General Magic Collapsed Because It Could Not Exclude Possibilities

David Epstein calls General Magic “the most important company nobody’s ever heard of,” not because it succeeded but because its collapse taught a generation of technologists what unconstrained ambition can do to a company.

Its central failure was not a shortage of capital, talent, or ideas. It had so much of each that it could defer the hard work of deciding what not to build. General Magic was developing a “personal communicator,” and almost any good idea someone proposed could be folded into the project. The result was substantial innovation—precursors to USB and emojis, among other things—but no durable focus.

The company defined its customer as “Joe Sixpack,” which Epstein treats as effectively no definition at all. Without a specific customer or a firm product boundary, the organization had little basis for rejecting work.

That was, in Epstein’s telling, the sentiment expressed by roughly three-quarters of the former General Magic employees he interviewed: they could not figure out what to leave out. Abundance made prioritization feel unnecessary until the accumulation of possibilities became unmanageable.

General Magic had unusual reason to believe it could build nearly anything. It was founded by three former Apple employees, two of whom had designed the original Macintosh. Its CEO, Marc Porat, had worked inside Apple on the question of what might come after personal computing.

David Epstein describes Porat as a visionary whose 1976 Stanford dissertation coined the term “information economy” on its first page and anticipated both technology’s promise and dangers involving automation and misinformation. In 1989, Porat drew a thin, buttonless glass rectangle with a touchscreen and rectangular apps—a device intended to serve as a phone, computer, fax machine, ATM, game machine, and messaging tool. The web did not yet exist, and only 15% of American households had computers.

That vision attracted money and talent at a scale that made General Magic’s ambition plausible. Goldman Sachs took it public in what Epstein calls Silicon Valley’s first “concept IPO”: the company had an idea rather than a product. It also assembled a 17-member alliance of international telecommunications companies, large enough that meetings began with an antitrust lawyer explaining what participants could not discuss.

The company had not merely imagined a broad future device. It had acquired the credibility and capacity to turn nearly every promising extension of that device into active work.

An engineer named Steve Perlman was assigned a calendar function for the communicator. He initially wrote it to run from 1904 to 2096, checked in the code, and considered the task complete.

A team leader then argued that someone might create historical applications, so the calendar needed to extend further back. Perlman changed it to run from year one into the future. Another team objected that this still anchored the feature in an arbitrary religious context. The calendar, they said, should begin at the start of astronomical time.

Perlman rewrote it to run from the Big Bang into the future. Epstein says the change consumed months, whereas the original 1904-to-2096 range would have taken four lines of code.

The episode illustrated the operating logic of General Magic. Porat said he had raised so much money to create heaven for engineers: a place where people were free to create and limited only by imagination. Epstein’s conclusion was that the company needed the opposite.

General Magic’s stock price doubled on its first trading day, Epstein says, and was worthless two years later.

General Magic’s legacy lies less in its own accomplishments than in what former employees took from its collapse. Alumni went on to co-found LinkedIn, eBay, and Nest; create Android; and work on the iPod, iPhone, Google Maps, and Safari. For Epstein, the consequential inheritance was a learned appreciation for limits that require choices.

David Epstein identifies Tony Fadell as perhaps the employee most marked by the experience. General Magic was Fadell’s first job out of college, filled with people he regarded as rock-star heroes. Its implosion was, Epstein says, traumatic. Fadell later led the design of the iPod and co-founded Nest, where he made constraints an explicit operating practice.

While researching constraints, Epstein was connected to Fadell by venture capitalist Bill Gurley. Gurley’s formulation was that “more startups die of indigestion than starvation,” a line Fadell immediately claimed as his own. The exchange did not settle its origin, but the phrase fit the General Magic problem: excess can destroy a startup when it prevents the company from ranking its opportunities.

At Nest, Fadell turned that abstract lesson into a decision rule. He made the team work inside a literal box and required it to prototype the packaging before the product. The package was what the end user would see on a shelf. If the team could not fit the product’s message and priorities into that space, then the idea was not a priority and belonged on the back burner.

The exercise did not merely constrain industrial design. It forced a decision before engineering effort could expand around every possible feature: what must the customer understand at the moment of purchase, and what can wait? The box created a practical test General Magic lacked.