Pixel's Magic Cue Isn't Just a Feature. It's Intelligence as OS.
This sounds crazy but hear me out: why Apple might be forced to run Gemini by 2027.
In January 2007, BlackBerry executives watched Steve Jobs unveil the iPhone.
At that time, BlackBerry had over 30% of the US smartphone market share, and growing Their users, mostly white-collared professionals, loved the tactile keyboard, the robust security features, and the push email service.
They saw the iPhone as a toy, not a threat.
“It's OK. We'll be fine”.
Famous last words.
For a short while after, they even looked like they were right. BlackBerry sales revenue actually tripled to over $11 billion in the next two years.
Five years later, things took a disastrous turn. BlackBerry fell from 56% of the US smartphone market share in 2019 to a disastrous 1% in 2024.
What the heck happened? Weren’t they doing everything right? They did everything their customers asked for. Better keyboards. More secure email. Faster messaging.
But they optimized for the wrong thing.
This week's Pixel release revealed something that made me think of BlackBerry's downfall.
Magic Cue, the new flagship AI feature on Pixel, pulls information from Gmail, Calendar, Messages, and Phone to surface what you need in the moment. Smart replies with restaurant addresses. Flight confirmations during airline calls. Live translation over phone calls.
Here's what everyone missed: Google isn't just integrating their apps better. They're demonstrating their nuclear option. Intelligence as OS.
Most iPhone users use:
Google Maps (70% of people prefer it to Apple Maps)
Gmail has 2.5 billion users globally (many on iPhones)
Google Calendar, not Apple Calendar
Google Docs, not Pages
Apple can build the most sophisticated AI ever created. But what happens when Google decides third-party AI can't access Gmail? Or Google Calendar? Or any Google service?
The Data Layer Is the New Platform
BlackBerry lost because they didn't control the new platform - the app ecosystem. They had the hardware, the enterprise relationships, the security. But none of that mattered when developers chose iOS.
Today's uncomfortable truth: Apple doesn't control the data layer on their own devices. Google's AI can analyze your entire digital life - your emails, documents, calendars, search history, location data, YouTube watching patterns. All on Apple's own hardware.
Apple owns the device but Google owns the data exhaust. And in the AI age, data exhaust is everything.
The Battle for Data is the AI War
The next tech war won't be fought over hardware or software. It'll be fought over API access.
Hypothetically - Google could, tomorrow, announce that only Gemini-powered devices get full access to Google services. Not blocking them entirely, that would trigger antitrust. Just degrading them. No smart replies in Gmail. No AI features in Docs. No intelligent suggestions in Maps.
What's Apple going to do? Build competing services? They've tried. Remove Google apps from the App Store? That hurts iPhone users more than Google.
The Nuclear Scenario
Here's the strange prediction: Apple will be forced to run Gemini on iPhones by 2027.
Not as an option. As the default.
The playbook is simple:
Google requires Gemini integration for full AI features in their services
iPhone users realize their $1,200 phone can't do what a $400 Pixel can with their Gmail
Users revolt, threatening to switch to Android
Apple capitulates, announcing Gemini integration as "giving users choice"
Sound insane? BlackBerry thought so too when they were forced to support iPhone apps. That "strategic partnership" was actually a surrender. They became a software company running on their competitor's platform.
Apple x Gemini collab?
Apple might actually make more money as a premium hardware company running Google's AI than competing in AI. Google pays Apple $20 billion a year just for search default. Imagine what they'd pay for intelligence default.
Apple might have to become a hardware manufacturer for Gemini. And they might stay one of the world's most valuable companies doing it.
Agree? Disagree? Tell me why I’m wrong in the comments below.
The Download —
News that mattered this week
OpenAI logged its first $1 billion-month but computing power demand is ‘voracious,’ CFO says: She adds that insufficient compute to meet the demand of AI is the company’s biggest challenge.
Google reveals how much energy a Gemini query uses: The average Gemini text prompt uses "0.24 watt-hours (Wh) of energy, emits 0.03 grams of carbon dioxide equivalent (gCO2e), and consumes 0.26 milliliters (or about five drops) of water," Google said, comparing the per-prompt impact to "watching TV for less than nine seconds."
Lovable projects $1B in ARR within next 12 months: Vibe coding startup Lovable aims to hit $1 billion in annual recurring revenue within the next 12 months, according to its CEO, Anton Osika.
AI is failing at an overwhelming majority of companies using it, MIT study finds. A staggering 95 percent of attempts to incorporate generative AI into business so far are failing. (MIT)