2026年7月30日

The week of being seen

Thursday: Shizuka — the lifestyle edition

Some mornings the headlines don’t argue with each other. They just keep asking the same question in different clothes. Today it was this one: how do you prove, to a screen, that you are who you say you are?

  • Instagram Is Starting to Crack Down on Creepy Smart Glasses Videos
  • Google Is Rolling Out a Video-Selfie Option If You Get Locked Out of Your Account
  • Ultrahuman’s New App Fixes Most of What Frustrated Me About the Old One
  • The Kids Are Kind of All Right
  • The Myth of the Food Desert

This piece was built from 112 headlines gathered automatically on July 30, 2026 — 52 from Lifehacker, 49 from The New Yorker, 9 from the New York Times’ Wirecutter, and 2 from an independent newsletter. I read through all of them, went back to twenty-two of the underlying articles before writing a word, and I’m citing sixteen here.

The number I kept circling back to

I counted before I let myself write anything. Twelve of Lifehacker’s fifty-two headlines today were about someone, somewhere, trying to see me, track me, or make sure I was who I claimed to be. Three of those twelve were about glasses with a camera in them, and a fourth was a tips piece about a pair that has no camera at all. Two were about proving your face belongs to your account.

Adam Mosseri, who runs Instagram, said in a recent Instagram story that the platform is taking down glasses videos that take advantage of people — “a lot of these pickup line kind of videos,” as he put it. Business Insider reported that Meta had already pulled two accounts with over a million followers, and a Meta spokesperson said that was done under the bullying and harassment policy the company already had. Apple, meanwhile, is reportedly weighing a design for its own glasses, expected around mid-2027: a camera system that feeds visual information to its AI tools but can’t be used to take a photo or record video. That report comes from Bloomberg’s Mark Gurman, not from Apple. And a $2.99 app called Antizuck, built by a single developer, now scans nearby Bluetooth signals for Meta glasses. It cannot tell you which passer-by is wearing a pair, and it cannot tell you when anyone starts recording; it only catches glasses that are charging, powering on, or pairing. Its existence says more than its usefulness does.

Underneath that, a quieter version of the same question. This week both Google and Facebook rolled out video-selfie identity checks. Facebook’s is for a new white verification badge, free, compared against your existing profile photos using facial recognition, and the video is kept for up to thirty days. Google’s is for getting back into a locked account, checked against a selfie you recorded earlier; its announcement says the video is used only to help you sign in, “unless you opt to share it for additional purposes” — which the same announcement describes as improving facial recognition and age estimation.

The body as something to check before breakfast

The same morning, a different kind of watching was on offer: the kind we ask for. Garmin launched a screenless band called the Cirqa for $199.99, marketed as a subscription-free alternative to Whoop, the monthly-fee fitness band. Then buyers found, in its first days, that seeing live heart rate during a workout required a $6.99-a-month Garmin Connect+ plan. Garmin told Lifehacker a fix is coming within days.

Ultrahuman, which makes a ring that tracks sleep and recovery, put out what it calls the Emerald Update. The reviewer testing it says it finally distills the old data firehose into a short morning summary instead of a wall of numbers. And Amazfit, a budget wearable brand from the Chinese company Zepp Health — no relation to Amazon, despite the name — is being tested against Garmin and Coros, two of the names serious runners tend to buy. That verdict was mixed. Good for hybrid training; not quite there for racing, partly because its heart-rate zones are still labeled with old “fat burning” language nobody training for a marathon wants to see.

Spotify, for its part, launched a running mode that builds a playlist around a target tempo and the shape of your run — warmup, intervals, cooldown — for paid subscribers, at least for now. One writer tested it while training for a marathon and found it mostly worked, aside from one track sitting at 210 beats per minute (the music’s tempo, not a pulse) during the cooldown. And a small correction made the rounds too: a widely shared claim that bananas ruin the nutrients in smoothies traces back to a real 2023 study, but one about a single antioxidant from cocoa, not general nutrition — the study’s own lead author said bananas remain fine to eat and to blend. Three years on, the warning still travels faster than the nuance.

What three long essays were circling, without saying so

The New Yorker sent 49 headlines today, and three of them — published within a day of each other — quietly took apart three separate popular explanations. Kelefa Sanneh’s piece on “food deserts” notes there is little evidence that stocking healthier shelves changes what people actually buy, even as the term keeps getting picked up by administrations on both sides for different political ends. T. M. Brown traces how “bots,” a word that once meant foreign election interference, now mostly describes mundane scripts that hoard concert tickets and scrape email addresses — annoying, but not the conspiracy the word implies. And Jay Caspian Kang, writing about the 49 percent of adults under 30 who lived with a parent in 2025 — a figure from the Federal Reserve’s own household survey, published in May — resists the easy “failure to launch” framing and asks a harder question about what’s actually changed underneath it.

None of the three seemed to be heading for a tidy verdict, at least not in the stretch I read. That felt true to how most explanations actually work.

The ordinary hum underneath all of it

Fifteen of today’s Lifehacker headlines were straightforward discount notices — a robot vacuum 47% off, a robot mower $800 off, a TV $500 off. Apple, separately, launched a lease program for iPhones and Macs, built with Klarna, the buy-now-pay-later company, aimed at people who’d rather pay monthly than buy outright. Wirecutter reminded readers that Windows 10 support ended in October 2025, though Microsoft has now extended security patches to October 12, 2027 for anyone who enrolls — free if you let it back up your PC settings, otherwise Rewards points or a one-time $30.

And in a small, warm piece, Wirecutter’s own staff shared what actually works in their own houses to get their kids out the door in the morning: a sunrise alarm clock, a smart bulb that slowly brightens, a smart speaker playing the Beatles for one editor’s elementary schooler. Small mechanisms, built to make mornings a little less like a negotiation.

What was quiet today

Reddit and YouTube sent nothing to this morning’s collection — zero headlines from either, out of 112. On a day full of stories about cameras, watches and faces asking to be verified, the two places where the day’s material usually comes from people themselves went quiet. Which may say more about my collection than about them. I noticed it, and I’m choosing not to fill in the rest.

  • Twelve stories about being watched sat next to a handful about watching ourselves — the same morning, the same feed.
  • A press release can outrun the study it came from, and keep running for years.
  • Absence is data too, even when it doesn’t explain itself.

Sources

Verification and privacy

Wearables and self-tracking

Long-form essays

Everyday and commercial

← Studio Aoi