Teams and open networks built every layer of computing

1

Poetical science first

Lovelace showed in 1843 that a calculating engine could manipulate symbols, music, and images—an imaginative leap Babbage’s unfinished hardware never matched alone.

Isaacson opens on Ada Lovelace, Byron’s daughter, steered toward mathematics yet fused it with poetic range. Annotating Babbage’s Analytical Engine, she argued the machine could follow rules for any symbol, not only numbers, and named that fusion poetical science.

Her notes included what is often called the first computer program: an algorithm for Bernoulli numbers. She also split a mere computer from a programmable machine that could switch tasks—an conceptual cut Babbage did not finish into working metal.

She did not work alone. She translated, annotated, argued, and extended his designs in constant correspondence. Funding collapse and Babbage’s gift for alienating patrons left the engine unbuilt; the chapter plants the caution that vision without institutions stalls for generations.

In practice: Babbage held the engineering ambition and Lovelace the symbol-processing leap, yet neither alone produced a working machine, and the world waited roughly a century for the pieces to reassemble.

2

Logic becomes switches

Shannon’s 1937 thesis made Boolean algebra buildable from electrical relays, and that insight spread only because it entered funded research networks Zuse never reached.

George Boole’s nineteenth-century algebra of true and false sat as philosophy until Shannon, at MIT and Bell Labs, showed relay switches could enact any logical or arithmetic operation. Isaacson treats the thesis as a quiet pivot: abstract reasoning gained a hardware blueprint.

Wartime urgency in the US and Britain pulled mathematicians and engineers into shared military-academic projects that could test and scale the idea. The same years held a stark foil.

In Germany, Konrad Zuse built binary calculating machines largely alone, cut off by war from Allied research traffic. Impressive machines, thin later influence. Isaacson uses the split to show that proximity to peers, money, and rival teams often decides whether an idea changes the world or stays in a drawer.

In practice: Shannon’s switches became a shared blueprint inside Bell Labs and wartime labs; Zuse’s comparable binary machines, built in isolation, did not steer the immediate path of electronic computing.

3

Programmers without credit

ENIAC’s 1945 success was as much the six women who invented plugboard programming as the hardware Mauchly and Eckert got famous for.

ENIAC at the University of Pennsylvania calculated artillery trajectories for the Army. Isaacson gives long attention to Kay McNulty, Jean Bartik, Betty Holberton, and three peers who wired plugboards and switches with no manuals and no precedent, inventing practice by trial and error.

They earned less than male counterparts, vanished from a famous 1946 public photograph’s credit, and dropped out of most histories for decades. Hardware took the glory; the work of making it useful disappeared.

Legal aftermath undercut singular origin stories further. Mauchly and Eckert’s patent fight ended in a 1973 federal ruling that key ENIAC concepts derived from John Atanasoff’s earlier Iowa State work. Isaacson positions ENIAC as one node among overlapping university, military, and corporate efforts that borrowed and disputed credit.

In practice: The six ENIAC programmers were excluded from the 1946 photo’s public credit even after physically creating the machine’s first working instruction sequences.

4

The lab as inventor

The 1947 transistor emerged from Bell Labs’ deliberate mix of theorists and experimentalists under one roof, not from one man’s solo flash.

William Shockley, Walter Brattain, and John Bardeen produced a solid-state replacement for bulky vacuum tubes. Isaacson centers the institution: Mervin Kelly housed physicists, chemists, metallurgists, and engineers in shared corridors and cafeterias expressly to force contact.

Bardeen and Brattain worked as an experimental pair; Shockley, more theoretical and famously difficult, pushed and sometimes blocked them. The device came from friction and correction. Shockley was initially off some key steps, then maneuvered for credit and his own junction design—ego and lab politics shaping whose name stuck.

Shockley left for Palo Alto to found Shockley Semiconductor. Recruited talent, including Robert Noyce and Gordon Moore, fled his management to found Fairchild Semiconductor in 1957, inadvertently seeding Silicon Valley. Government and corporate labs could absorb long, expensive uncertainty small firms could not.

In practice: Kelly’s cafeteria-and-corridor rule put unlike specialists in daily collision; the transistor came out of that collision and the later defections from Shockley’s firm.

5

Spinoffs as the engine

Fairchild’s repeated employee departures—the Fairchildren—plus stock options and venture capital made disloyalty the Bay Area’s innovation mechanism.

Robert Noyce at Fairchild and Jack Kilby at Texas Instruments independently built the integrated circuit in 1958 and 1959, each solving how to link transistors on one piece of material without fragile hand-wiring. Noyce’s planar method, from colleague Jean Hoerni, proved more manufacturable and became the standard.

Isaacson tracks what followed as the larger story. Fairchild became a training ground people left to found new firms. Noyce and Gordon Moore quit in 1968 to start Intel with Andy Grove; other alumni launched National Semiconductor and AMD.

Arthur Rock’s style of stock options and venture capital let engineers leave without losing everything. Ideas and people circulated across rivals even as firms competed—collaborative competition unlike East Coast loyalty culture at IBM. Noyce’s open floors and unreserved parking at Intel became a wider Valley template.

In practice: Noyce and Moore’s 1968 exit from Fairchild to Intel, carrying Grove, was one turn in a diaspora that also produced AMD and National Semiconductor.

6

Garages outrun labs

The personal computer came from Homebrew hobbyists and counterculture sharing, while Xerox PARC invented the GUI then failed to commercialize it.

Isaacson moves to the 1970s Homebrew Computer Club in Menlo Park, where Steve Wozniak and others swapped schematics under openness norms borrowed from 1960s counterculture. The Apple I grew from designs Wozniak already shared; Jobs pushed the Apple II in 1977 into a polished consumer product instead of a kit.

Xerox PARC had already created the graphical user interface, the mouse, and networked personal computing in the early 1970s. Xerox management did not bring them to market. Jobs’s 1979 PARC visit fed Apple’s Lisa and Macintosh windows-and-icons work—originators and shippers often diverged.

Bill Gates and Paul Allen founded Microsoft in 1975 around a BASIC interpreter for the Altair 8800. Gates’s 1976 open letter against hobbyist software piracy marked an early clash between hacker sharing and proprietary control. Isaacson frames the era as a contest neither ethic wholly won.

In practice: After the 1979 PARC visit, Apple shipped Lisa and Macintosh with windows and icons while Xerox still sat on its own research.

7

Packets without a center

The internet worked because ARPA funded a decentralized packet design and kept TCP/IP open, a Cold War choice that later allowed permissionless build-out.

Isaacson traces origins to the Defense Department’s ARPA in the 1960s. J.C.R. Licklider pushed interactive networked computing for human collaboration; Bob Taylor and Larry Roberts built ARPANET, linking four university nodes in 1969.

Paul Baran and Donald Davies, working independently, broke messages into packets that could take any route and reassemble, surviving single-point failure. Contracts spread across university labs and firms like BBN so many hands set standards without one commercial owner.

Vint Cerf and Bob Kahn’s mid-1970s TCP/IP let separate networks merge. The protocols stayed non-proprietary—the decision Isaacson calls decisive for later growth. Free-market Valley lore often erases that government-funded open plumbing. Design meant to ride out nuclear war also fit unplanned innovation by anyone with a link.

In practice: ARPANET’s first four university nodes in 1969 ran on packet routes and later open TCP/IP that no company could lock.

8

Openness then control

Each computing layer repeats the same loop: open collaborative platforms spark growth, then proprietary players consolidate, and the tension never fully ends.

Tim Berners-Lee built the World Wide Web at CERN in 1989 with HTML, HTTP, and URLs to link documents across the existing internet, then gave the work away unpatented—echoing TCP/IP’s open ethic. Mosaic, from Marc Andreessen and Eric Bina at Illinois’s federally funded supercomputing center in 1993, made the web visual and public.

Andreessen co-founded Netscape, which dominated briefly until Microsoft’s bundled Internet Explorer, backed by the Windows monopoly, displaced it and drew a major late-1990s antitrust case. Isaacson reads the sequence as the book’s pattern made plain: openness grows the pie; growth invites consolidation.

The same arc covers cheap transistor licensing, Xerox missing PARC’s spoils, and Larry Page and Sergey Brin’s PageRank at Stanford, shipped as Google in 1998 on the crawlable web Berners-Lee refused to lock. Isaacson notes the cycle rolling into social media and mobile, still unsettled.

In practice: Mosaic’s 1993 release triggered mass adoption; Netscape’s lead then yielded to Microsoft’s bundled Explorer and antitrust court.

9

Ingredients over heroes

Major digital breakthroughs arrived nearly simultaneously across teams, so Isaacson prescribes funded cross-discipline institutions instead of waiting on isolated genius.

The close gathers Lovelace and Babbage, the transistor trio, ENIAC’s programmers, the Fairchildren, Jobs and Wozniak, Berners-Lee and Andreessen. Isaacson holds that credit myths distort how to organize next.

Recurring structural pieces include ARPA-style government money for long-horizon work markets underfund; labs like Bell and PARC that force discipline mixing; Bay Area tolerance for failure and for leaving firms; and pairings of visionaries with executors who grind ideas into products. Connectors such as Noyce or Jobs often mattered more for assembly and translation than for originating every core technique.

Simultaneous invention—the transistor, the integrated circuit, packet switching—counts as evidence that shared intellectual environments reach readiness together. His prescription is concrete: fund basic research, build institutions that compel collaboration across lines, and stop concentrating credit and control in single heroic names.

In practice: Packet switching, the transistor, and the integrated circuit each had independent teams reach the result near the same time, which Isaacson reads as environment maturity rather than one unrepeatable mind.