Summary

Apple delivered a substantial iPhone cycle, but the flat response is really a verdict on its absence from AI, not its hardware. Ben Thompson still considers Apple the world’s best mobile-chip maker, and the iPhone 17 generation brings new form factors, AI-oriented GPU components, better cameras, more memory and battery gains. “There was a lot here. It just wasn’t what people wanted it to be.”
The iPhone Air may sell poorly yet still function as a valuable engineering platform. Thompson predicts that buyers prioritizing cameras or battery life will choose the Pro, but concentrating compute inside the camera plateau teaches Apple how to miniaturize electronics and manage heat in confined spaces—work that might eventually benefit glasses. “You make it thin because you can.”
The Air gives Apple permission to “make the Pro pro,” sharpening the product segmentation. The aluminum Pro is a thicker, tool-like device with 12 GB of RAM versus 8 GB, a larger battery and three full-resolution cameras; customers complaining about weight can “go get an Air.” Thompson’s preferred image is not jewelry but Milwaukee Tools: “Give me my big, chonky, thick, actually-runs-all-day” phone.
Apple’s retreat from undeliverable AI promises is healthy, but diminished excitement is “the price of retreating.” There was no timeline update for the new Siri, leaving the impression that Apple remains “stuck in 2015” while the industry’s most consequential technology advances elsewhere. The company may now underpromise and overdeliver, but it has forfeited the awe its events once generated.
Thompson ties Apple’s AI weakness directly to accepting roughly $20 billion annually from Google instead of competing in search. He says the figure may be higher as search volume changes; against about $100 billion in annual profit, losing the payment would hurt, but Apple would retain roughly $75 billion and ample capacity to invest. His harder conclusion: Google “bought them off,” and Apple stopped exercising the data, indexing and information-management capabilities that became essential to AI.
The investor thesis is a highly profitable technology company that still makes impressive hardware but has ceded leadership in the most interesting technology spaces. Apple’s products can be excellent while the company no longer appears to be changing the world.

Deep dive

1. Apple remains a technology company, but its ambition has narrowed

Thompson rejects the claim that Apple has become merely fashion or consumer packaged goods: it still makes the best mobile chips, impressive phones and, as Ben says John was raving, dramatically improved AirPods. The distinction is that Apple now concentrates on technologies “that most people don’t really care that much about anymore.”
What has disappeared is the sense of a company “going for it” and trying to change the world. Thompson can reluctantly admire Apple’s cash flow and low CapEx relative to peers, but fiscal discipline is not something enthusiasts should “sit here like seals and clap” for.
Sharp recalls when iPhone events captured people’s imagination. The new products can be genuinely good while Apple remains “nonexistent in the most interesting technology space”—the tension that explains both Thompson’s hardware enthusiasm and his strategic disappointment.

2. The iPhone Air is an engineering bet without an obvious mass buyer

Thompson predicts Apple will not sell many Airs: anyone who values the camera or battery will logically buy a Pro. The high-end processor makes it more attractive than the old Plus, but thinness is its clearest consumer proposition.
Its deeper value is experimental. “You make it thin because you can”: miniaturizing components and learning to confine electronics in small places creates knowledge whose eventual uses cannot be anticipated, potentially including future glasses.
The camera plateau resembles Apple’s Dynamic Island strategy—turning an unavoidable intrusion into useful architecture. Instead of tolerating a growing camera “wart,” Apple puts the compute inside the extrusion and leverages the space it has created.
Thompson’s critique is that Apple compromised its own pitch by retaining another camera bump atop the plateau. If thinness is the product, Apple should accept a degraded camera that fits flush: camera-first customers were buying the Pro anyway.

3. Aluminum and battery capacity make the iPhone 17 Pro a better tool

Separating the Air allows Apple to “make the Pro pro,” much as it once restored thickness and ports to MacBook Pros. Thompson wants a tool with all-day battery and “47 gazillion cameras,” not a fashion accessory; weight-sensitive buyers now have the Air.
Apple replaced titanium with aluminum, increased RAM from 8 GB to 12 GB and added battery capacity. The Pro is actually heavier despite lighter aluminum because battery is the heavy component—a trade Thompson readily accepts.
Jonathan’s unsent thesis captured titanium’s failures: unattractive colors that do not hold color well, excessive weight and poor heat dispersion. Sharp says his titanium phone becomes alarmingly hot and its battery “sucks”; the aluminum iPhone 17 Pro convinced him to buy rather than defect to Android.
Thompson goes further: his ideal tool would be plastic because it would not attenuate radio signals, improving antenna performance. Still, aluminum is “a gazillion times better” than steel or titanium, while the smaller glass section helps with signal requirements.

4. A meaningful redesign could not overcome Apple’s missing AI future

The event updated the base iPhone and the Pro while introducing the Air. This was a two-year redesign inside Apple’s usual three-year cycle, bringing a new form factor, three full-resolution cameras and a square-oriented selfie camera that preserves quality regardless of phone orientation.
The chip’s GPU now includes components designed for LLMs, unlike Apple’s neural processor for narrower machine-learning tasks such as photo identification. Thompson reads this as a low-key admission that Apple is now designing for LLMs, while considering 2025 a reasonable hardware timeline.
Yet fans immediately called the event boring and said they would wait, although next year may bring less industrial change. Thompson interprets that reaction as displaced disillusionment: there was no updated timeline for Siri, and Apple still feels “stuck in 2015.”
The keynote itself reinforced the problem. Sharp found Apple’s presentation stilted and antiseptic; Thompson preferred Google’s live presentation, including Jimmy Fallon and Giannis, because live delivery creates the possibility of mistakes and gives people something worth observing. Apple’s was a 70-minute recorded commercial that still felt long.
After promising capabilities it could not deliver, Apple needed to reset and focus on what it could execute. Underpromising may repair the company, but excitement falls with expectations: the audience’s angst and flatness are “the price of resetting” and “the price of retreating.”

5. Google’s search payments traded competitive muscle for profit

Thompson places the Google remedies decision beside Apple’s AI malaise. He puts the payment at conservatively $20 billion annually and says it may be higher as search volume changes; he also mentions a supposed decline over the last 12 months. Against roughly $100 billion in profit, removing it would hurt, but Apple would still have around $75 billion and money to spend on innovation.
The judge worried that Apple might innovate less, but Thompson argues that reduced innovation is what the agreement already produced. Apple was the only entity capable of confronting Google at Google’s scale, and Google “bought them off.”
Opting out of search meant avoiding the collection, cataloguing, sorting and management of enormous data sets—the same “muscle groups” needed for AI. Apple then entered the gym after ten years away, attempted its old weight and “screwed up their back” by promising Apple Intelligence capabilities it could not deliver.

6. Apple’s middle age is the disappointment

Thompson softens the “sugar water” analogy: Apple remains a real technology company, and its products keep improving. The more important half of Jobs’s challenge is “change the world,” and Thompson says the answer to whether Apple currently feels like it is doing that is clearly no.
The core complaint is therefore not that the iPhone 17 hardware is empty. Thompson says there was “a lot here”; the disappointment comes from the gap between excellent execution in mature products and Apple’s absence from the technology space that now commands attention.

Are We Past Peak iPhone? + Eliezer Yudkowsky on A.I. Doom OpenAI’s $10BN Secondary Sale, Ramp Hits $1BN ARR & Brex Hits $700M