Lifecording
The picks-and-shovels trade behind the rise of always-on AI wearables
A Black Mirror episode imagined a world where every moment of your life is recorded and replayable. That world is arriving, not as an implant but as a wearable, and the cleanest way to own it is to sell the picks and shovels.
In a 2011 episode of Black Mirror called “The Entire History of You,” most people wear a small implant behind the ear, a Grain, that records everything they see and hear. Any moment can be replayed on a screen or inside their own field of vision, scrubbed back, zoomed into, and studied. It is useful, and then it is corrosive. A man replays a dinner party over and over until suspicion hardens into obsession and he uncovers his wife’s affair. In one quiet, telling scene, airport security waves him through only after he lets them scroll through the last week of his life on a monitor. The episode was about memory, jealousy, and control. Fifteen years later, its premise is turning into a product category.
The question worth asking is not whether this is coming. It is. The better question is whether people will be forced into it or will line up to buy it. Our honest read is that the state angle, the officer asking to see a moment, comes later and matters less than most people expect. The thing that drives this is not compulsion. A perfect, searchable memory is one of the most seductive features anyone has ever been offered, and millions of people will pay for it and wear it gladly.
What lifecording is, and who named it
The person who has put the sharpest frame around this is Josh Wolfe, co-founder of Lux Capital. He calls it lifecording, and he means something specific: a wave of wearable hardware that records daily life, audio first, then images and video, with AI constantly listening, summarizing, and surfacing what matters, so your day becomes searchable in the same way your email is.
Wolfe has a habit of naming a shift before the rest of the market does. He publicly pitched Nvidia in 2016, when the story was still about graphics cards for gamers, and argued that it would become a simulation and AI company. The stock went on to return roughly 80x. In 2024, he called the turn in memory chips, arguing that value would shift from commodity DRAM toward the high-bandwidth memory that AI accelerators depend on, and backed SK Hynix. Lifecording, by his own account, is his third conviction call of this kind.
What makes it interesting is that it is pre-category. In his words, Wall Street “doesn’t have a name for this yet,” and to him, that absence is the whole point. By the time there is a clean label and a tidy ETF, the easy money is already gone.
His method is the part worth stealing. Wolfe does not try to guess which gadget wins. He backs the component suppliers that get paid no matter whose logo ends up on people’s faces. He calls it the arms-dealer trade. For lifecording, he built a model basket of nine names, held through his family office rather than the Lux fund and weighted across the hardware stack:
Nordic Semiconductor (OSE: NOD): Low-power wireless connectivity
TDK (TSE: 6762): MEMS microphones and motion sensors
Himax Technologies (NASDAQ: HIMX): Displays, optics, and always-on vision
Ambiq Micro (NYSE: AMBQ): Ultra-low-power edge processing
Infineon Technologies (XETRA: IFX): Power management and sensing
CEVA (NASDAQ: CEVA): Signal-processing and AI intellectual property
Synaptics (NASDAQ: SYNA): Edge AI, connectivity, and IoT processors
Enovix (NASDAQ: ENVX): Next-generation batteries
Cirrus Logic (NASDAQ: CRUS): Audio chips and codecs
His highest-conviction pick was Himax, and so far, that call looks right, which we will come back to. Two things to keep in mind up front: these positions sit outside the Lux fund, and this is an early, high-variance bet rather than a settled trade.
The hardware finally showed up
This pre-category thesis is worth taking seriously right now for one reason: the hardware arrived, all at once, from the biggest companies in tech.
Meta is the one to beat. Its Ray-Ban and Oakley AI glasses, developed with EssilorLuxottica, sold 7M+ pairs in 2025 alone, and Meta accounted for >76% of global smart-glasses shipments that year. In September 2025, it launched Meta Ray-Ban Display, its first pair with a full-color screen inside the lens, paired with a wristband that reads the faint electrical signals generated by muscle activity so you can control the glasses with small finger movements. In June 2026, it pushed downmarket, putting its own name on a line starting at $299 and shipping it with Meta AI powered by Muse Spark, a model built specifically for Meta’s products. Display-free smart-glasses shipments grew 210% YoY in Q1 2026, and Meta is reportedly aiming to sell 10M wearables in H2 2026 alone.
Google spent a decade in the wilderness after Google Glass, and now it is back. At its 2026 developer conference, Google showed smart glasses running Gemini on Android XR, the platform it developed with Samsung and Qualcomm. The frames are being designed with Warby Parker and Gentle Monster, so they read as eyewear rather than gear. The first versions ship this fall as screen-free audio glasses, with cameras, microphones, and speakers but no lens display. Screen-in-lens models will follow. The sequencing makes sense: get people comfortable wearing a camera and microphone first, then add the screen once the habit exists.
Then there is OpenAI, which acquired Jony Ive’s hardware startup for $6.5B to build a family of AI devices. The first is reportedly expected to be a screenless, context-aware device meant to sit alongside your phone and laptop. It is now expected in 2027, and the reasons behind the delay are also the reasons this whole category is not yet a lock. The team is reportedly still working through questions around privacy, compute, and the assistant’s personality and behavior.
Apple was late, but it is no longer absent, and that is why it belongs at the center of this story rather than the margins. The company sat out the first wave while Meta sold millions, but it is now reportedly pursuing three separate AI wearables rather than betting the category on a single form factor.
The first is a pair of display-free smart glasses designed to compete directly with Meta’s Ray-Bans. Production could begin around the end of 2026, with a launch targeted for 2027. The glasses are expected to include one camera for capturing photos and video and another that gives Siri visual context, allowing it to interpret what the wearer is looking at and answer questions about the surrounding environment.
The second is a new version of AirPods equipped with small cameras. These would not be used to take conventional photos or videos. Instead, they would feed visual information to Siri, giving the assistant a better understanding of where the wearer is and what is happening around them.
The third is a camera-equipped pin or pendant that could clip onto clothing or hang from a necklace. It is the earliest-stage product of the three and could still be canceled, but the idea is similar: give Siri a persistent set of eyes and ears without requiring the user to hold up a phone.
The common thread is not the shape of the device but the role of the camera. Apple’s privacy pitch would be to frame the camera as a sensor for Siri, not a recorder for the user. With camera-equipped AirPods in particular, the cameras are reportedly meant to give Siri visual context rather than capture conventional photos or videos. That allows Apple to position them as the assistant’s eyes, not as a conventional always-on recording device. It is a classic Apple move: enter a category late, then reframe it around clearer use cases, tighter integration, and privacy. That matters more here than usual because privacy may be the biggest barrier to mass adoption. A company consumers already trust with their phones, watches, and earbuds may be able to normalize cameras on the body faster than any first mover, while also setting the standards everyone else has to follow.
Vision Pro is separate from those three products. It is Apple’s existing $3,499 mixed-reality headset, or what the company calls a spatial computer, which places digital apps and screens inside the wearer’s physical surroundings. It is technically impressive but far too large and expensive to serve as an all-day social wearable. Apple even released an upgraded M5 version in 2025, so the product itself is not on ice. What Vision Pro demonstrates, however, is why the broader market is shifting toward something lighter, more social, and relatively cheap. Apple is now building for exactly that.
Underneath the giants sits a swarm of smaller always-on gadgets. Note-taking pendants and brooches were all over CES 2026. Amazon acquired listening-wearable maker Bee, while Meta acquired pendant maker Limitless. Not everyone is charmed. When a startup called Friend covered the New York subway with ads for its listening pendant, people scrawled “get real friends” and “surveillance capitalism” across them. That matters. The wearer is not the only person who has to be comfortable with the device.
Forced, or glad to?
So, back to the two roads.
The Black Mirror road, where the recording becomes evidence and someone in authority asks to replay your day, is not science fiction. Police already obtain doorbell footage through legal process, and connected cars already collect precise location and driving-behavior data. A device that remembers everything is also a device that can be made to testify. Over time, courts, insurers, and employers will learn these records exist and ask for them, changing how it feels to wear one. But that pressure is slower and arrives later. It is not what carries the category across the line.
The road that gets us there first is ordinary desire. A tool that lets you never forget a name, recall exactly what a doctor said, find the thing you set down an hour ago, or pull up the sentence someone actually spoke instead of the one you remember is close to irresistible. People traded away far more privacy than this for smartphones and social feeds, and they did it happily because the usefulness was immediate and the cost was abstract and far off. Lifecording will follow the same curve. Most people will not be forced. They will opt in, pay a premium for the better version, and defend it when someone objects.
The friction is real and deserves respect. Recording laws vary by jurisdiction, and several U.S. states require all parties to consent before a private conversation can legally be recorded. Social norms punish the obvious recorder, as the first Google Glass wearers discovered when they were called “glassholes” and banned from some bars and other businesses. Battery life is still measured in hours rather than days, continuous AI use can drain it much faster, and heat remains difficult to manage in something that sits on your face. These are speed bumps, and history says speed bumps slow a technology like this without stopping it. But they could also keep the category stuck for years, so any honest version of the thesis has to take them seriously.
Where to stand in the stack
If lifecording is coming, the investment question is where in the stack to stand. Wolfe’s instinct to sell the arms rather than pick the winning soldier is the right one. Guessing which device wins is difficult, and the winner may not even be public. Selling a component that every device needs is a cleaner way to bet on the category itself.
Below, we analyze each of the nine names in Wolfe’s basket and break down which ones look best positioned as the category scales. We also identify two major potential beneficiaries Wolfe left out, including one that may be the biggest arms dealer of them all. The market is only beginning to recognize this trade, and several of these names are still small enough for a real demand shock to matter.