Has Apple Lost Its Way? A Product Person’s Manifesto for Getting It Back
Every few months the same question bubbles up on the internet’s tech watercoolers: Has Apple lost its way? On Hacker News, one commenter sighed, “I think the ‘has apple lost its way’ is a meme now.” Another, less charitable, bluntly declared: “Apple’s software isn’t anywhere near optimal.” On MacRumors, long‑time posters trade laments about the fading “Apple charm,” while on Reddit you’ll find the recurring elegy: “Apple under Cook… is cooked.” (Hacker News, MacRumors Forums, Reddit)
The vibes are the vibes—but what do the numbers say? Apple’s business remains a juggernaut: roughly 51% of 2024 revenue came from iPhone (about $201B of ~$391B), with Services now a quarter of the pie and growing. The installed base of iPhones sits in the 1.3–1.5 billion‑user range, depending on who’s counting. This is not a flailing company. It’s a successful one facing strategic drift: uneven AI execution, an ambitious headset that hasn’t cracked mass‑market product‑market fit, and a maturing iPhone line whose annual treadmill increasingly feels… arbitrary. (Business of Apps, Visual Capitalist, DemandSage)
Below is the opinionated plan I’d run if I were in charge in Cupertino. It’s grounded in data, customer sentiment, and the stubborn physics of product adoption.
1) Put Vision Pro on the back burner—until a real UX breakthrough
Vision Pro is an astonishing demo and a niche tool. Enterprise traction exists—training, design visualization, field workflows—but consumer adoption is modest at best. Independent tallies put 2024 units in the ~370k–500k range; even bullish reads concede content and comfort gaps. Early return chatter hit the press in February 2024 (anecdata, but telling), while later analysis suggested return rates may be ~1%—comparable to iPhone—meaning buyers who stick with it tend to keep it, but the top of the funnel is small and content‑limited. In 2025, coverage still describes a device carving out niche business use rather than mainstream love. (productparty.us, AppleInsider, The Verge, The Wall Street Journal)
What to do: move Vision Pro to patient mode. Keep investing in core tech (lighter optics, thermal, battery, comfort, controllers/gestures) and enterprise SDKs, but freeze the consumer marketing drumbeat until there’s a breakthroughUX moment: a comfort delta, a must‑have app, a price collapse, or truly magical passthrough. As one Redditor griped about VP’s purpose—“over‑engineered… not made for gaming… few useful apps”—the burden is on Apple to prove why normal people should care. Don’t brute‑force that answer with ads. Earn it with the product. (Reddit)
HN, ≤25 words: “Very importantly [a PM] should be someone prepared to take decisions… that need making.” Sometimes the right decision is to wait. (californiaconsultants.org)
2) Revive ‘Project Titan’—but only as a partnership with Toyota
Apple canceled its decade‑long car project in February 2024, reportedly after spending over $10B; many staff rolled over to AI. It died for familiar reasons: autonomy is hard, EV economics are tight, and Apple never locked the right manufacturing partner. Toyota’s chairman warned in 2021 that if Apple builds a car, it must be in for “40 years of service and support.” He was right. (Reuters, MacRumors, The Guardian, AppleInsider)
What I’d revive is not an “Apple Car,” but a profit‑sharing platform with Toyota: Apple software + silicon (drive compute, UX, safety OS ideas that survived) coupled with Toyota’s manufacturing, dealer, and service network. Why Toyota? There’s history (reports of talks in 2021), and unlike some premium brands now pushing back against CarPlay Ultra to protect first‑party in‑car UX revenue, Toyota has generally followed demand on CarPlay because customers want it. Structure a revenue share on software and services that makes Toyota more profitable with Apple onboard than without it. Build mutual incentives and keep Apple away from sheet metal. (MacRumors, Carscoops, MotorTrend)
Will this be easy? No. But post‑Titan, Apple holds priceless IP (sensors, autonomy stacks, “safetyOS” concepts), and Toyota brings the scale and discipline Apple lacked. This is CarPlay‑as‑strategy, not CarPlay‑as‑skin.
3) Kill the annual iPhone treadmill and ship “iPhone Infinity”
“Another September, another iPhone.” The cadence has been brilliant for revenue—and corrosive for meaning. Replacement cycles lengthened to 3–4 years through 2023; a CIRP blip in late 2024 showed some shortening (36% of buyers had phones ≤2 years old), likely due to AI and USB‑C, but the long‑term arc still points to longer ownership. Europe is cementing that reality: from June 20, 2025, phones sold in the EU must meet new durability and repairability standards, with 5 years of OS/security updates and 7 years of spare parts availability. The world is rewarding longevity. So should Apple. (SellCell, MacRumors, Internal Market and SMEs, CLASP)
Plan: end the performative “every year is revolutionary” ritual. Announce iPhone Infinity—a continuously improved flagship line with quiet mid‑cycle hardware bumps and feature drops, plus clear Long‑Term Support channels. Sell by capabilities and support horizon (e.g., “Infinity, 7 years of OS + 10 years of service”). Tie upgrades to meaningful deltas (battery, cameras, modem, on‑device NPU) rather than calendar. Apple already explored a subscription path, then reportedly killed the ‘hardware subscription’ idea in 2024—good; instead of “phones as a service,” make support the service. (Bloomberg, MacRumors)
MacRumors, ≤25 words: “iPhone upgrade cycles are getting shorter again,” but only slightly, and after years of lengthening. Don’t fight physics—design for it. (MacRumors)
4) Do the same for software—introduce “iOS Infinity”
Apple already ship‑streams iOS with rapid point releases and “Rapid Security Response.” Make it explicit: iOS Infinity, a rolling OS with annual LTS branches. Align this with EU durability rules, publicly commit to 7 years of security updates on new flagships (Apple already effectively delivers 5–7), and lean into in‑place modular upgrades (visionOS‑style app frameworks, private APIs that decouple core services from monolith releases). This halves platform fragmentation and doubles customer trust. (compareandrecycle.co.uk)
5) Rebuild Siri as the leading AI assistant—and a developer platform
WWDC24 introduced Apple Intelligence, an on‑device first AI stack with Private Cloud Compute (PCC) for heavy lifts; Siri gained typing, on‑screen awareness, and device help. It’s the right privacy architecture. But users are impatient: timelines slipped, and press rounds report meaningful Siri upgrades sliding into 2025 and even 2026 windows. On Reddit, the sentiment is blunt: “They neglected Siri until she’s borderline worthless.” Apple must not just catch up—it must leap. (Apple, Apple Security Research, MacRumors, Reddit)
Direction:
Siri 2.0 = Planner + Tools + Memory. Keep Apple’s “Plan‑Act‑Summarize” architecture, but open a capability registry for third‑party tools (with fine‑grained, user‑scoped permissions). Think “Shortcuts on steroids,” model‑agnostic.
On‑device first, cloud when needed. PCC is Apple’s unfair advantage; keep the strongest privacy story in AI. Publish verifiable security docs (Apple already has) and expand community review bounties. (Apple Security Research)
SiriKit++ for devs. A modern, LLM‑aware SiriKit: structured intents and tool specs that let any app expose safe, reversible actions. Give devs a debugger for “why Siri failed” traces.
Choice of models behind the scenes. Where web/world knowledge or planning exceeds Apple’s models, be pragmatic—broker with OpenAI/Google under privacy contracts, but front it with Siri and attribute clearly. (Reports already suggest Apple is exploring Google’s Gemini for certain Siri tasks.) (The Verge)
Ruthless UX focus. Ship a dozen magical end‑to‑end flows and nail them: “Book the flight in my email into my calendar; rearrange meetings; submit my PTO; message the team; file the expense.” Fewer demos, more done.
HN, ≤25 words: “The best PMs are good mediators… get buy‑in from sales and development with minimum squabbling.” Replace squabbling with shipping. (CARLUEX)
6) Focus on real moonshots (with patient capital)
Apple’s next S‑curve likely isn’t a headset you wear all day. Pick 2–3 decade bets:
Ambient health: from watch‑first to Health OS—sleep apnea, hypertension, low‑glucose trends, mental‑health signals. Work backwards from clinical endpoints, partner with payers/providers, and own population‑scale prevention.
AI + privacy infrastructure: scale PCC into a trusted compute utility for developers—privacy attestation, external enclaves, zero‑knowledge receipts. If Apple can make privacy work at scale for AI, that’s a moat. (Apple Security Research)
Home energy + devices: make HomeKit not a hobby. Bundle local inference for energy optimization, grid interaction, and real security (not cloud‑latency “security”).
Human‑scale robotics & accessibility: leverage Apple Silicon NPUs, motor control, and perception for assistive devices—the robot that helps more people age in place is larger than a niche VR play.
Moonshots don’t have to shout; they have to compound.
But… has Apple actually lost its way?
The fairest answer is: Apple is still Apple—just not the Apple some of us remember. As one MacRumors poster wrote years ago, maybe nostalgia is doing as much work as reality. Yet the frustration is real: a headset without a mass‑market “why,” a car that died after billions, a voice assistant that lags rivals, and an iPhone calendar show that sometimes feels like a ritual. The cure isn’t a thousand features—it's a handful of right choices. (MacRumors Forums)
The scoreboard Apple should watch
Vision Pro: Pivot to niches until the UX breakthrough lands. Adoption likely sub‑million; enterprise value real; consumer value TBD. (The Wall Street Journal)
Titan: Canceled after ~10 years and >$10B; future should be Toyota‑style partnerships, not solo manufacturing heroics. (Reuters, MacRumors)
iPhone cadence: Replacement cycles lengthened; EU now mandates longevity—design for it with Infinityhardware + software. (SellCell, Internal Market and SMEs)
Siri/AI: Great architecture (on‑device + PCC), sluggish delivery. Open the platform, ship end‑to‑end flows, and be pragmatic on model sourcing. (Apple, Apple Security Research)
Revenue reality: iPhone is still ~51% of Apple revenue; Services ~25%. Use that cashflow to buy patience for moonshots. (Visual Capitalist)
Reddit (longer anecdote): “They grossly underestimated AI… neglected Siri until she’s borderline worthless.” That’s the internet at 11—but the core criticism is useful fuel. (Reddit)
Closing argument: Choose depth over ceremony
Apple hasn’t “lost its way.” It has choices. The choices I’d make are boring on the surface and potent underneath:
Pause Vision Pro’s consumer push until the comfort/content delta screams yes.
Reopen the car—but as software + silicon + revenue‑share with Toyota, not a badge.
End the September treadmill. Ship iPhone Infinity and iOS Infinity that honor long ownership and legislated durability.
Rebuild Siri as a tool‑using platform with airtight privacy and real developer leverage.
Bet on moonshots that compound quietly: ambient health, privacy‑first AI infrastructure, the real smart home, and assistive robotics.
If Apple does that, the internet will still argue—HN and Reddit are undefeated. But the product will speak. And in a year or two, those same forums might trade the “lost its way” meme for something rarer online: respect.