One signature erased Apple's entire headset future. John Ternus, who doesn't become CEO until September 1, personally signed off on cutting the whole Vision lineup — the second-generation Vision Pro, the lighter Vision Air, the display glasses — leaving exactly two products on the roadmap, both of them glasses, neither of them a headset.
That is the detail that turns an analyst note into a signal. Ming-Chi Kuo's report reads, as VR.org put it, less like supply chain analysis and more like an obituary. But the byline underneath the decision is what matters: this is the first major portfolio call anyone can pin on the man who will own Apple's product strategy for the next decade.
Last week we covered the handover itself — Cook out on September 1, Ternus in, the design team in crisis. This week the abstraction became concrete. We now have moves, not just a name: a canceled headset line, a glasses-only XR roadmap, and a Siri reportedly running on a competitor's brain.
Start with what got cut, because the logic is cleaner than the drama suggests.
Everything with a head strap is gone. Gurman reported the Vision Air was killed in October 2025 and the display glasses fell off the table in January 2025. Kuo's report removes every Vision Pro successor entirely. A proper follow-up is "on ice," which in Apple-speak is a polite way of saying dormant.
What survived are two bets on your face rather than over it. A display-free AI smart-glasses product — the Ray-Ban Meta category, cameras and speakers, no screen — is still tracking for the end of 2027. A more ambitious AR device using optical waveguides has slipped to 2029.
This is not visionary. Ternus is reading a sales chart. Smart glasses are reportedly outselling VR and MR headsets three to one, and nearly all of the XR category's 44 percent shipment growth in 2025 came from glasses, per VR.org. Apple watched a $3,500 headset miss twice — at launch, then again after the M5 refresh, per VR.org's analysis — while Meta moved Ray-Bans by the millions at a tenth of the price.
So the Vision Pro becomes what VR.org called a bridge product. Bridges are useful. They are also things you cross once.
Here is where the story stops being about a canceled headset and starts being about a coherent company.
Look at the second move: the design studio. Under Cook, it was hollowed out to the point that Gurman describes it as "a service bureau where other teams come to get what they need and get out" — stripped of a seat at the executive table, staffed increasingly with junior hires, drained of nearly every Ive-era designer who defected to LoveFrom, founded startups, or retired.
Ternus told staff the company is "going to keep focusing on design, because design is core to what we do at Apple." He put it more bluntly elsewhere: "The most beautifully designed thing that most customers own is an Apple product. We're going to make sure that stays the case."
Read those two moves together and they are the same bet.
Killing the headsets and rebuilding the design studio are not separate initiatives. They are one thesis: Apple wins on the thing it can still control — hardware you wear and actually want. Not a spec race. Not a headset nobody bought. A desirable object on your face at a few hundred dollars, backed by a design culture capable of making it feel inevitable.
Killing the headsets and rebuilding the design studio are not separate initiatives. They are one thesis: Apple wins on hardware you wear and actually want.
This is what the Cook-era Apple lacked, and it's worth naming the contrast. Per Apple's own statement, Cook's decade was scaling, services, supply chain — a $350 billion company grown to $4 trillion, more than 500 stores, a $100 billion services line. Extraordinary operations. But the design org shrank in exactly that period, and the iPhone looked largely the same for half a decade. The company got bigger and less distinct.
Ternus, a mechanical engineer by training who joined the product design team in 2001, is betting the other way. His first public act as CEO-designate was to subtract — to cut the product that no longer fit. His second is to rebuild the muscle that makes hardware worth wanting.
Coherent. Legible. The most focused Apple has looked in years.
And then, in the same window, the crack.
At Cook's final WWDC as CEO headliner, the standout announcement was not a new object. It was that Apple is reportedly leaning on Google's Gemini to power Siri, per a translated aggregator report. The company that has promised "Apple Intelligence" reportedly outsourced Siri's actual thinking to a rival. Testers reportedly said the upgraded Siri could access contacts' conversation records, and described the experience as smooth.
Sit with the contradiction. A company betting everything on making the most desirable object you own just rented out that object's mind.
This is the one move that betrays the strategy. Design-first works because Apple controls the whole stack — the aluminum, the chip, the feel, the software. That's what makes the hardware defensible. Intelligence is now the most important layer of that stack, and Apple just declared it does not control that layer. It leases it.
The glasses only make this sharper. A 2027 AI glasses product is, functionally, a face-worn interface to an assistant. If that assistant is Gemini, then the most personal Apple hardware yet is beautiful frames wrapped around Google's model. The object is Apple's. The mind is not.
The right way to read this is that intelligence is provisional right now, for everyone. Apple is choosing not to build the brain at all. That tells buyers the intelligence layer is unsettled, and nobody's current answer is permanent.
So what does this mean if you're deciding what to buy or wait for.
First: stop waiting for a Vision Pro 2. There isn't one. If you own a Vision Pro, treat it as the hardware you'll run through the end of the decade — software support looks solid through 2028 or 2029, but the M5 model is the ceiling. For the enterprises running these in operating rooms and training departments, that actually simplifies the math. There is no next one to wait for.
Second: the thing to actually wait for is the glasses. A design-led, headset-light Apple means the 2027 AI glasses and the 2029 waveguide AR device are the real roadmap. If Ternus rebuilds the design studio the way he says he will, those are the products where the "beautifully designed thing you own" thesis gets tested.
Third, and most important: treat the intelligence inside them as rented, not owned. The design you're buying will be Apple's. The thinking, for now, is Gemini's — which means the "Apple" smarts you're paying for are contingent on a competitor's roadmap, pricing, and priorities.
Ternus's Apple is the most coherent Apple in a decade. The subtraction was disciplined; the design bet is real. One signature made the whole strategy legible.
But a design-first company that rents its intelligence has drawn a line down the middle of its own product. Buy the object. Just don't assume the mind inside it is Apple's to keep.