Frank had lost most of his savings in a bank crash, but his wife somehow scraped up enough money to help Andy reach M.I.T. on a partial scholarship. Andy graduated in 1940 with a bachelors degree in chemical engineering, mathematics, and premedicine and worked for the Bendix Corporation and several other companies before founding Non-Linear Systems in 1953. Kay reached California through a job, although for years he’d been hoping to enjoy the warm climate there. He started his company in Solana Beach rather than Silicon Valley because he was already working in the area. Osborne himself would found his own firm while living in Berkeley—he’d locate it in Hayward, a medium-sized city within twenty miles of the Valley proper, the Palo Alto-San Jose area southeast of San Francisco.
For Non-Linear Systems Kay developed the world’s first commercially successful digital voltmeter. It was the result of some memories from World War II, when unskilled workers were constantly damaging voltmeters and having trouble reading them accurately. Kay’s new machine easily survived the careless. The instrument displayed numbers and didn’t require workers to make a reading off the thin needle of a conventional meter adjusted for the right voltage. “I don’t look at myself particularly as an inventor,” Kay said. “I’m a solver of problems. There are inventors who have many, many more ideas than I get.” Instead, he tried to refine ideas in “ways that make sense from an industrial-engineering standpoint and a customer standpoint.”
Kay’s digital voltmeter was perfect for the military and for the National Aeronautics and Space Administration (NASA) and their contractors; unskilled people could help test thousands of voltages on American missiles. “The silos are full of our instruments,” he said. But the NASA cutbacks of the Nixon era hit Non-Linear Systems hard.
“The seven vice-presidents went,” recalled a trade publication, “as did their white Cadillacs.” Frank Kopischiansky was willing to help see Non-Linear Systems through bad times, loaning thousands of dollars of painfully saved money. But that was hardly enough to sustain a corporation. What about the future? How to avoid the boom-and-bust cycle of government spending?
Kay decided to carve out a niche as a maker of rugged portable test gear for private industry. He wanted to offer many of the same wrinkles found in costly, bench-bound equipment, and one of those features was a microprocessor in a voltmeter: a miniature electronic brain that would make it easier to use. The idea came from an old friend, Bill McDonald, whom Kay had met decades earlier at M.I.T. Their gym lockers were next to each other’s there; both men had loved handball and hated team sports. McDonald, a husky man of basketball-player height, had worked for Non-Linear Systems in the early 1950s, then had returned East, where he had acquired a computer background. Then, in the late 1970s, he’d rejoined Non-Linear Systems, where his friend Kay hoped he would help develop a computer. Correctly or not, Kay believed that “ninety percent of the electronics business was computer and ten percent was test equipment, hi-fi, stereo, and stuff like that.” Non-Linear Systems had offered other companies’ minicomputers in systems it packaged for military contractors, but still wasn’t making machines of its own. Maybe a small business machine would be one way to enter computerdom.
“Two months after Bill came to work for me,” Kay recalled, “he showed me some microprocessor-based computers for doing accounts payable-receivable at the local computer dealer’s place.
“They were very reasonable, so I bought some and was very intrigued with them. And I wondered: ‘Why not make this stuff? I could make a better disk drive. And look at this: this connector doesn’t work too well.’ And Bill said, ‘We could make them, all right, but I don’t know where you’d sell them.’ This was in late 1979. I said okay and dropped the idea.”
Two years later, however, Kay watched his architect son-in-law, Michael Batter, struggling with an Apple and knew the time was ripe for an easier-to-tote portable.
Kay tried VisiCalc, the electronic spreadsheet, on the Apple and came away thinking the machine was too complicated, beyond being a hassle to move from home to office. So Kay told Bill McDonald, “Listen, we’re selling instruments to engineers. Let’s make a computer that we can walk in to an engineer and set on his desk and say, ‘Hey, look at this. You can use it.’”
Still, Kay worried about the new technology. “It’s a fairly stylized dance,” McDonald reassured him in spring 1981. “You buy the chips, and there is black magic in getting them to work—but not in how they work.”