As stupid as Apple’s Vision Pro is, Apple Intelligence is that … intelligent.
Innovation Is Overrated
Specifically, disruptive innovation — the kind that marks a “before” and “after” in our lives — is terrible for shareholder value. Some innovators that changed our lives:
- Seattle Computer Products
- Xerox PARC
- Grid
- Palm
- Netscape
- Friendster
- Blackberry
- Alta Vista
- Nokia
Combined market capitalization: $21 billion. And that’s only because Blackberry and Nokia somehow still exist. (Xerox donated PARC to a nonprofit in 2023, valuing it at $132 million.) Just four of the companies that capitalized on the innovations of the departed — Microsoft, Apple, Google, and Meta — register a market capitalization of $7+ trillion. The second mouse often gets the cheese. Tesla didn’t invent the consumer electric car (GM), Visa didn’t invent the credit card (Diner’s Club), McDonald’s didn’t invent fast food hamburgers (White Castle), and Coca-Cola didn’t invent soda (Dr. Pepper). One firm enjoys a market cap larger than all five combined — Apple is a second mouse the size of a blue whale.
I-brand
“AI,” has seen a brand equity implosion in 2024 similar to that suffered by the Supreme Court, Twitter, and elite colleges. The promise of the AI brand is that it will either save us/kill us all. It has done neither. Any time saved by AI has mostly been chewed up listening to breathless media reports on management (or lack thereof) drama at Microsoft AI (some people call it OpenAI). The brand is the offspring of capitalism and the Bravo Channel. Staggering increases in shareholder value mixed with IP theft, hallucinations, and constant catastrophizing. It’s as if Rupert Murdoch got married, for a sixth time, to SkyNet. I’m especially proud of the previous sentence. Apple’s move to shitcan the AI brand, and opt for Apple Intelligence, is the best brand move of 2024. The worst? A: A toss up between a chardonnay-fueled decision to hang the Stars and Stripes upside down or the zombie apocalypse of useful idiots on US campuses.
Apple Intelligence
“Apple Intelligence” is more than a great brand move; it encapsulates the company’s strategy. Take something invented elsewhere; make it more consumer friendly, easier to use, and more reliable; mix in world-class industrial design; and print billions. Artificial intelligence is for tech bros and data scientists. Apple Intelligence is AI for the rest of us. Shrewd.
The tech press has spent the past 18 months telling us Apple is behind on AI. While in the next breath reporting on the AI gaffes produced by its rivals. And that’s the point. Apple is always behind. Apple is a distinctly inventive company — its $30 billion R&D budget generates 2,000+ patents per year. But it’s mainly improving vs. inventing: ways to more precisely cut white cardboard boxes to deliver its new devices; new glues to bond layers of glass and plastic together in its phones. For the big stuff, like the mouse, digital music, and multitouch screens, it lets someone else traverse the Sierra Nevadas first.
What has Apple really been “behind” on? The Economist estimates that AI will generate $20 billion in revenue for AI “leaders” Alphabet, Amazon, and Microsoft in 2024, a whopping 2% of their combined revenue. Apple makes close to $20 billion just from Airpods. Think about that — all that hype and increased shareholder value, rivaling the GDP of Germany, and the AI industry is (so far) the same size as Airpods. The media tells us Apple isn’t an attractive employer for cutting-edge AI researchers because of its secrecy. But the company is content to let artificial intelligence scientists work elsewhere and publish their work and exchange ideas freely.
Siri, but Better
When Apple does try to be cutting-edge, it confirms the merits of its slower approach. When it launched Siri in 2011 (using technology it acquired), Samuel L. Jackson introduced the digital assistant science fiction had been promising us. Even today, Siri doesn’t work as well as promised in 2011. I just asked her the same question Jackson asked in 2011 — “Find me a store that sells organic mushrooms for my risotto” — and she gave me a list of UPS stores. Siri is Apple’s most glaring failure, even more so than the Vision Pro — the headset is a sideshow, an insurance policy Tim Cook purchased in case that mendacious fuck (one guess) was right about headsets. Apple spent couch change just in case Zuck knew hardware. He didn’t. Siri, however, is supposed to be at the heart of Apple’s most important products. Apple has the brand equity and capital to absorb mistakes like these, but the missteps show that being first is not the winning strategy.
Alexa and Google Assistant are better than Siri, but they aren’t better enough. They also came out too early, got bogged down in the Sierras, and were forced to eat each other. Too much? If Apple Intelligence delivers on the promises made this week, this version of Siri will be the product Apple should have waited to launch. Apple has learned to underpromise, in contrast to the rest of the AI committee, who tell us, in hushed tones, how concerned they are about AI. That’s an obnoxious humblebrag. The anti-Apple crowd growled that Android already has many of Apple’s new features. It does, sort of, if you have the right phone and can navigate a complex ecosystem. It has “artificial intelligence.”
Apple Intelligence’s integration of ChatGPT also elegantly eliminates an existential threat. Few firms have the consumer resonance and cheap capital to launch a handheld phone. It’s been reported that OpenAI is planning to develop an AI phone — with former Apple designer Jony Ive. But why would you want a ChatGPT phone when you can have ChatGPT on an iPhone?
Cabbage
On the day of Apple’s Worldwide Developer Conference, when Apple Intelligence was introduced, the stock declined 2%. However, as with an intelligent joke, it took a beat to absorb the real insight.
From a shareholder standpoint, Apple Intelligence has positioned the company to (again) create the (second)-most-profitable licensing deal in history. Much of Apple’s shareholder value hasn’t come despite competition from Microsoft or Google, but because of it. Specifically, Apple’s ability, as gatekeeper to the Earth’s billion most affluent inhabitants, to pit the two against each other. Alphabet pays Apple $20 billion each year to make Google the default search and gain VIP access to the premier club on the planet. It’s likely $19.8 billion of that hits the bottom line. At a P/E of 32, the Alphabet deal is responsible for 20% of Apple’s market cap. Incorporating OpenAI’s ChatGPT positions the business to put ball gags on both MSFT/OpenAI and Alphabet and molest them, as they’ll be forced to pay billions for direct access to Apple’s users. At launch, Apple is reportedly letting ChatGPT in for no fee, but don’t expect that to last, as Apple is likely erecting a toll booth — and it will surely take its usual cut from premium chatbot subscriptions bought via iPhone. The market figured this out, and over the next two trading days Apple added $300 billion in market cap. Apple Intelligence, by any other name, is this decade’s Apple monopoly tax.
Contextual Intelligence
“Generative AI” has been the anodyne buzzword in tech since ChatGPT launched, but there’s something more powerful: Contextual AI. I don’t need AI to know everything about Icelandic history or sinus medication. I need it to know about me. The most impressive AI feature on my iPhone is Memories. I don’t know how or when it was added, but it’s powerful. Out of nowhere I get notifications — “You have a new memory” — and there is a cropped image of my boys on their first day of school, the younger one clinging to his mother, that flows into another image of the older one not comforting him. All set to music that morphs my older son’s disposition into endearing from … indifferent. But I digress. I use AI for brainstorming and research; thus far that hasn’t affected my life nearly as much as the ability to more easily search my photos — a feature Apple Intelligence will enhance. A well-timed Memory renders me a chocolate mess, in a good way. No number of parameters in GPT-5 will generate that.
Integration into the Apple ecosystem was the theme of the Apple Intelligence announcement, giving the AI the context of our email and messages, calendar, browser history — the whole storehouse of information already on our device. Post-launch, Apple Intelligence will start to reach out to third-party apps and services beyond our devices. That’s where the real cheese lies — “Scott, how’s your shoulder pain? Do you want me to make an appointment with your physical therapist?”
Design matters, and Apple’s AI features will reflect that. But Apple’s greater advantage at this stage of its evolution isn’t design, or technology, or distribution, though all are best in class. Its advantage is that for the wealthiest billion people on Earth, an iPhone is the first device they see in the morning and the last before they go to sleep, and it’s never more than a few feet from them throughout the day. Everything I do is on my phone. The LLM that gets my discretionary spending will be the LLM that gets me. And that means it has to live on my iPhone.
Less Downside
Another advantage of the second mouse strategy is that when you fail, it’s a whole lot cheaper. The exit wounds are clean and heal quickly. Meta has burned $46 billion to stuff the same drawer that has your Nike Fuelband with Zuckerberg’s VR hallucination. Apple’s cheaper call option on the metaverse can be quietly killed in a few years with no lasting damage. In innovation-driven industries, how you fail is almost as important as how you succeed. I predicted the Vision Pro would be a failure when it launched, but I didn’t sell Apple stock.
2025
Next year will probably be the year that real winners and losers start to emerge in AI. We’re still in the Netscape stage, when the technology itself is the innovation. Because our government spent the last 40 years asleep at the switch on antitrust, the usual Big Tech giants will fight for AI supremacy, and Apple is holding a strong hand. Not just because of its second mouse strategy, but also thanks to its vertical integration. There is a lot of money to be made adopting an asset-light model — see Airbnb, Shein, Nvidia, more on that in another post — but Apple’s contrary approach has its advantages.
Apple Intelligence requires massive computational power to run LLMs on the device, but that’s key to its contextual awareness, speed, and reliability. When these features roll out later this year, they will only work to their fullest on the latest Apple devices — giving a billion or so users of older editions a reason to upgrade. Then they’ll work on every future iPhone — and iPad and Mac. This sharpens an edge over Alphabet, which can’t ensure every Android phone has the necessary hardware to run Gemini Nano, the Android LLM equivalent to Apple Intelligence, or even access to the updated OS. Android’s Achilles heel has long been this fragmentation — what you see running on Pixels in the keynote takes years to filter down to the phones most Android users actually own.
Future Memories
I spend a great deal of my wealth on homes in nice places. The goal is to live where my sons, and the people they collect, will come visit me. I think a lot about death. It gives me power/courage to live a bit louder and less fearfully. And when the ass cancer comes, I plan to be in a beautiful place, surrounded by people who will miss me terribly, a shit-ton of heroin, and Tom Petty, who’ll be joined by his best friends from the eighties. In addition, I plan to live my life over again, courtesy of Apple Memories. And that’s the real promise of technology: not to explore new worlds in a dildo or reduce customer service costs; but to save people time so they can spend more moments with loved ones … and feel closer to them. Tech’s promise isn’t artificial intelligence but native intimacy.
Life is so rich,
P.S. On Prof G Markets this week, Ed and I spoke with Morgan Housel, author of The Psychology of Money, about how FOMO, doom, and ego affect your investment behavior. Listen and subscribe here.
