EVENT · 01
The Apple I & the Homebrew Club: Hacker Ethic Meets Product WillApple I & the Homebrew Computer Club · 1976
Apr 1976Cupertino, CADawn of Personal Computing
Background & Key Figures
In January 1975 the Altair 8800 made the cover of Popular Electronics — the first microcomputer kit aimed at individuals. Bay Area hobbyists promptly formed the Homebrew Computer Club, first meeting that March, devoted to the hacker creed that information should be freely shared. Steve Wozniak, a 26-year-old HP engineer, was obsessed with building circuits from the fewest possible chips; 21-year-old Steve Jobs smelled a business.
What Happened
Woz single-handedly designed the Apple I — integrating processor, keyboard, and display onto one board, doing away with rows of toggle switches. He meant to give the schematics away, as the club did. Jobs insisted: turn it into a product and sell it. In April 1976 they sold their first batch of about 50 units at $666.66 each. Woz built the machine; Jobs built the company — and personal computing acquired, for the first time, a commercial form.
Counterfactual + Historians' Debate
Walter Isaacson's Steve Jobs (2011) notes the real fork was not technical but between the hacker's sharing ethic and Jobs's product will. Counterfactual: had Woz simply published the design, personal computing would still have spread through the Altair ecosystem — but it would have lacked the step of polishing technology into a consumer product anyone could use. The debate (see Steven Levy's Hackers): was the PC a bottom-up hobbyist movement, or the deliberate shaping of a few entrepreneurs?
Modern Parallel
The tension between open-model communities (Hugging Face) and polished consumer products (ChatGPT) today is the same story replayed: sharing spreads technology; productization delivers it to ordinary people.
One-Line Lesson + Question
Sharing a good thing and turning it into a product are two different virtues — most revolutions need the collision of both.
That good idea in your hands — is it better open-sourced and shared, or polished into something everyone can actually use?
EVENT · 02
The IBM PC: An Openness Born of DeadlineThe IBM PC & the Open Architecture Accident · 1981
Aug 12, 1981Boca Raton, FLOpen Architecture
Background & Key Figures
By 1980 the Apple II and a crowd of micros had Big Blue rattled. IBM built closed mainframes on multi-year cycles. To move fast, a team of about 12 in Boca Raton, led by Don Estridge, was authorized to bypass corporate process with one goal: ship within a year.
What Happened
For speed they used off-the-shelf parts: an Intel 8088 chip, an open bus, an outsourced operating system. In 1980 IBM approached a small firm, Microsoft; Bill Gates had no OS, so he bought QDOS for $50,000 and reworked it into MS-DOS — the crucial move being that he licensed rather than sold it outright. IBM could use it; so could anyone else. The PC shipped in August 1981, and its open architecture let Compaq and others legally clone it. "IBM-compatibles" soon flooded the world.
Counterfactual + Historians' Debate
Counterfactual: had IBM stayed closed like its mainframes — proprietary OS, locked bus — short-term profits would have been higher, but there would be no clone ecosystem, and neither Microsoft nor Intel would have risen to control the standard. IBM traded openness for a market explosion and handed dominance to Wintel. The debate: Paul Ceruzzi's A History of Modern Computing argues this openness was no master plan but a deadline-driven expedient — one that accidentally forged the industry standard.
Modern Parallel
Open the platform to win an ecosystem, or close it to protect profit — the Android-versus-iOS choice still re-runs the exact dilemma IBM faced.
One-Line Lesson + Question
A compromise you make for speed can redefine the power structure of an entire industry.
Where are you trading control for "fast" — and might an ecosystem grow there that ends up defining you?
EVENT · 03
Windows 95: The GUI Enters Every HomeWindows 95 & the GUI Goes Mainstream · 1995
Aug 24, 1995Redmond, WANetwork Effects
Background & Key Figures
The graphical user interface (GUI) originated at Xerox PARC and was first commercialized by the 1984 Apple Mac. Yet into the early 1990s the vast majority of computers still ran the command-line DOS, and Microsoft's Windows 3.x was little more than a shell on top of it.
What Happened
On August 24, 1995, Windows 95 launched — Start menu, taskbar, plug-and-play — putting a GUI into ordinary hands for the first time. Microsoft spent heavily on marketing (the Rolling Stones sang "Start Me Up" for it), and people queued worldwide to buy it. It also bundled the IE browser, laying the fuse for later antitrust suits. Almost overnight, the "personal computer" turned from a geek's toy into a household appliance.
Counterfactual + Historians' Debate
Counterfactual: had Microsoft failed to seize the desktop standard in the mid-1990s, the GUI war might have been split among Apple, IBM's OS/2, or an early Netscape. Windows won not by being technically best (the Mac was more elegant) but by ubiquitous compatibility and pre-installation — network effects overwhelmed experience. The debate (e.g. Martin Campbell-Kelly): was Microsoft's dominance a triumph of innovation, or lock-in that suppressed it for years?
Modern Parallel
The power of the "default" has not changed — pre-installed, bundled, ready on boot remains the most hidden and most decisive lever in platform competition.
One-Line Lesson + Question
What decides the winner is often not the best one, but the most widespread and hardest to replace.
The tools you rely on — is it because they're genuinely the best, or simply because "everyone uses them and switching is too much hassle"?
EVENT · 04
The iPhone, 2007: Putting the Computer in Your PocketThe iPhone & the Post-PC Turn · 2007
Jan 9, 2007San Francisco · MacworldTechnological Convergence
Background & Key Figures
By 2007 the PC paradigm was thirty years mature. Phones belonged to Nokia and BlackBerry — small screens and physical keyboards. Apple, revived by the iPod, was holding Steve Jobs's bigger move.
What Happened
On January 9, 2007, Jobs told Macworld: "Three products — a widescreen iPod, a phone, and an internet device… these are not three separate devices." Multi-touch, full screen, a desktop-class operating system stuffed into a pocket. The real detonation came in 2008 — the App Store opened to third-party developers, and the phone turned from a communication tool into a universal computing platform. The PC did not die, but the center of gravity of computing moved to the palm.
Counterfactual + Historians' Debate
Counterfactual: had Apple stayed closed like most phone makers and never opened the App Store, the iPhone might have been just another pretty handset — the open developer ecosystem was the detonator (yet another openness-versus-closure choice). The debate: was the mobile revolution Jobs's vision, or the inevitable convergence of capacitive touchscreens, lithium batteries, and low-power ARM chips (see Day 7) all maturing at once? Most historians of technology (e.g. Brian Merchant's The One Device) lean to the latter — a great man standing on an already-ripe technology curve.
Modern Parallel
Is AI in a similar "iPhone moment" — the underlying capability already mature, waiting only for a product form that folds it naturally into everyday life?
One-Line Lesson + Question
A true revolution is often recombining what already exists, not conjuring something from nothing.
In your field, which "already mature but not yet combined" parts are waiting for an iPhone moment?
Questions to Sit With
Q1: Is the history of the PC a pendulum between openness and closure?
From Woz wanting to publish the Apple I, to IBM's accidental openness, to Microsoft's bundled closure, to the iPhone's closed hardware plus open store — every link swings between open and closed. The pattern may be: openness accelerates diffusion and wins ecosystems; closure captures profit and controls experience. The real winners switch precisely between them — closed underneath to hold margin, open above to borrow the crowd's strength. To read any act of openness, first ask which layer of the value chain it opens.
Q2: What does the role-swap of IBM and Microsoft tell us? (echoing Day 7)
In Day 7, AT&T was forced by antitrust to open the transistor patent; here IBM opened its architecture to hit a deadline. Both were an incumbent's openness, and both handed dominance to a nimbler challenger (Fairchild; Microsoft/Intel). The lesson: openness is often the start of losing control for the incumbent, but a siege weapon for the challenger. The same act means opposite things depending on where you stand in the chain.
Q3: At each paradigm shift, where does the locus of power migrate?
In the mainframe era power sat with hardware giants (IBM); in the PC era it moved to OS and chips (Wintel); in the mobile era to platforms and app stores (Apple/Google). The pattern: when one layer is commoditized, value moves up to the layer controlling the gateway and the standard. Extrapolate: if intelligence itself is commoditized, does power shift to whoever owns the data, the distribution gateway, or the human-collaboration interface? That is the core bet of today's AI competition.
Q4: When can "better" beat "more widespread"?
Windows beat the more elegant Mac — a victory of network effects and lock-in. Yet the iPhone toppled the ubiquitous Nokia with a better experience. The difference: when switching costs are high and network effects strong, ubiquity is a moat; when the paradigm itself leaps and old network effects reset to zero, "better" gets its chance. To disrupt an incumbent, don't compete on quality in the same lane — open a new lane where its network effects no longer count.
Q5: Did Jobs make the iPhone, or did the era make it?
The great-man view stresses Jobs's product instinct; the convergence view (Merchant) stresses touchscreen, lithium battery, and ARM chip all ripening at once. They needn't conflict: the era laid out every part, but it took an obsessive integrator to assemble them into something coherent while others couldn't yet see it. For AI practitioners the implication: the parts (large models, compute, multimodality) may be in place; what's scarce is the judgment to integrate and define the form.