(Preview) Meta Doubles Down on Smart Glasses, Nvidia Partners with Intel, Notes on the Hidden YouTube Colossus
Summary
- Ben Thompson’s core call is that Meta’s $799 Ray-Ban Display matters less for its screen than for the Meta Neural Band bundled with it. The wristband reads electrical signals, enabling discreet gestures or even writing with a hand inside a pocket—an input breakthrough he sees as essential to augmented reality. Having previously tested it, Thompson said, “I have 100 percent confidence that this is going to work.”
- Shipping on September 30 at $799 moves Meta’s glasses effort from expensive prototype toward a manufacturable consumer platform. Thompson expected a four-figure price and said his “jaw literally dropped” at $799; Meta might be “taking a bath,” but after spending $100 billion on Reality Labs, subsidizing initial adoption could build demand, improve yields, and amortize tooling.
- The product’s severe constraints may be an advantage because Meta is not trying to rebuild the phone on users’ faces. It connects to iOS or Android, offers a small collection of Facebook/Meta-specific experiences, and even limits maps to walking directions, yet Thompson argues the eventual interface should be “AI-dominated”: contextual information appearing when needed instead of users navigating conventional apps.
- Meta can use early customers to discover the category’s beachhead rather than pretend it already knows the killer use case. Thompson thinks existing Ray-Ban glasses sold better than expected, while Oakley models and a centered camera make sports a possible focus. The objective is a few indispensable experiences—not “a collection of crap and people trying to dig through and find the jewels.”
- Andrew Sharp’s adoption pushback is that useful features still may not justify changing entrenched behavior. A nonglasses wearer must accept something on his face for roughly six hours of battery life, he thinks, with a 30-hour battery life for the case, while captions and live translation solve situations he rarely encounters. Thompson’s answer is that Meta may only be targeting “100,000 units or something like that,” leaving skeptics outside the initial market by design.
- Both hosts admire Meta’s ambition while conceding that success could normalize pervasive surveillance. Sharp called mainstream adoption dystopian if everyone can record conversations; Thompson agreed the issue is “under-discussed, under-appreciated,” probably “inevitable and inescapable,” and something he hates. They similarly resisted Zuckerberg’s vision that VR becomes the future of television because individualized headsets displace watching a shared screen with friends.
Deep dive
1. The Neural Band supplies the interface smart glasses were missing
Meta’s Ray-Ban Display, available September 30 for $799, puts a screen in the right lens for messages, video calls, directions, AI results, live captions, and translation. Users can pinch to select, slide a thumb across a gripped hand to swipe, double-tap for Meta AI, or twist a hand in midair to change volume.
Sharp initially dismissed the promised ability to handwrite words “in the air” as useless. Thompson’s explanation was the episode’s pivotal reveal: because the band detects electrical signals at the wrist, users can write with a hand inside a pocket—quietly replying on a subway without speaking aloud or visibly gesturing.
Thompson framed today’s alternatives as camera-based hand tracking, exemplified by Vision Pro and Oculus, or a physical controller. The band offers “augmented input” to accompany augmented display: “You can be looking at something and no one knows, and you can also be interacting with something and no one knows.”
His confidence comes from direct experience. The technology failed on his wrist several years ago, when he was told that different wrist types could be challenging, but worked perfectly when he tried it again last year; that reversal underpins his categorical claim that the interface works.
2. Every computing paradigm needs a native control system
Thompson’s historical analogy: PCs paired mouse and keyboard, the iPod had its wheel, and the iPhone made multitouch foundational. Pre-iPhone smartphones offered users miniature keyboards or styluses; smart glasses likewise need an interface native to their physical context rather than one inherited from another device.
Vision Pro illustrates the cost of getting input wrong. Thompson said Mike Rockwell viewed weak Siri performance as a serious product constraint because the fallback—“pecking on a virtual keyboard in the air”—was miserable, helping explain why an executive with direct frustration over Siri eventually took charge of that team.
Sharp’s pushback remained practical: even invisible pocket-writing requires new muscle memory, while a phone is already better for composing text. Thompson called that a V1 objection—“You have to start somewhere”—and argued that shipping a credible native interface is momentous even before it beats the phone broadly.
Voice alone cannot resolve the problem because it is socially disruptive and “modal”: a person walking around dictating messages appears to be talking to himself. The wristband’s value is therefore not merely precision but discretion, making everyday public use more plausible.
3. The $799 price turns an experiment into a manufacturing program
Thompson had assumed glasses bundled with the band would cost well into four figures, especially after Orion, a standalone prototype he said was still years away, with current thinking perhaps pointing to 2027, required its own computer, cellular capability, and an external brick. “My jaw literally dropped when he said 799.”
Meta might be “taking a bath” on each unit, but Thompson welcomed the aggression: after roughly $100 billion spent on Reality Labs, “what’s a few billion dollars more” to establish a user base? He sees the band as Meta’s proprietary advantage because neither Apple nor Google has demonstrated an equivalent, while conceding Apple may have undisclosed work.
Shipping itself lowers cost. Thompson’s simple tooling example: a $1 million investment spread across 1 million devices adds $1 each; across 10 million devices, only $0.10. Repetition also improves process efficiency and yields, reducing the number of partially finished items manufacturers must discard: “Making stuff makes it cheaper.”
Sharp added a supply-chain caveat: an FT report described Meta’s reliance on Chinese supplier Goertek. His broader point was that PRC manufacturing expertise lets cutting-edge electronics reach consumers at prices like $799; Thompson agreed Apple helped build a capability ecosystem from which the entire electronics industry now benefits.
4. Product limitations protect Meta from building a worse phone
Unlike Orion, Ray-Ban Display depends on an iOS or Android phone and runs a tightly constrained set of Facebook/Meta-specific experiences. Its maps appear to use OpenStreetMap or similar data and provide walking rather than driving directions; Thompson considers that narrowness “a good thing,” not an embarrassing absence.
His warning is that rebuilding “a phone on your face” is a trap Apple and Google may be especially liable to enter. Vision Pro’s support for iPad apps offered immediate inventory but also embedded experiences not optimized for the new medium; Meta, starting from zero, must build only what fits the interface.
Thompson expects the long-run product to be “AI-dominated,” with just-in-time information replacing app navigation. A user might leave home and automatically see the relevant route because the system may already know the destination—an ambient layer that sees what the wearer sees and requires little explicit interaction.
The phone remains the ultimate app platform: powerful, connected, and large enough to do almost anything. These glasses should therefore sell as an addition that performs a few tasks uniquely well, “not so you can get a worse version of a phone that happens to strap to your face.”
5. Meta needs a beachhead before it can claim the future of phones
Thompson compared the launch strategy with his own subscription rollout: Stratechery first offered two posts weekly, then made more frequent writing available by subscription rather than taking away existing access. New hardware should similarly make users feel they are gaining something desirable, not accepting inferior replacements for capabilities they already have.
Shipping lets Meta observe who buys, who abandons the glasses, and what persistent wearers actually do. Thompson thinks existing Ray-Ban models sold better than expected; he also noted Oakley releases, an explicit sports focus, and a centered camera as reasons sports might become the category’s beachhead, much as health and fitness became central to Apple Watch.
Sharp argued habit change demands a product that is dramatically better, particularly for someone who does not already wear glasses. He said the device may provide roughly six hours of battery life, he thinks, plus a 30-hour battery life for the case; live translation is “really, really cool,” but he cannot imagine enough occasions to use it.
Thompson replied that Meta need not convert Sharp yet. An initial target of “100,000 units or something like that” can be reached without skeptics like him, letting Meta find an additive, aspirational use case and layer on capabilities over time. The dry “Display” name appropriately “underpromises with the potential to overdeliver.”
6. The technological upside arrives with an unresolved social cost
Zuckerberg’s framing—augmented reality as the future of phones and virtual reality as the future of TVs—split the hosts. Thompson considered whether glasses could extend mobile computing from pulling out a device to merely shifting one’s eyes, but said phone replacement remains “a long time away.”
Both rejected VR television as a desirable default. Thompson called it “a little bit of a Bugman world,” imagining himself in a little hovel escaping to a different reality; Sharp preferred a large television, couch, and friends watching together rather than everyone occupying an individualized reality.
Sharp’s deeper objection was dystopian: if Meta succeeds, these glasses could become ubiquitous recording devices. Thompson fully agreed, extending the concern to always-listening AI pendants and noting that even non-display glasses can amplify nearby speech—useful in noise, but potentially socially inappropriate and, as Sharp joked, an “age-dropping” tool.
Their final tension stayed unresolved: Sharp saluted Meta for shipping boldly, accepting live-demo mistakes, and persisting despite years of criticism, yet feared the world created if that ambition pays off. Thompson’s blunt conclusion on the privacy tradeoff: “It’s probably inevitable and inescapable, and I hate it.”