Imagine a tech company so ambitious that it tried to take an idea public. That was the basic pitch behind “concept IPO,” a label that fit General Magic perfectly.
The company was built by three founders who had all worked at Apple. Two of them, Andy Hertzfeld and Bill Atkinson, were already Silicon Valley legends for their work on the Macintosh. Atkinson was also behind key interface ideas like double click and the drop-down menu. The third founder, Marc Porat, was the one who seemed to understand where computing was headed before most people did.
For his Stanford PhD dissertation in 1976, Porat studied a century of change in the American workforce and argued that the economy was shifting away from transforming matter and energy through agriculture and industry, and toward transforming information. He saw computers and telecommunications reshaping every sector. “We are entering another phase in economic history,” he wrote. On the first page of his dissertation, he introduced a term that would become famous: “information economy.”
He expanded on that idea in 1980 with a PBS documentary, The Information Society, where he described information technology as a force as disruptive as the plow or the steam engine. He also warned about privacy, information overload, misinformation, and rising inequality. At the time, most Americans had little sense that the ground was already shifting under them.
By 1988, Porat had joined Apple’s Advanced Technology Group, where the job was to imagine the next major platform after the personal computer. One day, he taped a Sharp Wizard electronic organizer to a Motorola analog cell phone. That rough prototype gave him the spark for a new category. He later built plaster models of a combined phone and digital assistant. In 1989, he sketched a product in a large red notebook that matched the future with startling accuracy. He called it the Pocket Crystal.
The Pocket Crystal concept looked like a thin glass rectangle with no physical buttons, just a touch screen. It would combine a phone and fax machine, and eventually be used for text messages, movies, video games, plane tickets, and downloadable apps. It would fit in a pocket and feel beautiful. Porat wrote that it should offer “the kind of personal satisfaction that a fine piece of jewelry brings,” along with the tactile comfort of a seashell and the enchantment of a crystal.
In 1989, only 15 percent of American households had a computer, and none of them were portable in the modern sense. The web did not yet exist. Still, Porat was essentially drawing what would later look like the iPhone.
The project was approved, but with one major problem: it was too large for Apple to absorb directly.
Mobile phones were still bulky and basic. Turning the Pocket Crystal into reality would require not just new hardware and software, but also the networks and communication standards needed to connect devices at scale.
In 1990, Porat and Apple CEO John Sculley agreed that Apple would invest and take a board seat, while the project would spin out as a separate company and begin looking for outside partners. The new company was named General Magic, a title meant to suggest both the power of major technology companies and the idea that advanced technology can feel like magic.
Sculley introduced the founders to Sony, and the pitch landed quickly. Sony invested and signed a licensing deal. Motorola followed, then AT&T. Soon the biggest names in telecom and consumer electronics were joining what became known as the Alliance. Philips came next, then Panasonic, NTT, Toshiba, France Telecom, and others, each investing millions. The partner group became so large and so powerful that Alliance meetings had to begin with an antitrust lawyer reminding everyone what they were not allowed to discuss. By the company’s own legal definition, it was one of the largest global business consortia ever assembled in the United States.
The momentum around the company, and the reputation of Atkinson and Hertzfeld, pulled in talent from across Silicon Valley. Former Apple employees joined, including Joanna Hoffman, who had led Mac marketing, and Susan Kare, who designed the Macintosh graphics. Jane Anderson, Sculley’s personal publicist, also came aboard. The company attracted young engineers with a strong desire to build something big, and some were even said to have slept outside the office in hopes of getting hired.
General Magic was custom-building nearly everything, and the pace of invention was intense. The team worked on an early form of USB, touch screens with a virtual keyboard, skeuomorphic graphics that represented functions like a filing cabinet or game room, and even a virtual street with a store for downloading new applications, an early app store concept. Messaging features included stickers and animated characters, which were early precursors to emojis. There were also interactive graphics, such as lips you could tap to record a voice message. The company developed Telescript, a programming language designed to let devices communicate across a virtual network for information sharing and ecommerce. The team called that remote environment “the cloud.”
When two young employees, Steve Jarrett and Tony Fadell, traveled to Japan to demo hardware for Mitsubishi Electric, the presentation was interrupted. Jarrett later recalled that a group of executives entered the room in factory uniforms and asked Fadell to explain the architecture again. When Fadell said the device had no modem because everything had been done in software, one executive was said to have buried his head in his hands. “You have just obsoleted Kato-san’s division,” the organizer reportedly said.
Back in California, the atmosphere was euphoric. The company seemed certain it would change the world. The work never stopped, with one programmer reportedly sleeping under his desk and asking whether meetings were “A.M. or P.M.” But the office still had a startup feel. A rabbit and a parrot wandered around freely. Water fights were common. At one point, a window was broken by slime launched from a giant slingshot.
The combination of vision, talent, money, and partners was intoxicating enough that General Magic moved toward an IPO even though it had very little revenue. One banker famously tried to persuade Porat to take part by offering him the shirt off his back, and then removed his suit in the middle of the office. In 1995, Goldman Sachs took General Magic public. The stock doubled on the first day of trading.
Porat later said that one of his goals in raising so much money so quickly was to create “heaven for engineers.” They were free to imagine, play, invent, and write. For engineers, he said, that was the best possible environment.
In the end, though, that freedom was part of the problem.
General Magic did not become the iPhone before the iPhone. Its rise was dramatic, but its collapse was even faster. All of the innovation inside the company never turned into a simple product that solved a real problem.
When the first device finally launched, it was a book-sized machine running an operating system that drained the battery. In six months, it sold only 3,000 units, many of them to friends and family. A year after the IPO, the stock had fallen 85 percent. Early employees were leaving, and Alliance partners began to pull away. Porat stepped down as CEO. What had once looked like a rocket ship for the information economy had blown up on the launchpad.
General Magic had every ingredient that should have produced success: a visionary leader, a talented team, a culture that encouraged experimentation, major funding, and extraordinary technical depth. Its alumni later became leaders at Apple, Google, Adobe, Samsung, and elsewhere. They helped build Safari, the iPod, the iPhone, the Apple Watch, Dreamweaver, Google Search, and AI at Apple. Former “magicians” also helped found LinkedIn, Nest, and Android. One of them later became the first female chief technology officer of the United States.
So why did the company fail so badly?
Part of the answer is unusually well documented because General Magic hired in-house videographers to record the effort as it happened. One of them, Sarah Kerruish, was 27 when she joined the company. She later said it felt surreal to land there, with the buzz surrounding the company, the unconventional office culture, and the sense that something historic was underway. She also met her husband there, Steve Jarrett.
More than a decade after the company collapsed, Kerruish dug through the footage and turned it into a documentary, General Magic, released in 2018. The film captures both the brilliance and the failure. Her view was blunt: the company had too much creative freedom and too much money.
That freedom showed up in the engineering approach. General Magic built almost everything from scratch while pushing hard in multiple directions at once. One engineer compared it to hand-building a car instead of using a factory. Another said the team was not standing on the shoulders of giants; it was building the giant from the ground up. That was exciting, but it was also risky.
Even with all of that internal innovation, the technology needed to make the Pocket Crystal real was still years away. The project kept slipping, and the company never settled on a smaller, clearer intermediate product. As Megan Smith later put it, the hardest part was deciding what not to do.
Because the team had the ability to do almost anything, it often did. If an engineer had a cool idea, it was likely to get built, even when it did not matter to the product or pushed deadlines further out. Kerruish recalled hearing marketing complaints late at night about features like a “walking lemon” animation while the device still had battery problems and instability. Those conversations showed how hard it was to impose discipline on such an open-ended effort.
Andy Hertzfeld described the same problem in a more personal way. He said he was still in creative mode, working on a flipping coin for the game room, even while the company needed to focus on the unglamorous work required to ship a product. He felt he was setting the wrong example.
Even the core target user was never clearly defined. The team talked about building for “Joe Six-pack,” but that was not a real product strategy. It did not provide a concrete set of needs, and it is not even clear that the typical consumer of the time had email. Tony Fadell later said the company lacked a clear message, a clear product definition, and clear answers to basic questions like who it was for, what it would cost, and when it would ship.
Another scene in the documentary shows Darin Adler, who had already led work on a Mac operating system, trying to explain the need for accountability. Engineers sat on the floor listening as he described moments when someone had to take responsibility and tell others to get to work. The response, according to Adler, was that managers were not needed because the company already had Andy and Bill. That lack of formal structure was treated as part of the magic.
At the same time, the company had to manage 16 global partners, many of whom were waiting to manufacture devices. General Magic went public partly to reduce its financial dependence on those partners, but the IPO also raised the stakes. More money brought more engineers and more scope. The project kept getting larger when it needed to become smaller. In the documentary, the IPO feels less like a milestone and more like a warning sign. As Michael Stern later said, the company had too much money and never had to worry about payroll, which is unusual for a startup.
In the end, the first device running General Magic’s operating system did not arrive until more than four years after the company and the Alliance were formed. It had an unusual interface built around a grayscale office, with a desk, calendar, and inbox. From there, users could move into a virtual hallway, a library, a game room, or “downtown” to browse stores and apps.
The interface was imaginative, but the product was overloaded. It required a 200-page manual, ran poorly on the hardware of the time, and delivered a choppy experience. It was bigger than an iPhone and did not include a phone. It had to be connected to a phone line to make calls or send messages. It was portable only in the loosest sense, and it was expensive. Demand was effectively nonexistent. The product was ahead of its time, but it did not solve a pressing consumer problem. Meanwhile, the internet exploded, and the idea of a proprietary cloud lost its relevance.
General Magic never used its environment to create useful constraints, and it never narrowed the project enough to focus on a clear user and a clear timeline. It had too much money, too much talent, and too much freedom. As one venture capitalist put it, startups more often die of indigestion than starvation.
