Apple Doesn’t Have an AI Problem. It Has a Games Problem.
And it’s been the same problem for 25 years.
Update, July 2026: This piece has been revised since original publication. The economics section originally overstated the PC market at $271 billion and understated Apple’s dollar share; it now uses IDC’s ~$170 billion traditional-PC figure, under which Apple holds ~9.5% of units and ~19% of dollars. The iPhone section has also been reframed: teen survey data shows iMessage — not the Mac — anchors US iPhone share, which makes Apple’s moat narrower, and the generational risk sharper, than I originally argued. A later pass added supporting data on gaming as a PC purchase driver (Pew, Statista) and tightened several causal claims. The thesis stands; the evidence beneath it is now stronger.
Wall Street has spent eighteen months working itself into a froth about Apple’s failure to “win in AI.” Apple Intelligence shipped late and thin, and the iPhone is now apparently a dumb terminal for somebody else’s models. I think this is mostly noise. Apple Intelligence will get to good-enough on its own clock, and most iPhone customers will never notice. The AI gap is a quarterly narrative looking for a catalyst. It is not the structural problem. The distinction is time horizon: on a five-year view, AI features may well move the needle more — but the AI gap is self-correcting, while the gaming gap compounds, and on a twenty-five-year view only one of them threatens the franchise.
The structural problem is that Apple has spent twenty-five years handing the PC gaming market to Microsoft. That miss constrains Mac TAM, locks an entire generation into Windows for life, and — the part nobody is discussing — quietly leaves the iPhone franchise undefended on a generational timeline. The questions these facts demand have never been put to Apple’s leadership in a public setting. They should be.
The head start Apple gave away
The clean start date is June 2000, when Microsoft acquired Bungie — the studio behind Marathon and Myth, both Mac-first franchises — to anchor the original Xbox. Halo, which Steve Jobs himself had unveiled onstage at Macworld a year earlier, shipped instead as an Xbox launch exclusive in 2001 and became the console’s killer app. Apple shrugged.
That shrug has compounded for twenty-five years. Microsoft spent roughly eighty billion dollars making gaming a strategic pillar — Activision Blizzard, Bethesda, Mojang, plus Xbox and Game Pass. Sony built PlayStation into a thirty-billion-dollar business with a stable of studios producing The Last of Us, God of War, and Spider-Man — works that win the cultural conversation, not just the sales charts.
Apple’s equivalent output: the Pippin (1996, dead on arrival), Game Center (2010, abandoned), Apple Arcade (2019, a family-friendly subscription explicitly aimed away from serious gamers), the Game Porting Toolkit (2023, a translation layer that handed developers a homework assignment: port your Windows game to our 2% user base at your own expense), and the Apple Games app plus Metal 4 (2025, a social launcher and graphics API update with no studio acquisitions, no gaming dock, and no first-party AAA titles behind them). Each chapter follows the same pattern: reactive, incremental, and carefully positioned to avoid genuine commitment. That is not a strategy. That is the absence of one.
The hardware is already there
This is what makes the miss inexplicable. Apple Silicon is competitive on raw GPU performance and dominant on perf-per-watt. Thunderbolt 5, which Apple ships on its M4 Pro and Max machines, runs at eighty gigabits per second bidirectional — enough bandwidth to run an external GPU at near-internal performance. The components for a category-defining product are already in Apple’s hands.
The product writes itself: a Mac mini-form-factor external gaming box that connects to any MacBook via a single cable. Mobile when you need mobile; drop into the dock at home and the laptop becomes a gaming workstation. Microsoft can’t build this because they don’t own the laptop. Nvidia can’t build it because they don’t own the OS. Only Apple can do the full vertical.
Instead, when Apple transitioned to Apple Silicon in 2020, they killed eGPU support entirely. No driver path for external GPUs. The stated reason was that Apple Silicon was “fast enough” — true for video editing and Xcode, emphatically not true for AAA gaming, where a single Nvidia RTX 4090 still embarrasses an M3 Ultra. Steam surveys put macOS at under 2% of its user base. That is the hardware gap in one number. And the retreat wasn’t limited to hardware: for years Apple’s App Store rules blocked native cloud gaming apps like Xbox Cloud Gaming and GeForce Now — the one path that would have brought the AAA library to Apple’s platforms at zero engineering cost to Apple. That was not a passive miss. That was an active retreat.
The economics of a self-imposed cap
Get the units-versus-dollars distinction right, because it is the whole story. The traditional PC hardware market — desktops, notebooks, and workstations, the IDC definition — runs about 260 million units and roughly $170 billion in revenue a year. Apple ships about 9.5% of the units. But because Macs sell at roughly twice the industry’s blended average selling price — about $1,370 per unit against an industry blend near $650 — that 9.5% of units translates into roughly 19% of the dollars, on Mac revenue of about $34 billion. Apple is fourth in units behind Lenovo, HP, and Dell, and arguably second in dollars. Mac grew 12% last year on the strength of the M-series — which only sharpens the point: the hardware is winning on its merits everywhere except the one segment Apple refuses to contest.
The gaming PC hardware segment — the machines, GPUs, and peripherals, separate from game software — was roughly $65 billion in 2025, growing at 13% annually, several times the broader PC market’s rate. The fastest-growing slice of the PC market is nearly twice the size of Apple’s entire Mac business, and Apple doesn’t compete in it. Yes, Apple collects a 15-30% toll on the world’s largest mobile gaming platform through the App Store — and a handful of prestige iPhone ports like Resident Evil Village and Assassin’s Creed Mirage notwithstanding, that’s not a gaming strategy. That’s a tax. Apple’s other answer, presumably, is that spatial computing via Vision Pro is the next gaming platform — a bet that requires a $3,500 headset to succeed where every previous VR device has failed, backed by a game library that doesn’t exist.
But the gaming PC dollar figure is only the visible part. The larger number is every productivity laptop, every kid’s first computer, every household desktop that gets bought as Windows because the gamer in the family decided. PC purchases lock in four to six years of platform commitment. And the mechanism is measured, not speculative: Pew finds 97% of American teenage boys play video games, and roughly half of teen gamers play on a PC. Explicitly badged gaming laptops alone run about a fifth of laptop shipments and roughly $31 billion in annual revenue — on the order of a quarter of worldwide laptop dollars, before counting the general-purpose PCs specced with gaming in mind. A single fourteen-year-old who wants to play with friends doesn’t just buy one Windows PC. He sets the household platform, which determines what mom replaces her laptop with, what dad gets for his home office, what the younger sibling unboxes at twelve. One gamer locks in three to five PC purchases over a decade, and Apple loses every one.
Now run the counterfactual, and be explicit about the frame. Suppose Apple held 25% of PC units instead of 9.5% — competitive with Lenovo, plausible if they actually tried. At Mac selling prices, 25% of units means roughly 40% of industry dollars: about $89 billion in annual Mac revenue against today’s $34 billion, or roughly $55 billion incremental — call it $40-45 billion if the expanded mix dilutes ASPs toward $1,100 — plus $15-20 billion in Services attach on the enlarged installed base. At Apple’s blended margins that stacks to something like $25-30 billion in incremental operating income, worth on the order of three-quarters of a trillion dollars at Apple’s current earnings multiple — before counting a single iPhone or wearable pulled along by each new Apple household. Count those, and a trillion dollars is the honest round number sitting on the table. And if you reject the upper bound entirely: even if serious gaming merely doubled Apple’s unit share into the high teens, the incremental revenue would still dwarf the AI narrative. The AI gap that analysts obsess about represents maybe $50-100 billion in five-year risk. The gaming gap is several times larger, and nobody is asking about it.
A moral judgment dressed as product strategy
If the financials are obvious and the hardware is sitting there, why hasn’t this changed in twenty-five years? The answer isn’t strategic. It’s cultural.
Jobs grew up in the Whole Earth Catalog / Bay Area counterculture, where “real” creative pursuits were music, film, graphic design, and serious literature. The original Mac was sold on Garamond and orchestral MIDI. Final Cut, Logic, GarageBand — the whole iLife suite — was built around his personal taste. Games were absent from that vision. When Microsoft bought Bungie in 2000, he didn’t think it mattered.
Cook is even further from gaming culture. Nothing in his public persona suggests he has ever engaged with games as a medium. The senior bench — Federighi, Joswiak, Williams, Ternus — same generation, same blind spot. Contrast Microsoft, where Phil Spencer reports near the top and is a visible, credible gamer. Or Sony, where PlayStation has institutional power. Or Nvidia, where Jensen Huang talks openly about games as the workload that built the company.
The clarifying test is the pitch meeting where someone proposes spending forty billion dollars on a top-tier game studio portfolio. At Microsoft that pitch wins. At Apple: “Is this really what we want to be known for?” That sentence is the whole problem. It treats the question as brand dignity rather than market opportunity. The charitable reading is capital-allocation discipline — plenty of non-gaming companies have bought studios and written them down. But discipline would look like a targeted alternative: selective first-party titles, funded ports, a hardware path for gamers. Apple’s alternative was nothing, for twenty-five years, which is how you know the objection was never financial. What Apple’s leadership believes, in the marrow, is that playing games is not a worthy adult activity. You can hear it in every keynote where games appear only as benchmarks — proof the silicon is fast — before the presentation pivots back to the “real” use cases.
This hierarchy is a generation out of date. Elden Ring, Baldur’s Gate 3, The Last of Us are works of comparable craft to the prestige television Apple is spending billions chasing on Apple TV+. The global games market hit roughly $197 billion in 2025 — larger than film and recorded music combined. None of this has penetrated Apple’s executive worldview.
The real cost: an undefended iPhone
Here is the part that should keep Cook up at night — and it requires being precise about what actually holds iPhone share in place.
It is not the Mac. American teenagers choose iPhone at rates near 87% while Mac sits in a small minority of their households; plenty of those teens run a Windows gaming rig and an iPhone side by side. What holds them is iMessage — the blue bubble, the group chat, the social network effect that lives entirely on the phone. Apple’s US iPhone dominance rests, for the majority of its customers, on a single social-network moat rather than on ecosystem depth.
Now look at what the Mac changes. In an Apple household, the iPhone is defended twice over: iMessage plus Handoff, AirDrop, Universal Clipboard, iPhone Mirroring, shared iCloud — integration that makes switching to Android genuinely painful. In a Windows household, the iPhone is defended once. Same phone, radically different switching cost. Mac → iPhone is nearly unswitchable. Windows → iPhone is one good reason away from Samsung.
That single remaining defense is under active siege. The Department of Justice’s 2024 antitrust suit made iMessage a centerpiece — the Attorney General called out the green bubbles by name — and the case survived Apple’s motion to dismiss in 2025 with every claim intact. Apple has already been pushed into adopting RCS, which closed most of the functional gap with Android messaging. The EU, tellingly, declined to regulate iMessage at all: outside America, the moat barely exists, which is one reason iPhone share runs from a third in Western Europe down to single digits in India. The blue bubble is a US-only network effect, and network effects — unlike integration lock-in — can collapse quickly when the friction sustaining them is removed, whether by court order or by a sufficiently good Android value play.
This is where the gaming miss compounds into franchise risk. The teenager whose games don’t run on a Mac defaults into a Windows household at twenty-two — an iPhone customer held by one thread instead of five. His kids grow up on Windows, and their phone choice is a coin flip. None of these transitions is inevitable; the point is that each one compounds the next. Apple’s gaming abandonment is not creating Android users today. It is manufacturing, at generational scale, the precise category of customer — iPhone-by-network-effect, ecosystem-free — that a legal remedy or a competitor can actually take. The market is pricing iPhone dominance as permanent. It is a single moat, currently in litigation, defending a customer base Apple declined to lock in when it had the chance.
The AI miss is a taste-of-the-week story. The gaming miss is a structural threat to the most valuable franchise in consumer technology, on a generational timeline, hiding in plain sight.
These are the questions Apple’s analysts should be asking on every earnings call. They won’t, because they share the blind spot. They too think video games are something kids do before they discover the things that really matter.