The Next Small Thing

So Hawkins took a job at GRiD Systems, a small Silicon Valley company that was exploring the frontiers of portable computing. GRiD had a simple, concrete business goal: to design a computer that you could actually carry. Hawkins wanted to go further: to design a computer that would actually respond to human impulses. And that meant figuring out the cognitive roots of those impulses. By day, he wrote code. But at night and on weekends, he immersed himself in trying to "map" the brain, in seeking the headwaters of intelligence and following their course through the learning process.

Over time, an interest became a passion. "My wife thought I was nuts," Hawkins says. "She worried that I was following in my father's footsteps. She suggested that I align myself with a university."

In 1986, Hawkins decided to take his wife's advice. He looked to the Bay Area's two flagship universities, Stanford and the University of California, Berkeley. The people at Stanford said he would have a hard time finding an intellectual home there. The people at Berkeley were somewhat more encouraging, so he decided to enroll in its graduate program in biophysics. "In three years," he says, "I became convinced that I had discovered and defined what intelligence is. It sounds wild, but I really believed it.'' (See the sidebar "This Is Jeff Hawkins on Brains," page 104.)

Unfortunately, Hawkins couldn't persuade any of his professors to champion his work. He left Berkeley in 1988 without a PhD - but with a big idea. His research into neural networks focused on how the human brain - and, by extension, the digital computer - could recognize patterns. He even developed an algorithm for handwriting-recognition software. He called it "PalmPrint." It has shaped everything that he's done since.

Hawkins left Berkeley jazzed and confused - jazzed about what he had discovered, confused about how he would earn a living. So he re-upped with GRiD, this time on different terms. He licensed PalmPrint (which he had patented) to GRiD and became its vice president of research, charged with developing pen-based hardware and software. Later that year, GRiD was purchased by Tandy, an electronics manufacturer and the parent company of Radio Shack.

A year and a half later, Hawkins and his team unveiled the GRiDPad, the world's first serious pen-based computer. The machine was big, slow, clunky, ugly - and a revelation. This could be the future of computing. Trade magazines raved about it. Companies descended on GRiD with proposals for smaller and faster versions.

A gold rush ensued. An array of high-tech giants - IBM, NCR, NEC, Samsung - announced plans to launch the next small thing. Kleiner Perkins poured money into GO. Apple, which had been developing a handheld device since 1987, shifted into high gear: In January 1992, then-CEO John Sculley coined the term "personal digital assistant," or PDA. The gold in the gold rush had a name.

For his part, Hawkins was not all that enamored of his invention. GRiD built computers for niche business customers: insurance adjusters, railroad inspectors, oil-rig workers. These were also the customers for GRiDPad. Hawkins wanted to build small, powerful, general-purpose computers. He wanted to change people's lives.

"It was about then that I decided I didn't want to be at Tandy,'' Hawkins says. "I knew what I wanted to do, and I was in a position to do it on my own."

Thus was born Palm Computing. It was January 1992. The good news: Hawkins had an idea and a company. The bad news: He had no plan, no product, no financing, and no partners (except for Tandy, with which he had negotiated a licensing agreement before leaving the company). But he did have a reputation. He met with Bruce Dunlevie, a venture capitalist who sat on the board of Geoworks, a software outfit that writes operating systems for portable computers. "I could tell right away that Jeff was a 'product picker,' '' Dunlevie says. "That's someone who is able to synthesize where technology is today and then advance it - someone who knows intuitively what people care about."

Dunlevie's firm, Merrill, Pickard, Anderson & Eyre, bid to be Palm's lead investor. Hawkins also talked with Sutter Hill Ventures, another prominent VC firm. He accepted $500,000 from each of them, plus $300,000 from Tandy. His dream had funding, along with a few partners. Now he just needed a plan and a product.