional business.

Above all, he's championing the obscure research of CSL tinkerers. It is a move at once symbolic (the message: Sony takes risks to invent the stuff we dream of) and practical (the AIBO robot dog sold out in Japan 20 minutes after it was made available). It's also an audacious business proposition: Idei is betting the store on people who spend their days and nights pondering questions like, "What goes on when people experience catharsis?" and "How does complexity emerge?"

When, in June, Idei flew to San Francisco for the opening of Sony's Metreon, an $85 million-plus entertainment complex, he had just come from a European business trip. He had 30 minutes to himself before a day of meetings that led up to the Metreon ribbon cutting, and emerged from his suite at the Sheraton Palace with dark circles under his eyes. Even so, he was uncharacteristically chatty as he entered a conference room.

Idei took a seat at the head of a long table lined with a couple dozen managers. His eyes scanned up and down one side of the table and then the other. At his right were Sony Men, suits from Marketing, Accounting, and divisions that make and sell the company's well-known consumer products. On his left were the geeks behind the new stuff - the Vaio laptop, the PlayStation 2. He smiled ever so slightly, and a moment later he let the gathering in on the joke.

"Here we are," he said. "Sony." Nodding toward one side, he said, "The conservative." Nodding to the other side, he said, "The progressive." He then looked ahead and said, "This is Sony now, our strength, our challenge." He shook his head.

An engineer who attended the meeting later observed, "Idei was clearly amused at what he saw: the potential clash. It was a coincidental division, but one that illustrated his new Sony, with its inherent potential for greatness, but also with its potential for - no, probability of - conflict and chaos."

When Idei was tapped to be president in 1995, almost everyone inside Sony felt he was a disastrous choice. According to Sony biographer John Nathan, Norio Ohga, the former president who picked Idei, said that 99 out of 100 people at Sony were appalled by the decision.

But Idei has nearly palpable confidence, and by 1998 he'd been named co-CEO with Ohga; he assumed full CEO-ship earlier this year. In 1997, he reduced the board of directors from 38 to 10 members and brought in more outsiders - traditionally unheard of at Japanese companies. This year he announced a 10 percent workforce reduction. More important, he began an instant deconstruction of Sony's hardwired internal structure. In order to make the organization pliable, he encouraged individuals to take the initiative, and divisions to share talent. Sony had to become, in his words, an "organism" that could learn and change in response to "the inspirations of independent thinkers." It could do this with "virtual companies" - ad hoc concept and product groups that work outside any existing structures. They could consist of 5 employees or 100 - whatever was needed - from anywhere in the company.

"In the year 2000 almost all our products will go to the Net without using the PC," says Sony president Nobuyuki Idei.

Driving the reorganization were the notions of emergent evolution and emergent behavior, concepts crucial to research at CSL. As the organism metaphors suggest, these theories come from biology and apply to life-forms that change through experience. In evolutionary theory, emergent evolution is the rise of a system that cannot be extrapolated from antecedent conditions. Emergent behavior, likewise, refers to unpredictable outcomes and applies to organisms that mutate, adapt, and overachieve.

In a small office in Sony's creatively titled Building 3, Toshi T. Doi - wire-thin, with graying hair, in a gray suit - arrives at a small, blank-walled office with a dog carrier and three assistants. On Doi's instruction, one of the K-9 specialists opens the tote's latch and carefully removes Sony's robot dog, AIBO (Artificial Intelligence Robot). He places AIBO on the gray carpet and it sits still, lifeless, like any inert battery-powered toy. But when he pats it on the head, the robot rouses, stretching, and bounds around almost like a capering lamb. "He's in a playful mood," the technician says.

Doi explains that AIBO, in performance mode, can be controlled by remote, but the autonomous function is far more exciting, since it incorporates emergent-behavior artificial intelligence. The robot reacts to stimuli and, in effect, learns, responding according to a programmed personality that develops freely.

AIBO's AI is rudimentary, but, Doi says, the potential is nearly limitless. It has a circumscribed ability to react with basic "emotions": delight, sadness, fear, and anger. It's programmed with attributes - likes and dislikes - and reacts based on them. AIBO "likes" pink, for instance, so it follows a pink ball, which it tracks with a color-sensing camera in its nose. When the technician hides the ball, the robot continues to look for it but then, like a puppy, grows bored after a while. As AIBO gets older, Doi says, it will have a longer attention span.

Hiroaki Kitano, who works on the cyberdog's AI, explains that the robot will follow a person who pets it - replicating attachment behavior - because it has been programmed to respond to attention. What it will make of other experiences, however, is somewhat unpredictable.

"It can do more than we could have ever programmed it to do," Kitano says. "It is impossible to predict all the possible reactions of a child playing with the robot, so the robot must build his knowledge from experience. It is impossible to predict the taste of every user, so the robot must learn to 'understand' it." When the technician knocks AIBO over, it looks hurt and growls. AIBO rights itself, shakes off the affront, and stretches and purrs. It walks away from the technician, "looking" back over its shoulder as if it is attempting to process its first experience of cruelty.

Doi might have created AIBO to satisfy his own curiosity, but Sony wouldn't have mass-produced a product that could do little more than yap and chase after a pink ball. As Doi reveals, there is a potential gold mine in the AIBO - suggested retail ¥250,000 ($2,500 in the US) - beyond the obvious novelty of a nifty, if pricey, toy. He envisions an AIBO industry, when the robot becomes a "walking game machine," a PlayStation on legs. With software designed by Sony game makers, as well as licensees, it is easy to imagine AIBO as a game system, becoming one's teammate or adversary in anything from soccer to chess.

"I think we may raise up an entertainment-robot software industry," says Doi, who also envisions applications beyond gaming. AIBO could retrieve email and the day's news, or even baby-sit. Programmed for the latter function, it could "watch" a child, maybe even entertain it with stories and tricks, and could send a message to parents when the child begins to cry or if an intruder appears.

Doi's visions don't stop there.

"I need an artificial muscle," Doi says. "I need skin that senses a variety of stimuli." For now, a dozen sensors are arrayed on AIBO's head, nose, footpads, and tail. In the near future, AIBO will respond to voice commands "heard" through microphones in its ears. Says Doi, "There's much we can do."

"Doi-san is inspired by God's creations," one of his colleagues says. "He strives toward the creation of creatures with beauty and grace. Some people think it's an affront to God, but to Dr. Doi it is a celebration, honoring Him."

Watching AIBO play, it's not surprising to learn Doi is the man most responsible for CSL's existence. In the 1960s and '70s, Doi, now 57, worked with one of Sony's famed founders, Masaru Ibuka, the genius engineer who developed early audio recorders. During the '70s, Doi encouraged Ibuka, cofounder Akio Morita, and their successors to move into computers. The results were early Sony PCs, including an 8-bit system released in 1982 just as IBM unveiled the 16-bit PC. The computer and an earlier Sony word processor were flops, though the latter project spawned the 3.5-inch floppy disk.

Sony made other attempts to break into computers - producing a multimedia PC and the NEWS workstation - but such devices were eventually surpassed by US manufacturers, and the divisions that built them were disbanded. Doi, however, encouraged then-president Norio Ohga to continue to pursue, if not computers, then at least the attendant technology.

"I knew that we needed a broad, large-scale initiative if we weren't going to be left behind," Doi says. "I knew we could continue to sell the Walkman and Sony televisions, but the future wasn't about stand-alone products. The future had computer science as fundamental." In 1987, Ohga gave him the OK to start the Computer Science Lab with a relatively small budget. His ambition for the lab was anything but small, however: "I told Ohga-san that we would exceed Xerox PARC."

The lab, at first, had two researchers and one manager, Mario Tokoro, an associate professor at Keio University. Tokoro, who worked part-time at Sony, was widely respected in the computer-science field, working decades ago on scenarios in which computer networks would transform everyday life.

"His research was unique, and I believed it had potential for Sony's future," recalls Doi. Tokoro built the Tokyo lab to its current size, 20 researchers, and started another in Paris with 5 engineers. "We drew on the skills of top computer scientists from around the world," Tokoro says. "That means paying extravagantly high salaries, depending on the engineer. We decided on yearly contracts - just like the ones paid to professional baseball players. If an engineer does not produce, he is let go. So everyone works very hard to show results."

Tokoro joined Sony full-time after Idei became president. "With Idei in charge, I felt Sony was moving to become an IT company, so I would be happy here," he says. Tokoro, who had become one of Idei's most trusted advisers, wanted his scientists to pursue pure research, but he also encouraged them to participate in the Sony culture and work in virtual organizations when they were inspired to do so. He explains: "I came over from academia partly because I was interested in the idea of practical applications of theoretical research."

With Doi, Tokoro set wide research parameters. Doi describes them as "the exploration of the billions of things that would be connected with the Internet, everything from PDAs to refrigerators." The researchers developed a wealth of computer technologies that were ready to be plucked and used in consumer products when Idei needed them. "We were there when Idei was ready to turn Sony into an IT company," Doi says.

"With Idei, finally Doi-san's vision was embraced," says Rodger Lea, a Sony computer scientist who caught Tokoro's eye when he was working in the UK on operating systems for HP. "His view was vindicated, and the computer researchers were viewed with appreciation."

After putting CSL in Tokoro's hands, Doi started his own lab, D-21, in Tokyo in 1996. D-21 focuses, among other things, on sophisticated speech-recognition research - tech that is key to AIBO, to Sony's new car-navigation system, and to several toys in development. Still, Doi's obsession was robotics, and he recently renamed D-21 the Digital Creatures Lab. "Doi-san was raised on cartoons like Astro Boy," notes an engineer in his lab. "He has always had a Jetsons dream."

While Doi pursues that vision, and perhaps develops the equivalent of Rosie, the housecleaning robot, colleagues back at CSL continue to focus on technologies integral to the Sony home network. Many of their breakthroughs are patented (last year Sony ranked in the top 10 companies receiving US patents), and all of them are designed for integration.

Applying emergent-behavior theory to Sony products, as well as to the corporate culture, Idei says, will enable the company to create new experiences greater than the sum of its parts. Or, as Idei puts it, "One plus one equals four." Everything Sony makes - whether a camcorder or a TV set - must be viewed as a component in a larger network. In other words, the old Sony sold stand-alone products, but the new Sony will sell connections.

To get there, Idei gave explicit marching orders when he arrived in 1995. In phase one, Sony products were to be transformed into digital devices (though, of course, huge-selling analog products will still be produced). The cassette and CD Walkmans became the MiniDisc Walkman. Eight-millimeter and Hi-8 handicams became digital-video handicams. VCRs became DVD players, and analog Trinitron TV sets became HDTV sets. There were new digital products, too: the Vaio laptop, digital cameras (the Mavica and the Cyber-shot), and cell phones.

In phase two, now under way, the stand-alone products become nodes - connected to one another, to other hardware devices, and to every imaginable type of content source. Norio Ohga's 1989 purchase of Columbia Pictures for $3.4 billion - criticized at the time for draining resources from Sony's main consumer-hardware business - fits right in with Idei's plan. Sony will deliver multimedia content to the home network running hardware, and, crucially, will sell the idea that Sony digital products can bring out the inner auteur in all of us.

To glimpse how this will all play out, leaf through the premiere issue of Sony Style, a new high-tone, full-color magalog devoted to Sony products, recording stars, venues, and previews. One section is unabashedly called a "manual for digital living"; another details how to make a feature film with Sony digital cameras.

To solve the home network's technical challenges - which involve software, not hardware - Idei turned again to Mario Tokoro and CSL. Processing content in real time between numerous Sony components, and those from other companies, required a new type of operating system - the project that is Tokoro's passion.

In fact, anticipating the needs of a new kind of network a decade ago, Tokoro set in motion a project to develop an OS that could play audio and video in real time. Tokoro, Hiroaki Kitano, and other researchers at CSL created the kernel of an OS named Aperios, which they handed off to the Sony Suprastructure Center, an R&D lab in Tokyo led by Akikazu Takeuchi. Takeuchi was assigned to take Aperios from a rough OS to a product, a massive undertaking that wound up requiring the attention of 100 software engineers in labs from Tokyo to Brussels to Bangalore. Meanwhile, Takeuchi asked Rodger Lea to establish Silicon Valley's Distributed Systems Lab, where Lea and another team of researchers are linking Sony products (the infrastructure) with its software (the "suprastructure") and loading the whole system with Sony entertainment.

Basically a space-age bachelor pad, Lea's lab in San Jose features a bunch of TVs, camcorders, MiniDisc players, and set-top boxes, all of which respond to one remote, not a pile of them. The set-top box supports Aperios applications.

Sony's ubiquitous computers will be "situation aware" - they'll do what they're supposed to without being asked.

The Aperios OS is revolutionary, Lea says, because it can do in real time what others can do only with stored data. One Aperios program, for example, can run a movie as it is downloaded from a satellite. (Windows must download the entire file before the OS can play the film.) Besides its real-time capabilities, Lea adds, Aperios solves storage problems. Apps built on the OS can use any number of components to store data - anything from a set-top box to a MiniDisc - using whatever is available within the network when it needs it.

Aperios also expands the number of sources of information you can use at once. As a result, when you're watching a movie and you wonder where else you've seen that actor, you could pause the film and do a quick Web search, using the actor's image itself if you can't remember his name. Before you go back to the movie, you might click another button on the remote to check in on the kids, viewing camcorder images (or, conceivably, baby-sitter AIBO's perspective) on a screen within a screen.

Beyond its versatility in running the home network, Lea boasts that Aperios has a number of custom interfaces, one of which can "zoom" left to right and up to down, and it "scales" - adds new functions - easily. "Say you add something new to your home network that is two years newer than the rest of the products and has features that weren't anticipated by the older version of the operating system you have," Lea says. "Aperios will detect it and upgrade itself and learn whatever new properties it needs - therefore giving it the ability to use the same features on the older products as well as the new one."

When Sony got into the OS business, an obvious question was whether the company was taking on everyone's favorite adversary, Idei's golfing buddy Bill Gates. Aperios has different functions than Windows does - basically, it controls content, whereas Windows runs the machine itself - but there are overlaps, and some cable companies are already deciding between set-top boxes that run Windows CE or Aperios. A few, like AT&T Broadband and Internet Services (formerly TCI), are electing to use both.

In any event, it almost goes without saying that first-mover status will be crucial as the digital home network becomes a reality. And in Japan, Sony set-top boxes running Aperios can already do more than receive traditional cable broadcasting: They pick up digital satellite TV (including Sky PerfecTV!, Sony's joint venture with News Corp. and Softbank) and download music, playing tracks in real time or storing them with Sony's ATRAC MiniDisc technology.

Of course, homes will always have appliances from many manufacturers. To be certain its products work with all the rest, Sony has helped develop HAVi - as in home audio-video interoperability - middleware that lets diverse systems communicate. Eight electronics firms, including Sony and Philips, have committed to HAVi as a middleware standard. HAVi, which can be updated each time you add a new device, will compete with the protocol of another alliance that includes Microsoft, Intel, Mitsubishi, and Honeywell - and Philips, which isn't choosing sides. Called the Home Application Programming Interface, or Home API, this rival software has a different basic premise - one that could be a tragic flaw, according to Sony's Akikazu Takeuchi. Whereas HAVi doesn't require a central controller, Home API relies on a PC to manage the other home appliances.

"In the year 2000 almost all Sony products will go to the Net without using the PC," Idei says. Which raises another question: If a home network doesn't require a central computer, why did Sony reenter the computer business with the Vaio PC?

Partly, Idei says, because a PC will be a component in many home systems - used in video editing, for example - at least for the short term. But a more fundamental motivation is that CSL's imagined home of the future will be filled with computers - but computers embedded in several other devices.

"The PC was born for business and productivity," explains Keiji Kimura, head of Sony's Information Technology group. "Everyone else brings a PC designed for business into the home. If we are going to be successful, it will ultimately be with a completely new PC concept, designed for the home itself."

Kimura's group created the Vaio, which, though a late entrant into the computer sector, has almost half of the laptop market in Japan and 15 percent (and growing) in the US. Yet in spite of the Vaio's success, even Kimura is quick to add that it's not the true home computer Idei envisions.

During his San Francisco trip, Idei elaborated on why Windows machines - including the Vaio, perhaps the sexiest slipcover Windows has ever donned - still aren't consumer products. He told the story of how that morning he had opened his Vaio and plugged his modem into the hotel's telephone line. Then he had turned it on. And waited.

"Consumer-electronics people know that is unsatisfactory," he says. "Before the PC will be a truly accepted consumer product, it must be as easy to use as a Sony telephone or camcorder. You plug them in and turn them on."

Next, to check his email, Idei clicked on the Windows dialup networking icon. It didn't work. From a hotel phone, he had to reconfigure the software - in this case, to tell it to dial 9 to make a call outside the hotel. This step, too, must be eliminated, Idei says. "Now, we have to download music. You don't download running water - you just buy a bottle. I don't think 'download' is a good idea. You just want the music. That's what we want to provide."

So, yes, Idei says, the Vaio is a critical part of Sony's future, but "it is unessential for many things." He intimates that in the future, the home PC will be the network itself. "We understand the home," Idei insists. "The PC companies don't. We will do it right."

Hiroshi Yoshioka - a 20-year Sony veteran who sped development of the progressively slicker versions of the camcorder: 8-mm, Hi-8, digital - is building on the home-network concept in Japan. The camcorder, he says, can work directly with the cell phone, for instance, so you can send video to your friends across the world. Similarly, the phone as an image-transmitting component may soon be superfluous; instead, you will send images from your camcorder directly to the Internet. The functions of many of these products, too, will become blurred. The new Cyber-shot can take video clips as well as still photographs; the Vaio PictureBook includes a built-in camera that captures and sends a picture or video as an email attachment. The idea is similar to Sun's plans for Jini-enabled office systems - a variety of peripherals that run together or independently - and this similarity has not been lost on the bosses of the two companies, who, along with Philips, have announced their intentions to make the Jini and HAVi standards compatible.

Another Sony creation, the Memory Stick, aims to replace the disk as the portable storage medium. The size of a piece of chewing gum, but able to offer more storage space than a CD, it can easily move data - music, video, and other types of information - between appliances. And Sony's components will connect via Apple-designed FireWire, which Sony has branded i.Link. Currently, Sony designers are busy ensuring that all its new digital products can take full advantage of the Memory Stick and i.Link, while making sure Sony doesn't lose its reputation for slim cool. (Speaking of which, Satoshi Suzuki has a concept for a new Walkman that is little more than a headphone with a Memory Stick.) Mitsuru Inaba, who manages the Sony Design Center (see "Product Revolution"), supervises a team of eight "superdesigners," and is the maestro of Sony cool.

Vaio designer Teiyu Goto, one of these superdesigners, likes to point out that he had the home network in mind when he named the Vaio. He looked not only to an appropriate acronym (originally Video Audio Input/Output) but also to the visual impact of the logo. The cursive "VA" resembles analog waves; the "IO" recalls a 1 and a 0, the digital building blocks.

Even CSL's most obscure inquiries provide foresight into the coming home networked entertainment system. Take Jun Rekimoto's table computer. While the all-in-one remote Rodger Lea demo'd is a step in the right direction, Sony eventually wants computer interfaces to be as simple as a light switch - or, better yet, as intuitive as a motion detector. Besides Rekimoto's projection-based system, there is work on alternative systems that will make computers "situation aware" and "assistance orientated": Computers will do what they're supposed to without being asked.

"They will be part of our physical world and as common as wristwatches and eyeglasses," Rekimoto says.

With a prototype device called the Navicam, Rekimoto has come up with a way for these computers to receive coded information. The Navicam scans personal icons - like those near the doors to each CSL researcher's office - as if they were barcodes. It reads the coded information from an IP network address and displays a variety of content in many media, including text - in the case of the personal icons, the researchers' names and descriptions of their interests - and graphical images.

The Navicam's basic technology is already used in Sony's PictureBook, and it will have a wide variety of other uses, particularly reading coded information beamed from cellular systems or global positioning satellites. The small device could serve as a route finder, using a navigational satellite, and could bring us relevant information depending on what we're doing and where we're doing it at any moment in any location on earth, Rekimoto says. To what end? "These systems will provide us with information, expedite communication, guide us, and attend to our needs."

Cybermask developer Kim Binstead, too, is working on a project that fits nicely into Rekimoto and Sony's vision of ubiquitous computing. It began with a philosophical query, she explains: What goes on when people experience catharsis? The answer will allow game designers to use the relevant psychophysiological feedback to influence the action. In other words, using Binstead's research, games of the future could adapt to one's heart rate, galvanic skin response, breathing, and other measures.

"If you have direct access to a user's state, you can have more relevant games," Binstead says. Beyond gaming, she adds, there are other entertainment applications: "An emotionally responsive Walkman playing what you most need to hear at any given moment might be kind of cool."

There are several immediate reasons Sony's share price is at an all-time high even as profits have plummeted: its brand, one of the best recognized in the world; the PlayStation, which accounted for 40 percent of the company's operating profits in 1998; and its digital products, such as the Mavica digital camera. But if you take the long view, it's Idei's vision that has revitalized Sony. "It's because Sony is once again in the hands of the people with the best ideas about the future," says Shizuo Takashino, manager of Sony's Home Network division. "It's because Idei has put his money on the thinkers."

"At Sony now, I can see the character of every person," says Idei's friend and adviser Mario Tokoro. "Sony is not a collection of people, but a collection of each one of the individual people. It is fundamentally why Sony can continue to keep its original creative spirit and change for the time."

"Most large companies don't do pure research anymore," Rodger Lea says. "It costs Sony a lot - $6 billion a year - but it's needed for a valid, long-term vision. US public companies, worried about their shareholders, can't afford it," Lea adds, "but Sony is thriving because of it."

"Most large companies don't do pure research - it costs a lot," says Rodger Lea. "But Sony is thriving because of it."

Idei says, "If you look at the incorporation prospectus for Sony put together by Masaru Ibuka, under whom the audio and video tape recorders were developed, and Akio Morita, who made Sony a world-class company, you will see a clear-cut adventure spirit. It's in our DNA. But times change, and we had to remake Sony.

"Sony is a personal-experience company," he explains. "Movies, Vaio, Walkman, PlayStation - they are not essential for human beings' survival, but maybe they are essential for the survival of our emotional side. To us, our challenge is to continue to be vital in people's lives in this way. We want to provide our dream for others. To do this, we must retain the minds of children, filled with wonder and curiosity, and we must keep in mind our goal: to give people the enjoyment that is like the enjoyment of children."