Transistor

The transistor didn't emerge from Bell Labs in 1947 because Shockley, Bardeen, and Brattain were geniuses—it emerged because the adjacent possible had finally aligned. Vacuum tubes, invented in 1904, had proven that electron flow could be controlled. Quantum mechanics, developed in the 1920s, revealed that semiconductors behaved strangely at junctions. Purified silicon and germanium, produced for radar during WWII, sat waiting in stockpiles. Bell Labs, facing the limits of mechanical telephone switching, had both motive and resources.

The convergence was so complete that multiple teams were racing. Purdue University was months behind. Westinghouse had parallel efforts. Herbert Mataré and Heinrich Welker in Paris demonstrated a working transistor in 1948, unaware of Bell Labs' announcement. This convergent emergence proves the invention was inevitable—the adjacent possible had opened.

The cascade was immediate and explosive. Transistors replaced vacuum tubes: smaller, cooler, more reliable, requiring no warm-up time. Texas Instruments commercialized the silicon transistor in 1954. Fairchild Semiconductor was founded by defectors who saw the opportunity. One Fairchild defector—Robert Noyce—co-founded Intel, which rode the microprocessor to dominance.

Each step created new adjacent possibles: the transistor made the integrated circuit conceivable (1958), the IC made the microprocessor conceivable (1971), the microprocessor made personal computing conceivable (1970s), personal computing made the internet accessible (1990s), the internet made the smartphone conceivable (2007).

By 2026, silicon transistors approach their physical limits at ~3nm. The adjacent possible now includes quantum computing and neuromorphic chips—but those inventions await their own convergence of conditions.