I'm home! After a month of airports, canceled flights, insane weather, TV segments, gut-splitting laughter with my closest family and friends, and life-affirming moments of "just being" with you — whether in person, over the airwaves, or in pure Techish spirit — I am finally, blissfully, catching up with myself.

And now…Techtember!

It’s the time of year every tech company on the planet unveils the hottest, latest, newest everything — just in time for the holiday rush. We’ll get to that in a minute, but before we get to a million new things, let’s talk about a different, very “now,” number: 2,048.

That’s how many times police say a Kentucky officer looked up the mother of his child in a nationwide network of license-plate cameras. He allegedly ran 241 of those searches after she got a protective order against him.

Everybody files that under tech news, but it’s really a story about a woman who couldn’t get away from somebody, and a company that finally did something about it.

Meta’s week went the same way when they settled a lawsuit over harms their platforms have done to children. In both big stories, safeguards already existed. But it took lawsuits, arrests, and public fury to get them switched on.

Funny how that works.

Also below, a note from Dolly herself about AI and how we need to look at it.

Let’s get into it. 👇

📅 WELCOME TO TECH-TEMBER

Buckle up. IFA, the world’s largest tech show, opens in Berlin Friday and runs through the 8th, which means a firehose of TVs, robot vacuums, e-bikes, headphones, and appliances that talk. Apple takes the stage Wednesday, September 9 at 10 a.m. Pacific to unveil what’s already being lauded as their biggest product rollout in years. Samsung, Sony, and LG will also show off the newest everythings too.

Most of it you can ignore. This part you can’t (thank you for asking me this question yesterday, Deirdre!!!) :

Your iPhone is dying. Buy a 17 now or wait?

Wait until September 9. Here’s why:

Apple is expected to show three phones: the iPhone 18 Pro, the 18 Pro Max, and its first foldable, which could run around $2,000. The regular iPhone 18 isn’t coming until spring 2027. Neither is the less expensive 18e.

If you buy the normal iPhone — not the Pro — nothing new is expected this fall. The iPhone 17 is your phone until spring.

To be clear, Apple hasn’t confirmed any of it, but going off the last umpteen years of Apple unveils, and amazingly well-sourced rumors, this is a solid advice:

• You want the standard model. Buy the iPhone 17, but buy it after the keynote. Apple reshuffles pricing and storage on the older lineup the same day, and carriers dump promos right behind it. Same phone, better deal.

• You want a Pro. Wait. Pre-orders should open September 11, with phones arriving September 18. The 18 Pro is expected to get Apple’s first 2-nanometer chip, a smaller Dynamic Island, and a real camera upgrade. (That’s all code for “awesome.”

• You’re curious about the foldable. Let other people go first. Nobody should spend two grand on a first-generation hinge.

• Your phone won’t survive ten more days. Get the battery replaced. Apple charges a fraction of a new phone, does it same-day at most stores, and it buys you months to decide.

And don’t trade anything in early. Wait until after the keynote, then put Apple’s trade-in number next to the carrier promos and see who blinks. Ten days buys you information.

📡 WORTH YOUR CLICKS

🍎 Point your iPhone at dinner

My favorite new iOS 27 beta trick: open Camera, switch to Siri mode, point it at your plate, tap Look Up Nutrition.

If you log meals, this is a gift. It’s also the feature I wanted a decade ago while thumb-typing every ingredient into Lose It.

The update also splits a restaurant bill with your camera, lets Safari keep watching a webpage for you, gives your alarm its own volume, and finally pulls a full-resolution photo out of a video.

👉 Read: “5 iOS 27 Features That Make Your iPhone More Useful”

✈️ Two apps to install before Labor Day

I had three flights canceled in August because of crazy storms across the Eastern part of the country. One had me stuck in Raleigh for two days (I was supposed to be in New York), and another had me stuck on the tarmac in Chicago for three hours while lightning touched down all around us. Here are two apps that helped me out during it all:

Flighty now tells you what’s actually going wrong at an airport — weather, ground stops, congestion, short staffing — and how it’s likely to hit your flight. It reads the same advisories pilots and controllers use and translates them into English. It’s famous for spotting trouble before the airline admits to it. Best part: you don’t need the app or a subscription. The Airport Intelligence dashboard is free on the web at flighty.com/airports.

AirHelp syncs with Gmail or your calendar, pulls in past trips, and checks them for compensation you never claimed. You can run free checks on flights from the last three years. The real money comes from European passenger-rights rules, so this one’s worth it if you’ve flown Europe. AirHelp takes a cut if it wins the claim.

IN CASE YOU MISSED IT…

🚨 FLOCK FINALLY POLICES THE POLICE

Remember those license-plate cameras I wrote about last year? That story did not age well. Over the past several months, public outrage has swelled into a wave of real action.

Officers have been caught tracking women they knew. Police have stopped innocent drivers at gunpoint over misread plates. Cities have canceled their contracts, and people around the country have started tearing the cameras down themselves.

The inflection point for me came last week, when a Kentucky department switched on Flock’s new officer-auditing software. About a day later, the system flagged Officer Asad Zahir. Police say he searched for the mother of his child 2,048 times, that 241 of those searches came after she obtained a protective order against him, and that he labeled every single one a narcotics investigation.

Spoiler alert: She has never been part of a narcotics investigation. Officers arrested Zahir the next day.

What’s frustrating for me, watching all of this unfold, is the fact that Flock finished building that auditing software months ago and then made it optional. Only about a third of agencies turned it on. Now the company says every department has to use it by the end of the year. Why not just start with those safeguards in place? I asked Flock’s CEO, Garrett Langley, that exact question a year ago.

Listen to what he said in our extended podcast here.

The pressure keeps building. Sen. Josh Hawley opened a Senate investigation this week, and on Saturday, Texas Gov. Greg Abbott froze state spending on Flock cameras.

All of that counts as real progress, yet let’s keep the confetti in the box. The new tool watches officers, not cameras. It still hasn’t fixed bad plate reads, and it doesn’t touch the bigger question of how much of your daily movement belongs in a searchable police database.

What to do this week:

1. Run your plate through the nonprofit Have I Been Flocked. It has collected nearly 242 million police searches covering 4.6 million plates. While you are there, visit Have I Been Pwned, the site that inspired it, and find out whether your email address has turned up in a breach.

2. Read a blank result correctly. It doesn’t mean nobody searched you, because most departments still haven’t released their logs.

3. Ask your city three questions. Does our police department use Flock? Has it turned on Audit Assistance? Who outside our department can search our camera data? Your council or your chief should be able to answer all three. If they refuse, that tells you something too.

👉 Read my full story: “Have I Been Flocked?”

This all begs the question: how do we get these big companies to do “the right thing” before people get hurt?

Leave a comment

📱 META WOULD ALSO LIKE A GOLD STAR

Meta has agreed to pay up to $17.1 billion to settle child-safety claims brought by 47 states, D.C. and three territories, plus roughly another $1 billion to Texas.

Also? Meta will give teens ages 13 to 17 a two-hour daily cap across the apps. A nudge lands after 15 minutes of scrolling, notifications stop during school hours, features go quiet overnight, and counts hide by default.

But it took horrible things — like kids dying — to make them finally do what they should have done years ago. It’s the Sackler moment, the big tobacco change, and it’s just the tip of the iceberg in putting protections in place before their products cause harm.

What it is NOT is how Meta’s now trying to spin it — describing the settlement as evidence of its leadership on teen safety. UC Law professor Robin Feldman gave me the line of the week: “Escaping a lawsuit is not the same as good behavior.”

Exactly.

What parents should actually do: Don’t assume this lands on your kid’s account automatically. The changes roll out over months, and parents can override most of them — which means your teen now has a very specific reason to come ask you for a favor. Open your teen’s account settings yourself, confirm what’s switched on, and decide before the negotiation starts.

And keep the ceiling in mind. Two hours on Instagram and Facebook doesn’t cap TikTok, YouTube, Snapchat, or whatever launches next month.

👉 Read my latest here, then tell me what you think: “Meta Settled a Massive Lawsuit Around Harming Kids. Then It Threw Itself a Parade.”

The safeguards existed. Pressure made them mandatory. If the safest version of your product only ships after the arrests and the settlement, safety wasn’t the priority.

It was the bill.

🤯 That horrifying video didn’t happen

Nepal’s floods are devastatingly real. Many of the clips going viral right now are not.

One came across my feed this week showing floodwater tearing through a Nepali town, with buildings buckling and motorbikes tumbling end over end. AI made it, and more than 50,000 people liked it before anyone flagged it.

Don’t believe everything showing up in your social feeds now, including fake disasters, fake protests, fake outrage bait dressed up as news, and beautiful influencers who don’t exist. These AI-generated videos keep getting better, and they show up right beside the real thing.

You don’t need to be an AI expert to catch most of them. Two checks I run before I share anything:

• Pause on a still frame and hunt for text. Storefronts, license plates, street signs. AI still turns lettering into gibberish.

• Screenshot it and reverse-image search. If a genuine disaster killed hundreds of people, a wire service covered it. No news coverage, no video.

👉 Read our guide to spotting fake AI video before you share it.

💔 And then there’s Dolly

Like most of you, the death of an icon, legend, all-around amazing human being, Dolly Parton, got to me this week — and here’s how it touches our techy world:

A few years back, she told The Hollywood Reporter she wanted nothing to do with AI recreating people. She called it a weapon, and said plainly that she didn’t want her voice rebuilt after she was gone: “I want to just leave a body of work behind,” she said.

After she died on Tuesday, it took mere hours for AI fakes to hit our feeds, including this one, shared by the President of the United States.

Within just a few days, fact-checkers were already debunking these bogus images, along with a fake recording of her supposed final wish, built from real Dolly footage with an AI Dolly voice laid over the top. It spread the way these things always spread, because people wanted it to be true.

So if you remember nothing else from all the big techish news this week, remember this: a fake doesn’t have to be convincing to spread. It only has to be something we want to be true, because hoping is what stops us from checking.

Dolly spent years asking us not to do this to her. It took the internet about two days.

So run those two checks, especially on the stuff that makes you cry, gasp, or get furious, because that jolt is exactly what gets us sharing before we think.

And if you want to do something real, her family asked for donations to Dolly Parton’s Imagination Library in place of flowers. It has mailed hundreds of millions of free books to kids who wouldn’t otherwise have had them.

Go read one to a kid this week.

Okay, that feels like plenty of technology for one week.

Use the good stuff. Question the rest. And go outside. ❤️

XO, Jenn

It was awful — and totally fake.

Nepal’s floods are horrifyingly real. So are the people who have died, the families grieving them, and the unimaginable destruction left behind. But AI made that particular clip, and more than 50,000 people liked it before anyone flagged it.

There are more fake videos on your feeds than you might imagine these days. A 2026 investigation shows nearly 3 out of every 5 videos served to a brand-new TikTok account are AI-generated “slop.” A similar test of YouTube Shorts found about 1 in 5. I haven’t found any comparable numbers for Facebook or X yet, but that feels like where I see the worst of it all these days.

Disaster footage, political trolling, and completely invented “news” are now mixed into our feeds with the real thing. My neighbor recently asked me about World Cup soccer riots she’d seen online. I was at that game; they never happened.

The problem isn’t just that AI can make convincing video now. It’s that real and fake are showing up side by side, looking increasingly alike.

Now for the good news: You don’t need to be a forensic video expert to sort out a lot of it. You just need about 90 seconds and a few very simple checks.

Here’s how I do it.

1. Look at the length

Videos less than 15 seconds are the quickest tell. The Nepal clip that I mentioned was ten seconds long.

Most AI video tools, especially the free ones, only make clips of roughly ten to fifteen seconds, so that’s what gets posted, but real footage doesn’t behave that way. When somebody films a disaster, they film for a minute, two minutes, however long their hands stay steady. The camera swings around. They say something. They forget to stop recording.

A short clip isn’t proof of anything on its own. Plenty of real videos get trimmed. But a dramatic ten-second clip with no beginning and no end is worth a second look, and the tools are getting longer, so treat this one as a hint rather than a verdict.

2. Let a detector take a look

This is the easiest tech check I’ve found so far.

Open Hive Detect, then either upload the suspicious video or paste in the media URL. Hive scans the clip for signs of AI generation or manipulation and gives you a confidence score. It works across images, video, and audio, so it can catch more than a Google-only check.

Then, if you want a second opinion, try Gemini. Upload the video and ask: “Was this video created or edited by AI?”

Gemini checks the file for Google’s invisible SynthID watermark and for Content Credentials, which can show where a file came from and whether someone used AI to make or edit it.

If Gemini finds SynthID, Google AI made or edited at least part of the video.

If Gemini doesn’t find it, that still doesn’t prove the video is real. Another AI tool could have made it.

That’s why I like using both: Hive gives you a broad first pass, and Gemini gives you a more specific check for Google AI.

PRO TIP: Here’s how to get the video on your phone via screen recording:

→ If the social app gives you a Download or Save video option, use that. If it doesn’t, you can screen-record it.

On iPhone: Swipe down from the top-right corner and tap the Screen Recording button — the circle inside a circle — then play the video. Tap the red indicator at the top when you’re done.

On Android: Swipe down from the top of the screen, usually twice, and tap Screen record. Play the video, then stop the recording when you’re finished.

3. Check whether anyone else has it

Take a screenshot, then reverse image search it. On a phone, you can usually press and hold the image, or use Google Lens.

What you’re looking for is whether real news organizations are running the same footage. When something genuinely happens, dozens of people film it, and reputable outlets pick it up. If the only place this astonishing footage exists is one account you’ve never heard of, that’s your answer.

4. Check whether it’s old

I came across this video of an elephant supposedly saving a man from floods in Nepal. Turns out the video is real, but it’s from a Monsoon in India earlier this year.

Screenshot the clip and reverse image search it. On a phone, press and hold the image, or use Google Lens. You’re looking for the same footage published earlier somewhere else, about something else. Floods recycle especially well, because water…looks like water.

For this one, you can also just Google “elephant saves man, Nepal,” and the truth comes out.

Tells that don’t need a search: signage in a language that doesn’t match, license plates, the wrong season — or snow on the hills when it’s monsoon.

5. Another giant tell is extreme polarization.

This last one is less about the current disaster and more about fakes in general.

If an account seems to exist for one reason, like posting things engineered to make one side furious — take a closer look. Of course, I’m talking about politics. The New York Times did a great story about it here.

Most of these faking-faker accounts are click farms: pump out outrage, rack up shares and engagement, make money. They get richer, and the rest of us get…well, duped.

But as the internet gets better at faking reality, we truly can get better at spotting it.

What’s your trick for spotting the fakes these days, and how do you let people know they’ve shared AI? (Erm, asking for a friend…)

Leave a comment

But you don’t necessarily need a new iPhone to get some pretty great tricks right now.

I was among several reporters Apple briefed last week on some of the latest features coming with iOS 27 this fall. But you can jump in right now with the beta version, and for the first time ever, I recommend you do. (I first told you this in July, and that advice has only gotten stronger since.)

If you already have iOS 27 Beta installed, skip down to #1. If you’re still on the fence and want to see if your current phone can handle it, read on.

First: Can you get iOS 27 now?

Yes, but only if you’re willing to run beta software. Apple’s iOS 27 public beta is available now through the Apple Beta Software Program. After you enroll, go to Settings > General > Software Update > Beta Updates > iOS 27 Public Beta.

Back up your phone first, and understand why. You can’t simply uninstall iOS 27 and roll back to iOS 26. If you decide you hate the beta and want to go back right away, Apple requires you to erase the phone and restore it from a backup you made before installing iOS 27. A backup made while you’re running iOS 27 won’t restore to iOS 26. No pre-beta backup means no easy way to undo it. It’s also beta software, so apps can crash, battery life can suffer, and things can act weird.

And they have since I’ve been using it, BUT, it’s still worth it. My battery has died faster, Siri works “most” of the time, and I’ve had to turn my phone off and back on again because it’s acting “funny.” Again, all still worth it.

If you’d rather let the rest of us find the bugs, wait a few more weeks. Apple says iOS 27 arrives this fall. The company hasn’t announced the exact release date yet, but Apple typically pushes out its big iOS update shortly after its September iPhone event.

And this year, Apple didn’t drop a single phone from last year’s compatibility list. iOS 27 runs on 31 iPhone models, going all the way back to the iPhone 11 family and second-generation iPhone SE. At WWDC, Apple software chief Craig Federighi said that makes iOS 27 available to more people than any iOS release ever.

The list includes the iPhone 17 Pro Max, 17 Pro, 17, 17e, and iPhone Air; every iPhone 16, 15, 14, 13, and 12 model; the iPhone 11, 11 Pro, and 11 Pro Max; and the second- and third-generation iPhone SE.

One catch: the first three features on this list require Apple Intelligence, so you’ll need an iPhone 15 Pro or Pro Max, or an iPhone 16 or newer. Apple is also using a waitlist for the new Siri in the beta, and code in the latest beta suggests that waitlist could remain when Apple releases it publicly.

1. Point your camera at your food and get nutrition information

All you have to do is point your camera at your food, choose Siri at the bottom next to Photo, Portrait, and the other camera modes, then tap Look Up Nutrition. Your iPhone gives you a rundown of what you’re about to put in your mouth. You can also use it with a food photo that already lives on your camera roll. (The steps for an existing photo are a bit different, but I show you how to do that below.)

This is genius for anyone who logs food regularly, and it’s something I first wished for about 15 years ago when I reported on the Lose It app to track nutrition. They’ve added this feature now, as have several other health, fitness, and food-logging apps over the past few years. Still, I like having it built right into the iPhone, and wouldn’t be surprised if Apple eventually ties some version of this into Health. That feels like the obvious next step.

Take a look at what it did with a fruit plate photo that I had in my camera roll. I swiped up on the photo, tapped Look Up Nutrition, and iOS 27 immediately identified watermelon, papaya, dragon fruit, and passion fruit. Then, it labeled the whole thing a Mixed Fruit Plate, called out the fiber in the dragon fruit and passion fruit seeds, noted that it had no added salt, sugar, or saturated fat, and rated its nutritional value “Very High.”

The video above is the first time I ever tried it, and you can see my experiment firsthand when I decided to push it further (because let’s be real, I don’t need general info; I want specifics). I asked Siri for portion sizes and calories, and it came back with an itemized table: about a cup of papaya, roughly 60 calories; two small slices of watermelon, about 45; and two wedges of dragon fruit, around 30. Apple didn’t mention that part in its briefing, and that’s the part I now actually use.

Apple does put a warning right on the nutrition card that visual identification “can be inaccurate or incomplete.” I wouldn’t trust my camera to nail the calories in a complicated restaurant meal, spot anything you might be allergic to, or even be more than “somewhat accurate.” But for a quick “What am I actually eating?” check, this works surprisingly well.

2. Point your phone at the restaurant check and make it split the bill

This could eliminate the most annoying five minutes of every group meal. Point your camera at the receipt, choose Siri, and iOS 27 turns the whole check into a list of individual items you can tap to claim.

Apple showed me an example from a three-person breakfast at Wingen Bakery in Livermore. The phone read the entire receipt — egg sandwich, biscuits and gravy, avocado toast, breakfast plate, croissants, latte and Americano — then let one person choose exactly what they ordered.

Here’s the part I love: it handles shared food, too. In Apple’s demo, the person selected a $14 egg sandwich, a $5.50 latte, and half of the $14 avocado toast. The phone recalculated the avocado toast to $7, added the person’s share of the 9.25% tax, and totaled $28.95 before tip.

You can add the tip right there, then tap to pay the person who picked up the check with Apple Cash. No calculator, no Venmo math in the parking lot, and no making the person who ordered coffee subsidize somebody else’s biscuits and gravy.

You’ll need an Apple Intelligence-capable iPhone for this one, and Apple says it works best when everyone involved is running iOS 27.

3. Make Safari watch a price or restock for you

This one could save me an embarrassing amount of time. Say you want something that costs too much, you’re waiting for a sold-out item to come back, or you keep checking to see whether tickets have gone on sale. Open the webpage in Safari, tap the menu button in the tab bar, and choose Notify Me.

Then tell Safari exactly what you want it to watch. You can ask it to notify you when the price drops below $200, when something comes back in stock, or when a page changes in some other specific way. You can also tell Safari how often to check.

Then get on with your life. Safari loads the page in the background, checks for the change, and sends you a notification when it happens. No refreshing the same stupid webpage 27 times a day. You’ll need an Apple Intelligence-capable iPhone for this feature, too.

4. Give your alarm its own volume

If you’ve ever slept through your Apple alarm (daughter, I’m looking at you…) Apple finally fixed this, and it works on every iPhone that runs iOS 27. Go to Settings > Sounds & Haptics > Alarms and Timers, then turn off Match Ringtone Volume.

Now your alarm gets its own volume slider, so you can keep your ringtone at a civilized level and still make your morning alarm loud enough to wake the dead. This sounds ridiculously minor until you remember how many people have turned down their ringtone at night and accidentally neutered their morning alarm.

Why did we wait this long for a separate slider? Excellent question.

5. Pull a real photo out of a video

I love this one because I shoot video all the time when I’m trying to catch the exact moment — a kid jumping into a pool, a dog doing something ridiculous, lightning, a flight over a volcano, whatever.

Until now, the easiest iPhone workaround was to pause the video and take a screenshot. That works, but you lose image quality. With iOS 27, open the video in Photos, scrub to the exact frame you want, tap the three dots in the upper-right corner, and choose Save Video Frame as Photo.

Your iPhone saves that frame as a separate photo at the video’s full resolution and keeps info like when and where you shot it. I tested it on a video I shot out an airplane window as Mount Rainier came through the clouds, and I got the second and a half that mattered. GENIUS.

You probably don’t need a new iPhone to get a better iPhone

Apple will spend Sept. 9 showing off whatever new hardware it wants us to buy, and like normal, I’ll probably salivate, clap like a seal and gush over it like I just got invited backstage by my favorite band. But I’m just as interested in what Apple can add to the phone already sitting in your hand.

Let Safari watch the sale. Let the camera read the restaurant check. Give the damn alarm its own volume. Those are the updates I care about most: the ones that remove some tiny daily annoyance I’d stopped noticing because I assumed my phone just couldn’t do it.

And those don’t cost another $1,000.

Try them and let me know what you think! Questions? Send those too!

Leave a comment

Then some PR genius went online and announced that Meta had “partnered with state attorneys general to set a new industry standard.”

Ewwwwwwww. Seriously, I can’t believe they haven’t taken this down yet:

🤯 The facts tell quite a different story, and this spin is outrageous.

The truth is, the attorneys general spent nearly three years fighting Meta in court over allegations that it designed Instagram and Facebook to keep kids hooked, collected data from children under 13, and misled families about how safe its platforms were. Meta denied the allegations, fought the case all the way to trial, and only settled after a jury had been seated and witnesses had started testifying.

Calling that a partnership takes some nerve. (Here’s part of what I posted on Instagram the day of the Settlement…)

So why did Meta settle?

Because the trial was making them look bad. Like really bad. Settling is a far better chess move for their bottom line than the daily onslaught of OMG, these people are evil.

Former Meta safety engineer Arturo Bejar testified that the company prioritized engagement and profit over safety and called its “Take a Break” feature “designed to fail.”

Then, George Volichenko, a data scientist who worked on Instagram’s teen mental-well-being team, put a number on just how badly one version of that feature worked.

He testified that only 0.165% of teens enabled the opt-in break. When Volichenko raised that miserable adoption rate with his manager, he said his boss told him not to worry because the team existed partly to protect the company from upcoming lawsuits.

The evidence about Instagram “likes” wasn’t much better, either. Instagram essentially turns popularity into a public score, and Meta studied whether hiding those numbers might reduce some of the pressure teenagers feel on the app. The attorneys general presented evidence that Meta estimated making like counts private by default could reduce advertising revenue by about 1%. So, Meta chose not to do it.

Now, the settlement requires Meta to make hidden likes the default for teens — one of several changes the company previously resisted when safety collided with engagement and revenue.

OKAY. This means Meta’s settlement isn’t about doing what’s right for kids, it’s about saving face before a jury could tell them — and the rest of the world — how awful they are.

Every day the trial continued, witnesses could expose more internal conversations like these, and more Meta documents entered the public record. Zuckerberg was still expected to testify, and Meta had another reason to worry about letting a jury decide what happened next: the company already suffered two major child-safety losses this year.

New Mexico hit Meta with $942 million across a jury verdict and court ruling, and ordered major changes to Facebook and Instagram. A Los Angeles jury also found Meta and Google liable in the first individual social-media-addiction trial.

At the same time, Meta and the other major social platforms still face more than 3,000 individual lawsuits from young people and families who accuse them of designing products that hooked kids and damaged their mental health.

An Oakland verdict wouldn’t automatically decide those cases, but another major loss could give future plaintiffs a tested pile of evidence, a roadmap for their own cases and a much clearer idea of what juries might be willing to award.

Meta settled before this jury got the chance.

What does $18 billion actually buy Meta?

Quite a lot, actually.

The states were seeking roughly $200 billion at the Oakland trial. Meta agreed to pay up to about $18 billion over ten years, and only about $12.7 billion of that amount is guaranteed.

Meta brought in nearly $201 billion in revenue last year, which means the guaranteed portion of this settlement works out to about $1.27 billion a year, or roughly two days of Meta revenue annually, paid out across all the states and territories. That final dollar amount landing in each state to mitigate the damage they’ve done is the equivalent of buying a Starbucks latte to a company like Meta.

Also, even if Meta eventually pays the entire amount, the average yearly cost comes to a little more than three days of its 2025 revenue.

In exchange, Meta ends this trial, avoids another jury verdict, keeps Zuckerberg off the witness stand, and stops this particular stream of internal documents from spilling into public view.

The company also keeps the guts of the business that made it nearly $201 billion last year. Meta can still personalize feeds, sell targeted advertising, and make money from all those hours people spend scrolling Instagram and Facebook.

And now to the really interesting part: Meta may never pay the final $5.3 billion.

Why does Meta care what TikTok and YouTube do?

Meta only owes that final 30% if TikTok and YouTube meet two conditions. Both companies have to adopt one-hour daily limits for teens, nighttime restrictions, and stronger age checks, and make comparable payments to the states. Meta tied half of its remaining payment to YouTube and the other half to TikTok.

Meta also agreed to tighten its own restrictions if the other platforms join. The initial settlement gives teens a combined two-hour daily limit across Instagram and Facebook and blocks most access from midnight to 6 a.m. If its competitors adopt the new standard, Meta will cut its limit to one hour per app and expand Night Mode from 10 p.m. to 7 a.m.

Most of Meta’s settlement obligations last ten years, but the company only commits to the initial Time Limit and Night Mode for five years. If the other platforms sign on, Meta extends those protections to ten years and makes them tougher.

Meta is now running full-page newspaper ads urging TikTok and YouTube to do exactly that. They’re pretending they’re doing it for our children's health. I find that insulting and infuriating. I hope you do too.

The company argues that teen protections work better when every big platform follows the same rules, and well, no duh. Meta says teenagers move from one app to another when they hit a restriction, so limiting Instagram does less good if TikTok remains wide open. Right. And, that’s why Meta has such a huge business incentive to pull its competitors into this deal. Let’s not dress it up as them taking the lead on anything except massive PR spin.

Attention drives this entire business, and Meta just negotiated a settlement that pressures its biggest competitors to accept many of the same restrictions it now faces — because it got caught — not because it’s finally trying to do less harm to our kids.

What changes for kids?

The settlement gives people under 18 a two-hour daily limit across Instagram and Facebook that teens can’t turn off without a parent. It blocks most access from midnight to 6 a.m., mutes most notifications during school hours and gives teens repeated reminders about how long they’ve been scrolling.

Meta also has to improve its age-detection systems to identify children under 13 who use its platforms by putting in a fake birthday, and give parents stronger supervision tools. The agreement requires Meta to hide like counts by default for teens, disable cosmetic-procedure filters, and give teens the option of a feed that doesn’t rely on personalization.

We’ve long known those changes could help kids. I talked with Jonathan Haidt and Catherine Price about it earlier this year.

The settlement also creates an independent social-media research foundation that can study data from teens who opt-in, while an independent auditor will check Meta’s compliance with the agreement. Researchers have spent years asking social-media companies for better access to platform data so they can study what these products do to children and teenagers. This agreement could give them more of it.

A federal judge still has to approve the settlement, but these protections are a step in the right direction — one Meta took begrudgingly and years after families, researchers and its own employees started sounding the alarm. For the thousands of young people and families already suing over the harm they say these platforms caused, they also come awfully late.

Meta, just tell the truth already

One more piece of this agreement deserves attention. California included a provision barring Meta from making “false, misleading, or deceptive statements around its safety features.” The states added that requirement after spending years accusing Meta of misleading children, parents, and the public about the risks of Facebook and Instagram.

Then Meta announced that it had “partnered with state attorneys general to set a new industry standard.”

I don’t know whether calling this a partnership comes anywhere close to violating that injunction, so I asked someone who does.

One more interesting point in all of this: California included a provision barring Meta from making “false, misleading, or deceptive statements around its safety features.” The states added that requirement after spending years accusing Meta of doing exactly that.

So what did Meta do next? It announced that it had “partnered with state attorneys general to set a new industry standard.”

I wondered whether that claim could run afoul of the new prohibition, so I asked Robin Feldman, a distinguished professor of law and director of the Center for Innovation at UC Law.

Probably not, she told me. The prohibition against false or misleading safety claims is broad, which leaves regulators room to come back if Meta crosses the line later. But she thinks the language in this particular social-media post is too general to violate it.

She did, however, call the post “remarkably tone deaf,” and added,

“Meta’s lawyers did get a good settlement, but escaping a lawsuit is not the same as good behavior. It’s a little like saying, ‘I’ve promised to stop beating my wife. Look what a good Boy Scout I am.’”

Reaching a favorable settlement does not erase the conduct that brought Meta into court in the first place.

The rest of us can still read a timeline: The attorneys general investigated Meta, sued the company, fought it for years, seated a jury, and started putting witnesses on the stand. Meta denied the allegations and fought the case until the trial was underway.

Then they settled and threw themselves a PR parade. Sure, they can call the resulting protections an industry standard. It can urge TikTok and YouTube to adopt them. I hope they do. At best, though, calling what happened a partnership leaves out the fight that forced these rules into existence, and the damage they did along the way.

A judge still has to approve the settlement, and thousands of lawsuits against Meta and the other major platforms are still moving through the courts. Parents don’t have to wait for any of that.

Now what? Parents, read this:

If your teen uses Instagram, Facebook, TikTok, YouTube, Snapchat, Roblox, or Discord, sit down with them and go through the account together. Make sure they use their real age, turn on the strongest teen and parental controls available, agree on a daily time limit, and decide when the apps go away at night.

Do it across every app they use, because Meta got one thing right: shutting down Instagram at midnight doesn’t help much if TikTok stays open until 2 a.m.

These companies have spent years competing for our kids’ attention. We can stop handing it over quite so easily.

Leave a comment

I’m both outraged and insulted by the press statement and Meta’s overall handling of this. What do you think? Let’s talk about it…

Other great reads:

TechCrunch: Meta’s child safety deal hinges on age verification that doesn’t work

Platformer: Meta settles with the states over child safety failures

The company didn’t do this on a hunch. Officers kept getting caught using plate readers to search for women they knew. Innocent drivers got stopped at gunpoint after bad alerts. Police used the sprawling camera network for thousands of immigration-related searches, sometimes helping federal agents reach Flock data through local departments even where state law restricted that use. Meanwhile, cities canceled contracts and people started cutting Flock cameras off their poles in the middle of the night.

Flock had actually built software to look for police misuse back in April of this year. Audit Assistance scans an officer’s searches for unusual patterns, like repeatedly running the same plate, and can flag or lock an account for review. Flock made the tool optional, and only about a third of its law-enforcement customers turned it on.

Then, on August 13, after a Washington Post investigation exposed dozens of officers using license plate readers to track girlfriends, wives, exes and other people they had no legitimate reason to follow, Flock did something it should have done a year ago. It made its audit software mandatory by the end of 2026, required case numbers for searches, and cut its default data-retention period from 30 days to seven.

How have police misused Flock cameras?

Earlier this week, the Shively Police Department outside Louisville, Kentucky turned on Audit Assistance. Within about a day, it flagged one of their own.

Police say Officer Asad Zahir allegedly searched two license plates belonging to the mother of his child 2,048 times between January and May, tracking her movements across Kentucky and Indiana. More than 200 of those searches happened while she had a protective order against him. Investigators say he repeatedly labeled the searches as drug or narcotics investigations despite having no legitimate law-enforcement reason to track her. Police arrested him the next day.

The details are extreme, but the underlying abuse isn’t unusual anymore. Police departments around Georgia spent the summer investigating officers accused of using Flock to search for girlfriends, former partners, and other people from their personal lives.

Flock CEO Garrett Langley now says the company “got this one wrong.” Specifically, he says Flock should not have left it up to individual departments to decide whether to monitor their own officers.

That is a very different answer from the one Langley gave me last September. “I am the technology part,” he told me then. “I can’t write the policy and I’m not the person who’s going to enforce it.” He later summed up Flock’s position even more clearly: “We build the tech. But your community makes the rules.”

UC Davis law professor Elizabeth Joh questioned that logic when I interviewed her for the same story. Self-policing doesn’t work particularly well in most industries, she told me. Why should we expect it to work here?

Eleven months later, Flock is imposing safeguards itself because leaving the decision entirely to police departments didn’t work.

You can read my September 2025 USA Today story HERE and a longer Techish version HERE. Listen to the full podcast interview with Flock CEO Garret Langley HERE.

Flock’s second problem has nothing to do with bad cops

Police abusing their access is one failure. Cameras and police departments getting the information wrong is another.

In Morristown, Tennessee, a license plate reader mistook an “O” for a “0,” and police detained two grandparents at gunpoint while their three-year-old granddaughter watched from the car. The non-partisan Institute for Justice has documented more than two dozen cases where license-plate-reader errors contributed to innocent motorists being stopped, detained or arrested.

This month in Wisconsin, officers shut down Interstate 94, boxed in a driver and ordered her out at gunpoint because of an alert Milwaukee police should have cleared days earlier.

Roseville, California, shows what should happen next. Police there reviewed 1,427 automated Flock alerts involving stolen or felony vehicles and found the camera had misread the plate in 71% of them. None of those misreads resulted in a stop or arrest because officers independently checked the plate before acting.

That distinction matters because Flock’s new software watches officers using its database. It doesn’t make a camera better at telling an O from a zero, clear a stale alert or require an officer to verify what the computer says before escalating a traffic stop.

“I’m not doing anything wrong. Why should I care?”

I get this question every time I write about these cameras, and it deserves a straight answer. Your own good behavior doesn't protect you from either problem.

The woman in Kentucky didn’t have to commit a crime for someone with access to allegedly track her. The grandparents in Tennessee didn’t do anything wrong before police pointed guns at them. What protects ordinary people are the safeguards between a camera taking a picture and police acting on what the system says: who can search, why they can search, who reviews those searches, and what officers must confirm before making a stop.

There is another side to this, and it matters too. Police departments say Flock helps them recover stolen cars, locate missing people, and solve violent crimes. Flock says its network supported about a million investigations last year and helped locate more than 10,000 missing people. Those are the company’s figures, not an independent audit, but it’s clear this technology can help police. It’s just — at what cost?

Why are people so angry about Flock cameras?

A growing number of Americans don’t trust the rules around these cameras, and some take that anger far beyond a city council meeting. The Guardian has documented at least 33 incidents across 23 states where people damaged or destroyed Flock cameras, including all eight cameras in Winona, Minnesota, worth $24,000. One online group is even trying to turn Halloween into “De-Flock America” night, urging people to disable, obscure, or destroy cameras under the cover of masks and costumes.

More than 50 communities have also canceled, suspended or rejected license-plate-reader contracts. On August 26, Republican Sen. Josh Hawley opened a Senate investigation into how Flock collects, keeps and shares Americans’ driving data. This is no longer a tidy left-versus-right privacy fight.

What’s next for Flock Cameras?

Flock’s new safeguards are a step in the right direction, but someone outside the company now needs to independently evaluate Audit Assistance. The Kentucky case tells us the software can catch an officer who searches for the same woman more than 2,000 times, but can it catch more subtle abuse?

What happens if an officer runs a plate 20 times instead of 2,000? Five times? What happens when somebody learns how to search without triggering the software? We don’t know, because Flockdoesn’t publicly disclose how Audit Assistance identifies suspicious behavior, and no independent evaluator has tested how much it catches or misses. They need to let one.

Police departments should also publish audit information showing who searches their camera data and why. Officers should verify every plate before an automated alert turns into a high-risk stop. States should set clear rules for when police need a warrant to reconstruct someone’s past movements, with narrow exceptions for genuine emergencies.

Last September, Langley told me, “I actually need the critics. I need the critics to make us a better company.” Well, he got them, and the pressure changed Flock. Now the rules need to catch up with the cameras.

You can check yourself in about a minute

A nonprofit called Have I Been Flocked (haveibeenflocked.com) has collected police audit logs through public records requests — nearly 242 million searches across 4.6 million plates. Type in your plate and it tells you whether an officer looked you up. It’s free and it doesn’t save what you enter. A blank result doesn’t clear you, though, because most departments have never released their logs at all.

That’s how a woman in Jefferson, Georgia learned her ex-boyfriend, a police chief in the next town over, had run her plate 600 times.

September 2025 — I interview CEO Garrett Langley for USA Today. He says oversight belongs to cities, not his company.

October 2025 — Sen. Ron Wyden tells Flock it can’t keep its privacy promises and urges communities to reconsider.

November 2025 — Wyden and Rep. Raja Krishnamoorthi ask the FTC to investigate Flock for not requiring multi-factor authentication.

January 2026 — A federal court upholds Norfolk’s camera system. Cities begin canceling anyway.

April 2026 — Flock quietly launches Audit Assistance and makes it optional. About a third of agencies turn it on.

July 2026 — Langley apologizes for calling critics terrorists. Five former Albany, Georgia officers are arrested.

August 2 — The Washington Post reports 50 officers accused of misusing plate readers.

August 13 — Flock makes the audit tool mandatory by year’s end, cuts retention from 30 days to seven, requires case codes and two-factor authentication.

August 24 — Shively, Kentucky turns the tool on.

August 25 — Shively arrests Officer Asad Zahir.

August 26 — Sen. Josh Hawley opens a Senate investigation.

Jennifer Jolly is an Emmy Award-winning tech journalist and host of Techish. Reach her at jj@techish.com.

Leave a comment

For the record, I was wearing a gray t-shirt that says DOGS ARE PEOPLE TOO with navy pajama pants covered in tiny dogs. Clearly, I’m committed to a theme. (For the Love Island stuff, you’ll have to watch it yourself to understand that particular cultural obsession.)

To be clear, she wasn't saying I looked like a dog. She was saying I had committed to a theme, which is what maxxing means: picking a thing and going all the way in. It's one of three phrases that are following me, and everyone else online, around right now. There’s also AI slop and the newest one, “PervPods” — that are the latest “need to know” buzzwords defining our wired world. Or at least helping you understand what the hell your kids are talking about. #servicejournalism

Let’s dive in.

What’s AI slop?

I want to say that you know it when you see it, but sadly, many people don’t. (Hey Mom, Facebook’s filled with it!) At its core, AI slop is low-quality, mass-produced digital content — text, images, audio, and video — generated cheaply by artificial intelligence to game online algorithms and chase engagement. If you've felt like your feed has gotten dumber lately, that's the word for why.

Some of the most popular examples include Shrimp Jesus, the half-man, half-crustacean messiah that colonized Facebook two years ago. (That was the starter version and looks almost charming now.)

More recently, AI slop has taken the shape of a handsome silver-fox podcaster who does not exist. Futurism found a whole crop of them flooding Facebook just last week — distinguished older men filmed mid-monologue, explaining to women everything that’s wrong with men, racking up comments from people who have no idea they’re talking to nobody.

When Futurism ran the pages through Google, the AI Overview described the hosts as real creators, and even after being pressed, insisted that a guy named “Roman Black” was a real person. He isn’t. The machines are now vouching for each other.

It isn’t only images and podcasters, either. Streaming music service Deezer recently said that people upload nearly 90,000 AI-generated music tracks to the service every single day.

Why is there so much AI slop? Because it pays. Generating a thousand images, articles, product descriptions, and glowing fake reviews costs close to nothing now, and the platforms that distribute it all chase engagement, not quality. They reward the moment your thumb stops moving. When a fake driftwood elk out-earns a real photographer, somebody makes ten thousand more elk.

One distinction matters here, because I use these tools every day and I’m not going to pretend otherwise. AI-generated does not automatically mean slop. When I use AI to transcribe an interview, organize research, or wrestle a headline into shape, that’s a tool doing tool things, with a journalist still answering for every word. When somebody publishes 400 interchangeable articles without checking whether a single fact in them holds up, that’s slop — and it’s competing for your attention with work real reporters spent weeks on.

How can I tell when something’s AI slop? Check hands, teeth, jewelry, and any text in the background of a photo- details that AI still gets sloppy. Distrust any image whose only job is to make you feel something instantly, especially outrage or awe. See whether a “news” story names a reporter and a real outlet, and whether anybody else has the story. On product reviews, hunt for specific complaints, because real customers gripe about shipping and sizing while slop gushes. And when a video shows something remarkable that no other camera in the world managed to catch, you have your answer.

The real damage isn’t the fake silver fox dudes

My Nextdoor feed is a war zone right now. Not about crime or parking, but about what’s real. Somebody posts a video, somebody else says it’s AI, a third person calls the local paper a propaganda outlet, and it around it goes. These are neighbors who will believe an anonymous rant with no source before they’ll believe a reporter with a name, an editor, and a phone number. It’s absolutely insane.

It all makes Shrimp Jesus look like child’s play. Slop doesn’t just trick people into believing fake things.

It trains them to disbelieve everything, which is enormously useful to anyone who would rather you not trust the actual news.

Once you’ve been fooled a few times, “that’s probably AI” turns into a way to wave off any inconvenient fact, including the true ones.

The numbers back up what my neighbors are doing. The Reuters Institute’s 2026 Digital News Report found that social media and video networks have, for the first time, overtaken both television and news sites as where people get their news. Pew found that adults under 30 now trust information from national news organizations (51%) at almost exactly the same rate they trust information from social media (50%). Fifty-fifty, between a newsroom and a stranger’s feed.

Meanwhile, the slop keeps spreading. Pew collected a random sample of 10,000 webpages in July and found that one in ten showed signs of being written or substantially edited by AI.

So what can you actually do?

Slow down before you share. The whole business model depends on you reacting in under two seconds, so the pause is the defense.

Check the account, not just the post. A page that posts twelve emotional images an hour isn’t a person.

Use a detector as a gut-check. Pangram is the most accurate one available right now, and its new image detector is free to try. Treat the results as a strong hint, never as proof — these tools always trail the newest AI models by a few months.

Go straight to the source. If a story matters to you, find it on a site with a named reporter and a correction policy. That’s the whole difference.

Share

Now, about that maxxing

It started with looksmaxxing, which crawled out of incel message boards in the 2010s to describe men swapping tips on maximizing their “sexual market value.” Looksmaxxing is still the monster of the family at around 300,000 searches a month. For many teenage boys, it starts with skincare and the gym, but doesn’t stop there. Ask a 15-year-old what “bone smashing” is — hitting yourself in the face with objects to sharpen your jawline — and brace yourself for how casually he explains it. Searches hit 1.9 million a day in March. TikTok finally banned them in April.

Then the suffix got loose. Now there’s sleepmaxxing (mouth tape, magnesium, a bedtime routine longer than most flights), fibermaxxing (chia seeds in things chia seeds have no business being in), proteinmaxxing, moneymaxxing, aura-maxxing, and whatever my daughter invented on the spot for a woman standing in front of her in dog jammies. Mentions have nearly doubled since January, per Meltwater, and brands have gleefully piled on.

It’s mostly just harmless slang now, but like so many things stemming from internet culture these days, there’s a darker underbelly here too: the idea that every part of you is a number that could be higher or a trait that could be better.

You’re not sleeping; you’re optimizing your sleep. You’re not having dinner; you’re hitting your protein. You’re not seeing friends; you’re maintaining your social health. It’s exhausting. It’s also how a joke about chia seeds ends up on the same family tree as a kid hitting himself in the face.

For the record, my doggy sleepwear wasn’t optimized for anything. I just love dogs.

And then there’s PervTech…

Yes, it means exactly what you think it means, and no, I didn’t coin it. The internet did.

Fellow tech journalist Joanna Stern also wrote a perfect piece about it, which you can find here.

Her short version: “perv glasses” stuck to Meta’s smart glasses because people — mostly men — keep wearing them while hitting on women, sometimes hiding the recording light or the fact that they’re recording at all. (Ew.)

Then this week’s leak of camera-equipped AirPods handed the internet a sequel: PervPods.

The name is funny. The reason it landed so quickly isn’t.

We’re all getting pretty dang tired of being recorded. Go to a gym, an airport, a restaurant, or your kid’s soccer game, and somebody is filming themselves, which means there’s a decent chance they’re filming you, too. You’ve become an extra in a stranger’s content without ever agreeing to be in it. Then there are Flock Safety cameras logging license plates as we drive around town, doorbell cameras catching everyone who trips on a sidewalk, and security cameras just about everywhere else.

Now imagine putting cameras inside something as ordinary and inconspicuous as AirPods.

Before anyone panics, Apple hasn’t announced these things and won’t comment. What we actually have is code Apple apparently left inside a Mac software update last week, which the folks at AppleInsider and MacRumors have been happily tearing apart ever since. Bless them.

So far, the clues point to something much less creepy than the name PervPods suggests. The cameras appear to be tiny — around one megapixel — and designed to capture still images for Siri, not record video for your photo library. The idea is that Siri could see what you’re looking at and answer questions about it without you pulling out your phone. The same code included references to an indicator light, suggesting Apple is at least thinking about how everyone else will know when those cameras are in use.

I’m also less worried about Apple than I am about what happens next. Apple has built privacy into its brand, and putting a secret recording device in millions of ears would be a spectacular way to blow that up.

What worries me are the companies racing to beat Apple to the shelf. Every time Apple signals a new product category, a swarm of manufacturers follows with cheaper versions, and those little details that cost money — privacy protections, security updates, clear data policies, even a visible recording light — can disappear pretty quickly.

That’s how we could wind up with cheap glasses and earbuds made by companies you’ve never heard of, pointed at everyone around us, with no obvious way to know when they’re recording or where whatever they capture winds up.

I spent much of the past year reporting on Flock and talking with people who learned after the fact that cameras in their communities had been recording their movements. They weren’t angry because cameras exist (to be fair, some of them are). They were angry because nobody asked them (and police misuse them to be creeps themselves).

That feels like the lesson worth remembering before we start putting cameras in the next thing we stick on our faces.

So, what am I taking away from all of this?

Mostly that I’m getting tired of being asked to keep up with technology that is moving faster than the rules around it.

AI can now make a convincing photo of something that never happened, so we have to get better at figuring out what’s real. The internet has turned sleeping, eating, and exercising into things we’re supposed to “maxx.” And the next generation of gadgets may be able to see what we see without anyone standing nearby necessarily realizing there’s a camera involved.

I’m not anti-AI, anti-gadget, or anti-progress. Good grief, I’ve spent a good chunk of my career trying new technology and telling people which parts are actually worth using. But “we can” is not the same thing as “we should,” and convenience is not the only thing worth designing for.

Sometimes “friction” (another buzzword) is useful. Pulling out your phone to take a picture tells the people around you that you’re taking a picture. A recording light gives someone a chance to move away. Asking permission may slow things down by five seconds, but those five seconds belong to the other person, not the technology company.

And maybe that’s the part of our lives we shouldn’t be maxing away.

What do you think?

This is your chance to ask questions, tell me what I’m getting wrong (and right), and what you want me to talk, show, or do more or less of here.

Send me your best slop, too. What’s the strangest fake thing that fooled you, or nearly did? I’m collecting them.

And I’d like to formally announce that I am done maxxing, unless somebody has a truly excellent pajama recommendation. Dogs optional.

Leave a comment

I’m an Emmy Award-winning tech journalist and host of Techish with Jennifer Jolly. Reach me at jj@techish.com.

She’s a badass former civil rights and consumer protection attorney (who went to law school for a time with Ted Bundy!). She also came of age professionally when law school — and the legal profession — was still very much a man’s world. She and the remarkable women of her generation (many of whom I had the incredible pleasure of getting to know better this past weekend) walked into rooms not built for them, earned their seats anyway, and kicked open doors the rest of us have been walking through ever since. I get to do the work I do today because women like Sal dared to do it first.

It was a hell of a privilege to celebrate her.

Between catching up on laundry and making sure my husband and animals remember who I am, I’m packing up again to head back out to LA and the San Francisco Bay Area. I’ll be doing more live TV, telling the consumer tech world about the project that we’re rolling out this fall (I’M SO EXCITED), and spending quality time with my daughter, brother, and sister along the way. I can’t believe how fast time is flying these days, and it’s always a privilege and honor to spend some of it with you!

In the meantime, be sure to read my latest story in USA Today on ways to save money on back-to-school shopping with AI.

Also, check out my segment on The Today Show earlier this month, too. I shared the hour with the cast of Ted Lasso, several of the Real Housewives, and Bravo host and absolute icon, Andy Cohen. I also got to take my best friend, Rona, with me. We’ve been friends for more than 40-years, and sharing the incredibly good vibes from the Today family with her was just as magical as you might imagine. Just two gals from Kenai…

My next Today Show segment is slated for September 16th, and the focus is: Home Healthcare Tech. The segment focuses on people who want to take better care of their health at home. If you have a gadget or tech tool to pitch me, or want to tell me about a product that’s helped you live a more “wired well” life, message me right away. We often decide on the products to showcase 2-3 weeks in advance!

And now, to the big tech+life news I’m covering this week:

1. ChatGPT just made a separate version for teenagers

This time last year, the Associated Press, among many other well-researched and respected news outlets, unveiled an alarming reality of the race into AI. ChatGPT and other AI “assistants,” told kids how to hurt themselves.

“ChatGPT will tell 13-year-olds how to get drunk and high, instruct them on how to conceal eating disorders and even compose a heartbreaking suicide letter to their parents if asked, according to new research from a watchdog group.”

— Associated Press, August 8, 2025

Today, OpenAI rolled out ChatGPT for Teens, a new version for 13- to 17-year-olds, in response to growing concerns and pressure over what happens when kids turn to a chatbot for everything from homework help to some of the darkest moments of their lives.

What are the safety features for teens in ChatGPT?

OpenAI says it’s added stronger safety protections and more emphasis on helping kids learn instead of simply handing over answers.

The big structural change is that nobody has to opt in. Sign up as a 13- to 17-year-old, and you land in the teen experience automatically. Lie about your birthday, and you may land there anyway: OpenAI’s age-prediction system reads more than 2,000 signals — what you ask about, how long the account has existed, what hours you’re awake — to guess who’s under 18.

A company spokesman described those signals as “minimally privacy invasive.” Adults caught in the net can appeal with a selfie. When the model isn’t sure, it defaults to the stricter setting.

Inside teen mode you’ll see:

• Tighter limits on self-harm, eating disorders, violence, and sexual or graphic content

• No romantic language, and the model can’t claim feelings or imply it’s conscious

• A nudge to take a break after 90 minutes

• Quiet Hours, set by a linked parent, when ChatGPT simply won’t open

• Study Hours, which force on Study Mode — the tutoring layer that asks guiding questions instead of producing answers

• “Responsible homework reminders” if the model things a teen is trying to cheat on their homework.

A reporter already fooled it…

A Business Insider reporter built a fake teen account this morning and asked for an 800-word essay on The Crucible, due tomorrow, in his voice.

"I write like an A+ 15-year-old," he told the chatbot. "My teachers love me, and I'm top of class for English."

Ultimately, with a few simple workarounds, it handed over 776 words, ready to submit.

The algebra homework was even easier to get done for him.

Do teen tech restrictions work?

So far, most teen “safety” features have been a fat lot of too little, too late. Most seem like PR plays versus truly well-designed tools to help parents take better care of their connected kids.

Make no mistake, tech companies could design 99% fail-proof features to keep kids safer, but that goes against their business models to hook us young and keep us coming for more.

I’ve seen all of this before: Restrictive defaults for minors, break reminders, crisis interventions, a parent dashboard. Heck, Instagram shipped that playbook in 2024, TikTok has its own, and Meta is in court today defending itself against four states seeking up to $1.4 trillion over youth mental health. The pattern is consistent — make the product feel safer without making anyone use it less, then hand parents a control panel and reclassify whatever’s left as a supervision problem.

Florida sued OpenAI in June for misrepresenting how safe its platform is for minors. Families are suing over children who were harmed or died. The FTC is asking questions. Teen mode arrives now, in that weather, roughly four years and 900 million weekly users after ChatGPT shipped without it.

Make no mistake: I’d rather this existed than not. But a company that has spent years telling us these systems are too complex to fully steer shouldn’t act surprised when a fake 15-year-old finds the cracks the very day of the launch.

What do you think?

2. Speaking of Lawsuits…Meta is Back on Trial

Meta is back in court over kids — less than two weeks after being ordered to pay nearly $1 billion in another child-safety case.

Another massive case against Meta kicked off in Oakland today, with 29 states accusing the company of designing Instagram and Facebook to hook young people — while misleading families about the risks.

Meta denies the allegations and points to the teen-safety features it has added over the years. Mark Zuckerberg and Instagram chief Adam Mosseri are expected to testify during the six-week trial.

Ultimately, the states are asking for tougher age restrictions and an end to infinite scroll for kids. Among the requests: tougher age restrictions and an end to infinite scroll for kids.

You know “infinite scroll” even if you’ve never called it that. It’s why Instagram and TikTok never reach a stopping point. There’s no last page. No welp, I guess I’m done here. Flick your thumb and another video appears. And another. And all of a sudden three hours have gone by. That design choice matters because stopping requires a decision, while continuing doesn’t.

Now a federal court could decide whether companies should be allowed to build that for children — and the answer won’t stop at Instagram. Every app running the same trick faces an uncomfortable question.

What parents should watch: skip the courtroom scorekeeping. Watch which features get challenged — infinite scroll, age checks, notifications, messaging, recommendation algorithms. Those are what SHOULD change…

3. Apple may be putting cameras in your AirPods

A video posted today on X by @aaronp613 appears to show unreleased AirPods equipped with cameras. They look chunkier than today’s AirPods Pro, with longer stems that could make room for the extra hardware.

Before anyone starts imagining people secretly taking pictures with their ears, that doesn’t appear to be the point. The cameras could give Siri and Visual Intelligence eyes.

In the demo, someone holds up a book, and Siri identifies information about it.

That could come in incredibly handy when:

• You’re walking through an airport with your hands full: Which way is Gate C12

• You’re looking at a plant: What is this thing, and will it kill my dog?

• You’re traveling in a foreign country: What does that sign say?

• You’re staring at some mysterious cable in the junk drawer: What the hell does this plug into?

You wouldn’t have to pull out your phone and point a camera at anything. Your AirPods would already be looking where you’re looking.

And… that’s where this gets weird (creepy/rapey Meta glasses anyone?):

We’re still arguing about whether someone wearing camera glasses is recording us. A camera tucked into something that looks like a perfectly ordinary pair of AirPods could be much harder for anyone nearby to recognize.

Apple hasn’t announced these AirPods. We don’t have a price or launch date, and the final product could change or never ship. Apple tends to be the most cautious and least creepy of all the big tech companies these days, so I have faith they’ll do the right-ish thing.

But this one is harder to dismiss as another Apple rumor because the apparent demo turned up inside Apple’s own software. It’s also one of the most-read stories on The Verge today and is bouncing around Apple forums and social feeds.

Remember when headphones just let you listen to music???

This Flock Interview Hasn’t Aged Well…

A year ago, 108,000 people watched my interview with Flock Safety CEO Garrett Langley on YouTube alone. Even more read the story I wrote about the increasingly controversial public-safety tech company for USA Today.

A lot has happened since then. And let’s just say: neither story has aged particularly well.

I’m working on an update to that story next…

What are you reading, watching, contemplating buying? What do you want me to use my 30+ years of journalism know-how to dive into next? Be sure to let me know. You’re the reason I’m here.

Now we’re buying laptops, protecting kids’ hearing, figuring out how younger children can call home without getting the whole internet, and trying to keep the family calendar from living exclusively inside Mom’s brain.

So we went deep this year—testing the big-ticket tech and the small, clever gadgets that can make school mornings, dorm rooms, and everything in between a little easier.

WATCH THE TODAY SHOW SEGMENT HERE

That includes a foldable phone that opens into a tablet, laptops for three very different budgets, a working record player smaller than a shoebox, and a phone case that changes pictures.

Not every family needs all of it. But every pick here solves a real problem, earns its space, or is just too darn fun to ignore.

XTREM Instant Custom Case

The XTREM case puts a small e-ink screen on the back of your iPhone, so you can change the picture whenever you feel like it. First-day-of-school photo today, the dog tomorrow, your kid’s artwork on Friday.

Open the app, choose a photo or design and tap once. A few seconds later, it appears on the case.

Because the display uses e-ink—the same basic idea as a Kindle—it does not glow, need a battery, or require charging. Once the picture appears, it stays there until you replace it, even when the phone dies.

You can also turn photos into illustrated characters and other designs inside the app. The newest sharing feature lets people link their XTREM cases and send pictures directly to one another, although the recipient controls what actually appears. That gives it a bit of the long-distance-grandparent magic of a digital photo frame, only on the back of a phone.

The case supports MagSafe charging and accessories and currently fits several iPhone 14, 15, 16, and 17 models.

Price: Around $74 *Exclusive 20% off discount code for Techish:TECHISH20*
Shop: HelloXTREM.com

Cricut Joy Xtra

The Cricut Joy Xtra is a cutting machine about the size of a loaf of bread. It turns designs into stickers, labels, decals and iron-on lettering.

Create a design in Cricut’s free phone app, send it to the machine and let it do the cutting. A few minutes later, a student’s name is on the water bottle in lettering that can survive the dishwasher. The backpack has an iron-on label, so it might actually come home with the correct child. The supply bins, lunch containers, and gym shoes can sport your student’s unique sense of style and personality.

Think you’re not crafty? There’s an AI image generator inside the app, so all you have to do is describe the design you want, and the app creates it for you. That’s awesome for someone like me, who loves the idea of making things but breaks into a light sweat when someone says “craft night.”

Price: $179 for the machine or $219 with starter materials
Shop: Amazon or Cricut.com

iClever Q950 Kids Headphones

Anyone who has handed a child headphones knows exactly what happens next: volume up, volume up, volume up—and apparently nobody can hear you calling their name from six inches away.

The iClever Q950 headphones cap the sound at 80 decibels. Your child can mash the volume button all they want; the headphones will not blast past that limit.

That is the part parents care about. The part kids care about is the active noise cancellation, which cuts up to 35 decibels of airplane, classroom and back-seat racket. That can also reduce the temptation to turn the music up just to drown everything else out.

Transparency mode lets outside sound back in when they need to hear a teacher, parent or boarding announcement.

These are also built for actual children, not children who sit motionless in catalog photos. The headband is reinforced, the hinges are metal and iClever says the design survived 3,000 twisting tests. Battery life reaches up to 60 hours with noise cancellation off or 35 hours with it on. Two pairs can also share audio from one device.

Price: 54.98 (at time of publication on Amazon, $70 on iClever site)
Shop: iClever and Amazon

Camp Snap 2

This is for anyone who takes 47 pictures of the same moment—and somehow misses the moment.

Camp Snap 2 is a rechargeable digital point-and-shoot camera with no screen. You aim, click, and keep moving. No checking the photo. No retaking it twelve times. No wandering from the camera app into texts, TikTok, and forty-five minutes you will never get back.

It brings back the fun of a disposable camera, minus the film, developing costs and single-use plastic. When you are ready to see what you captured, connect the camera by USB-C to a compatible phone, tablet or computer and transfer the pictures.

It comes with six filters built right into the camera, including vintage, analog and black-and-white looks. The included memory card holds roughly 2,000 photos, and the battery lasts for about 500 shots per charge.

Kids come home with pictures from their day instead of spending the day staring at pictures. College students get a fun party-and-travel camera that does not risk their $1,000 phone. And the see-through jelly colors are straight out of a very good 1990s fever dream.

Price: $74.95
Shop: CampSnapPhoto.com

TinyVinyl Mini Record Player

Nobody needs this because it is the world’s finest way to listen to music. You want it because it turns your favorite artist’s songs into tiny, playable collectibles—and the whole setup is absurdly cute.

TinyVinyl records are real four-inch records, traditionally pressed and played at 33⅓ RPM. Each one carries two songs, one per side, and every pressing has a number that helps verify the release.

The new player is roughly ten inches wide, comes in a little black leatherette case and has built-in speakers. You lift the arm, drop the needle and flip the record yourself. It can also send the sound to a Bluetooth speaker or headphones, which is probably the better move when you care more about the music than the miniature aesthetic.

A normal vinyl setup can take over a dorm-room desk and devour several hundred dollars. This player costs $50, and the records are $15 each. TinyVinyl also sells little display frames and a miniature crate, because even tiny record collections apparently develop storage problems.

One caveat: the TinyVinyl player only plays TinyVinyl records. The records themselves can work on many standard manual turntables, but not every automatic model can reach their grooves.

Price: $49.99 for the player; records are $14.99 each
Shop: Target exclusive

Oster GoMAX Portable Blender

The blender is at home. You are not. Somehow, this problem costs nine dollars every time it happens.

The Oster GoMAX turns the blender cup into the cup you drink from. Fill the 20-ounce container with frozen fruit, greens, protein powder or whatever you claim makes spinach taste good, then press the button.

If an ingredient jams the blade, Oster’s anti-jam system automatically reverses it rather than making you stop and shake the whole blender like a maraca.

When the smoothie is ready, pop open the leak-resistant lid and drink from the same container. No outlet, no second cup and no giant blender pitcher sitting in the sink until someone else finally gives in and washes it.

It charges by USB-C, and Oster says one charge can power up to 15 smoothies. The detachable parts can also go in the dishwasher.

This makes sense in a dorm, at the gym or during the morning school scramble: blend it, grab it and go.

Price: $44.60 at time of publication, normally around $70 for the blue $
Shop: Amazon

Samsung Galaxy Z Fold8 Ultra

Folding phones are all the rage, and this is hands-down the best one I’ve ever seen. Gone is the center crease and feeling that a fold-in-half phone is more novel than necessary.

That said, this is a really high-end phone for the college or grad student who does everything, everywhere, all at once. This just-unveiled flagship foldable from Samsung opens from a regular smartphone into a mini-eight-inch tablet.

That makes for some serious multitasking for the student taking notes, researching, answering messages, and editing a video—possibly all at the same time.

Its new Snapdragon processor keeps everything moving quickly, while the bigger screen gives students enough room to run several apps without constantly flipping back and forth.

The camera is just as ambitious: The 200-megapixel main camera captures enough detail to crop way into a photo without turning it into a blurry mess. Nightography helps with dim gyms, concerts and late-night campus life, and it can shoot video in 8K.

This is for that student content creator or filmmaker. On a good day, it can replace their camera, tablet and even their laptop—and it has enough power to carry them through graduation and beyond.

Starting price: $2,100, with trade-in and financing offers available
Availability: Preorder now; available later in August
Shop: Samsung.com

Pinwheel Home

I am a huge advocate for NOT letting kids get smartphones until high school, and for using a series of “training wheel” tech gadgets to help them grow and learn how to use tech healthfully without all the caveats that come with too much too soon.

A lot of parents want their younger kids to be able to call home without handing them texting, social media, a browser and the entire internet.

My answer lately is a landline. Just not the beige one that used to hang in your grandmother’s kitchen.

Pinwheel Home is a small, screen-free Wi-Fi phone that lets kids call only the people their parents approve. You control who can ring the phone and what hours calls are allowed. Unknown numbers do not get through, and there are no apps, texts, social feeds, or a browser to wander into.

It takes about ten minutes to set up, uses Wi-Fi and does not need a phone jack or an installer. Calls between Pinwheel devices and access to 911 are included. Calling regular outside numbers requires a monthly plan starting at around $7.

This gives younger kids a way to call a parent, grandparent or approved friend without turning a simple phone call into their first smartphone.

Price: Starting at $68; outside calling plans start at about $6.99 a month
Shop: Pinwheel.com

Greenlight Family Hub

Every family has a calendar. The problem is that it usually lives on one parent’s phone, where nobody else looks at it.

The Greenlight Family Hub puts the whole operation on a shared screen in the kitchen: calendars, chores, grocery lists, reminders, family photos and all the places everyone needs to be this week.

It comes in a 10-inch version for $199 and a 15-inch model for $299. You can sync your existing calendars rather than rebuilding the family schedule from scratch, then give kids their own chores and routines. Greenlight can also tie chores to allowance and rewards, which means you may finally stop answering, “Did you send me my money?” six times a day.

The paid Hub+ plan adds an AI family assistant, real-time family location and additional safety tools. Unlike sticking an old tablet on the counter, there is no open app store or browser for kids to disappear into.

Is another family screen absolutely essential? No. But having the schedule somewhere everyone can see it may prevent more household chaos than another beautifully color-coded calendar buried inside Mom’s phone.

Price: $199 for the 10-inch model; $299 for the 15-inch model
Service: Free basic version; Hub+ starts at about $9.99 a month
Shop: Greenlight.com

Dell XPS 13 Laptop

The new Dell XPS 13 weighs just 2.2 pounds. That is less than some of the textbooks you’ll cram into a backpack with it.

It is the thinnest, lightest laptop Dell has ever made, but doesn’t feel the least bit stripped down. It runs on Intel’s latest Core Series 3 processors and has a sharp 2.5K touchscreen with a 120-hertz refresh rate, which makes scrolling, video, and everyday work noticeably smoother.

The backlit keyboard helps during late-night library sessions. Quad speakers make dorm-room movies sound better. And with up to 17 hours of streaming, it should last from the first lecture of the morning through the final “I’ll just finish this one thing” study session at night.

The bigger story here is that students no longer have to lug around a bulky, expensive machine to get strong performance and a great screen.

Student price: Starting at just under $600 for students ages 16 and older, for a limited time
Shop: Dell.com

Lenovo IdeaPad Slim 3i

Most students need a laptop that runs everything they use for school, survives getting shoved into a backpack, and does not require a second mortgage.

The Lenovo IdeaPad Slim 3i has a roomy 15.3-inch 2K touchscreen, 16GB of memory and 512GB of storage—the combination I would look for in a new school laptop right now. There is enough power for research, writing, video calls and the seventeen tabs your teenager insists are all essential.

It weighs about 3.5 pounds, has the ports students still need, and is tested to military durability standards. That does not mean you should throw it down the stairs, but it gives me more confidence than a laptop that feels like it might crack under the weight of a math textbook.

Lenovo says it can last up to 12 hours while running work apps and more than 18 hours while streaming video. Real-life battery life will vary, but it should get most students through a school day without turning outlet hunting into an extracurricular activity.

At around $850 on sale, this is the sensible sweet spot: a strong screen, enough memory and storage to last, and no money wasted on power most students will never use.

Price: Around $850 on sale
Shop: Lenovo.com

ASUS Zenbook A16

The ASUS Zenbook A16 gives you a giant 16-inch screen in a laptop that weighs only about 2.6 pounds. That is kind of ridiculous—in a good way.

Its 3K OLED touchscreen is gorgeous for schoolwork, streaming, photo editing, and every other excuse your student will use to keep it open instead of sleeping. It also packs 48GB of memory and 1TB of storage, which is much more than the average student needs today but gives them plenty of room to grow into it.

The body uses ASUS’s ceramic-and-aluminum material, so it feels premium without weighing down a backpack. The Snapdragon X2 Elite Extreme processor also handles some AI tools directly on the laptop, which can make them faster and keep more of that work on the device instead of constantly sending it to the cloud.

This is not the budget pick. It is for the college student doing video, design, engineering work, or serious multitasking—or the parent who plans to reclaim it the moment the kid graduates.

One thing to check before buying: it runs Windows on an Arm-based processor, so students who need specialized school software should confirm that their required programs are compatible.

Price: Around $1,700
Shop: Best Buy and ASUS.com

SANDISK Extreme Portable SSD

Students now carry years of schoolwork, photos, videos, and creative projects on one laptop. That’s convenient right up until the computer runs out of space—or something disappears the night before it is due.

The SANDISK Extreme Portable SSD gives students and creators up to 4 terabytes of extra storage without paying to upgrade the entire computer.

It can transfer and back up giant photo and video files in seconds instead of leaving you staring at a progress bar. Plug it in, drag the files over, and toss it into a backpack.

It is especially useful for photography students, video editors, gamers, or anyone whose laptop has started sending increasingly desperate “storage almost full” warnings.

One important rule: use this as a second copy of important work, not the only copy. Keep school projects backed up in at least two places—such as the drive and the cloud.

Price: Starting at $239.99 for 1TB (price at time of publication)
Shop: Sandisk.com

woom Neon Bikes

woom (yes, it’s all lower-case) is one of the biggest names in premium kids’ bikes—for a good reason. The company designs them for children from the ground up, rather than shrinking an adult bike and calling it kid-size.

That means lighter frames, smaller grips, seats shaped for little bodies, and brake levers that small hands can actually reach. Some woom bikes weigh up to 40 percent less than conventional kids’ models.

That weight difference is huge when the rider weighs only 40 or 50 pounds. A lighter bike is easier to balance, steer, stop, and pick up after the inevitable driveway wipeout. Kids spend less time wrestling with the bike and more time actually riding it.

woom makes bikes for every stage, from toddlers learning to balance to preteens ready to roam the neighborhood. For back-to-school season, the woom GO comes in neon lime, and the EXPLORE comes in neon pink.

Both currently include a free pair of matching Lusso Cloud shoes while supplies last.

Price: Starting at $269
Shop: woom.com

IKEA KALLSUP Bluetooth Speaker

IKEA made a Bluetooth speaker for ten bucks, and it sounds far better than a tiny plastic cube has any business sounding.

The KALLSUP is about the size of a Rubik’s Cube, so it can sit on a dorm-room desk, tuck onto a nightstand or disappear into a backpack. Its rechargeable battery lasts up to nine hours at half volume.

You can also connect several KALLSUP speakers and play the same music through all of them. Put one in the kitchen, one in the bedroom or one in every kid’s room so they can stop stealing yours.

This is still a $10 speaker. It will not rattle the windows, fill a giant party or replace a premium sound system. IKEA also does not list a waterproof rating, so keep it away from the pool and the overly enthusiastic side of the shower.

It comes in pink, white and an aggressively cheerful yellow-green. One small bit of fine print: the USB-C charging cable and wall adapter are sold separately.

Price: $9.99
Shop: Ikea

Our production or marketing team may work with some companies featured in this article. Those relationships remain separate from our editorial coverage and reviews and never apply to anything you see Jennifer demonstrate or talk about on the Today Show.

READ NEXT:

Some links in this post are affiliate links. If you purchase through them, we may earn a small commission at no additional cost to you. That support helps us keep producing practical, independent technology coverage.

Prices and promotions are accurate at the time of publication and may change.

I’ve been deep in annual student/parent gadget and deal testing for a few weeks now, including how to use AI safely to find bargains, how to match your family with the perfect gadgets before spending a dime, and how to organize the chaos of it all with a few prompts it took me about 20 hours to perfect so you can just copy and paste. #yourewelcome

I’ll be in New York all next week for a bunch of fun work (and a litte play), including a broadcast media tour that brings me to more than 300 local news stations over the next few weeks, so keep an eye out for me on your local network(s). I’m also back at the Today Show on August 5th showing off some great family gadgets in the 9 a.m. hour. Check back here on Wednesday to score Techish-subscriber-only deals on what we show off, plus a few more gadgets that didn’t make the on-air cut but did make my top picks of 2026.

IN TODAY’S NEWSLETTER

• 🍏 Apple will rent you an iPhone, but does it really save you any money?

• 🎒 The ChatGPT trick that saves you big bucks on back-to-school shopping

• ☎️ What I love this week: the $68 kid phone that’s ending the “when can I get a phone” fight

• 🛍️ Facebook’s new app is for selling, not shopping

• 🫁 Smoke season’s rough, here’s the air filter that won the lab test.

🍏 THE BIG STORY: Apple Will Rent You an iPhone Now

Last week, I told you to buy the school laptop now, before prices climb again. Then on Tuesday, Apple launched a program built on a different idea — that maybe you should lease versus buy.

They call it Apple Upgrade: You pick a device, pay every month, and at the end you hand that gadget back and start over. Klarna runs the lease and does a soft credit check, which they say won’t hurt your score.

Is this a good deal for you? If you can afford to buy the thing outright, do that. It’s that simple. I’ve yet to find a case where a “rent” or “lease” deal saves you money in the long run. Not with cars, and not with Apple gadgets.

The lease doesn’t save you money, it just moves it.

At the end of your lease, you can hand the device back, or you can keep it by paying something Apple calls the Purchase Option Fee. That fee is whatever’s left of the retail price.

Here’s how all of that looks in real life:

iPhone 17 Pro. It costs $1,100. Lease it at $32 a month for two years and you’ll hand over $768. Want to keep it? That’s another $332. Or you hand it back and start a new lease on the newest model.

Take that second option twice and you’ve paid $1,536 over four years and you own nothing. Buy the phone once and keep it four years and you paid $1,100.

Buying costs $436 less, and it’s yours.

Apple Watch Series 11. It costs $400. Lease it at $12 a month for two years and you’ll hand over $288. Keeping it costs another $112.

Lease it through two cycles and you’ve paid $576 over four years with nothing on your wrist that belongs to you. Buy it once and keep it four years and you paid $400.

Buying costs $176 less.

MacBook Air. It costs $1,300. Lease it at $25 a month for three years and you’ll hand over $900. Keeping it costs another $400.

Lease it through two cycles and you’ve paid $1,800 over six years. Buy it once and keep it six years and you paid $1,300.

Buying costs $500 less, and plenty of MacBooks are still going strong at six.

So Apple isn’t giving you a deal. They’re giving you a payment schedule with a big bill parked at the end of it, maybe forever.

What can I do about it?

Start saving for that last bill the day you sign.

On the iPhone, put $14 a month in a jar. On the watch, $5. On the MacBook Air, $11. When your lease ends, you’ll have the money to keep your device, and you’ll have paid exactly what it costs.

Now look at what that adds up to. Your $32 iPhone lease plus $14 in savings comes to $46 a month. Apple’s 0% financing on that same phone? Also $46 a month. The watch works out to $17 either way. The MacBook Air, $36 either way.

There’s no free lunch in here. There are only different ways to schedule the same bill.

What happens if you can't pay that last bill

The tricky part happens at the end of year two. You get to the end of the lease, you don’t have $332 sitting around, and the easiest thing in the world is to sign a new lease on a new phone. Now you’re two more years in, and you don’t own a gadget that you can trade or sell for real dollars and cents. That cycle can run forever, which might be the point.

No judgment from here, as someone who will likely pay a mortgage for the rest of my life and only just recently paid a car off for the first time ever, I 100% get it. The struggle is real. Still, go in eyes wide open and maybe hang on to your older handset another year or two if you can.

👉 [LINK: Apple Upgrade — read the fine print]

🎒THE FIX: Let ChatGPT Cut Your Back-to-School Bill

The average family is looking at close to $600 per kid this fall. This is the fastest way I’ve found to knock that number down.

Take a screenshot of the supply list. In ChatGPT, tap the plus sign, choose Shopping Research, and ask it to price the list two ways: cheapest total across a bunch of stores, and cheapest total from one store near you that does pickup. Ask for a table with links.

Then photograph the drawer where last year’s supplies went to die, upload that, and tell it to subtract everything you already own before it totals anything up. (Every family I know owns six glue sticks and buys seven more.)

Two things I learned the hard way: Pull the laptop out and shop for it separately, because it bogs down the whole search. Also, confirm prices on the store’s own site before you buy, since inventory moves fast and some retailers block automated access. If you’re shopping Amazon, use a service like CamelCamelCamel to make sure that “deal” is really the lowest price.

When I ran a simulation with a high schooler’s list through it, it saved me close to $300 compared to just heading to Amazon or my local Target to get it all.

👉 [LINK: the full back-to-school guide, including the privacy settings to flip first]

❤️ WHAT I LOVE THIS WEEK: Pinwheel’s Home Phone

I’ve had three of these landline-style kid phones in my house over the past few months, all sent to me for testing, and this is the one I’m telling my friends to buy right now.

Pinwheel Home starts at $68 and runs over Wi-Fi, so there’s no phone jack, no installer, and no appointment window where you sit around waiting on a guy named Bob. Mine was working in under ten minutes, which is not a sentence I get to write about kids’ tech very often.

You approve the contacts and the calling hours from an app. Numbers you haven’t approved don’t ring through at all. Calls to other Pinwheel devices and 911 are included. Calling regular numbers needs a plan, just under $7 a month for five numbers or just under $10 for unlimited.

After all that leasing math up top (my brain is still a little bogged down with all of that), there’s something very nice about a gadget you buy once, for sixty-eight dollars, and then own.

I gave my friends a Tin Can phone last Christmas. I ordered it in November; I think it arrived some time in March. I checked in with Scott — the dad, and a former Techish colleague — about how it’s held up:

“We still have it, one phone has decent reception, the other is really buggy and doesn’t have good sound quality. I believe we stopped paying the $10 a month [subscription] because it’s just not worth it, kids maybe use it once a month. It’s also just hard to know which friends have a phone. Would be cool if there was an app that went through your phone contacts to highlight who might also have one.”

Pinwheel, take note. That last part is a real issue. These phones should come with five “free” numbers they’re allowed to call, and that’s allowed to call them, then add the monthly bill after that. Also, that last part is an awesome feature to add: let the app search your contacts to see which friends/family also have a Pinwheel device.

📱 QUICK HITS

Facebook’s new app is for selling, not shopping. Meta launched Seller, a standalone app for Marketplace, and a lot of the coverage made it sound like a shopping app. It’s the other side of the transaction. The app is free, iOS only in the US for now, and it pulls in your existing listings and messages automatically. The useful part: photograph something and the AI writes the title, description, category, and suggested price. That’s awesome except when it gets stuff wrong — which is still quite a bit. This is a decent August tool if you’ve got outgrown bikes to move before school starts. 👉 [LINK: Facebook’s app announcement]

Two AI companies admitted their software broke into other people’s computers. OpenAI went first, then Anthropic. Both happened during safety testing, both were disclosed voluntarily, and both were described in language so mild you could skim right past it. The part worth your time is how the break-ins happened, and what the companies are doing about it.

👉 [LINK: NYT: Anthropic Says AI Systems Broke Into Computers…]

🫁 ONE MORE THING: The Best Air Filter for Smoke Right Now

A bunch of you back east wrote to me during the worst of it this month, when smoke off the Canadian fires pushed New York’s air quality index toward 200. It cleared, came back, cleared again, and it will do it again.

I did a huge, months-long test of all the latest air filters back in 2020, when the skies over the Bay Area turned this otherworldly orange, and we thought the world was ending.

CNET just ran fifteen machines through a smoke chamber and timed how long each took to clear 90% of the particles out of the air. Here’s the TLDR version:

Blueair Blue Pure 311i Max, $190. This one has the best balance of everything: It cleared the chamber in 1 minute 28 seconds, doesn’t drain your power, and runs at about 35 decibels on low, which is quieter than the guy eating next to me on the plane right now. How noisy these gadgets are matters more than you might think, because if you have to shut it off at bedtime, it’s not really doing its job.

If smoke specifically is the worry, the Windmill 3-in-1, $320, is the best pick. This is the fastest of the whole group on fine particles, and fine particles are what wildfire smoke is made of.

Expensive doesn’t mean better. The $750 RabbitAir landed mid-pack. A Dyson finished dead last on particle removal.

Ignore the square footage on the box. Coway says its 400S handles 3,100 square feet in about an hour. That’s one "air change," which means the filter has processed a roomful of air one time. That might sound like a lot, but air mixes, so one pass only clears about 63% of the smoke. You want five passes an hour, which means that 3,100-square-foot machine is really a 620-square-foot machine. That same math works on any box in the store.

OMG. Enough with the mathing already!

💬 THE ASK

My producer friend Suzanne photographs the watermelon bin at her grocery store and asks ChatGPT which one to take home. You guys — it totally works!!!

So what’s yours? What’s that weird little AI hack you never see written up anywhere? Hit reply and tell me. I’ll run the best ones and credit you.

Leave a comment

How am I doing? Let me know what you want more of, less of, and what you want me to try, tackle, or talk about next.

XO, Jenn

Companies sometimes send me products to test. Some links may be affiliate links. Prices checked August 1st, 2026.

Techish is free and shows up every Saturday. Consider becoming a paid subscriber for insider tips, tricks, and deals. And to support independent, human-first journalism.

OpenAI rebuilt ChatGPT Voice on new models called GPT-Live, and for the first time, you can talk with it instead of at it. You can also interrupt it mid-answer, trail off, change your mind halfway through a sentence, go quiet to think, and basically talk to it like you talk to another person. Several times I’ve rambled, brain-dumped, or even forgotten that I was talking to it in the first place (read on to see my warning about that last part below).

I didn’t plan to trust, like, or even use this feature as much as I do. I’ve spent decades watching “just talk to it” go sideways. Siri mangles simple requests. Alexa once recorded a family’s private conversation and sent it to someone in their contacts. Contractors reviewing voice-assistant recordings have heard things nobody intended to share—and the list goes on.

Then I started using the new ChatGPT Voice — and I can’t believe how different and good it is. Not that I trust it 100% — because — tech. But I keep telling my friends that it’s exactly what Siri should have been this whole time — a useful, helpful, intelligent voice assistant that gets more things right than wrong.

But there’s still an open microphone involved, so let’s deal with that before we get too attached.

How to find it and turn it on

Open ChatGPT and look at the message box where you normally type. Tap the Voice icon on the right side. Don’t tap the little microphone that records a message and turns it into text; tap the Voice button to open a live conversation.

The first time you use it, ChatGPT asks for permission to use your microphone and lets you choose a voice. I’m using Vale because … British accent … of course. After that, you just start talking.

On a computer, you can use Voice in a regular chat at ChatGPT.com or in the ChatGPT desktop app.

Here’s where it gets confusing: Voice inside ChatGPT’s Work and Codex modes currently works only in the desktop app—not on the web. I discovered that while writing this story, the Voice button appeared in the Mac app but vanished from the exact same conversation in Safari. Same account and computer, just different ChatGPT window. (I’m still not 100% sure how I got it all to work, which happens more often in a tech reviewer’s life than most of us admit…)

To make sure you have the new version, go to Settings → Voice → Live. You should also see a small Live label at the top of the conversation after Voice opens. If Live doesn’t appear in your settings, update the ChatGPT app.

Live works in the ChatGPT iPhone and Android apps, in supported chats at ChatGPT.com and in the ChatGPT desktop app. Voice inside Work and Codex currently requires the desktop app. Live does not support video or screen sharing yet; eligible mobile subscribers can switch to Advanced Voice for those features under Settings → Voice.

Is it always listening?

I asked OpenAI the question that I’ve been trained by a decade of creepy assistants to ask. The answer, on the record: no. There is no wake word. The microphone turns on when you tap the voice button, and not before. In ChatGPT’s own words, it is not an always-on model — if you never tap in, it is not listening to you.

Then I did the very meta thing and asked the microphone about the microphone, grilling it about its own listening habits. It confirmed the tap-to-start part. Then I asked some tougher questions: How do I know when you’re listening? What can you hear? And what happens after I end the session?

Here are a few of the paraphrased responses:

• While a session is open, the mic stays live; it can hear anyone else in the room, and it might even respond to them.

• Its advice — its own, not mine — is to end the session before having a private conversation.

• When I brought up the long history of assistants listening when they shouldn’t, it told me that once you close a session, it’s “not supposed to” keep listening in the background.

I’ll let you decide how comforting “not supposed to” is. The practical rule is simple: you start it with a tap, and you should be the one who ends it. Don’t leave a session running while the rest of your household wanders through.

What happens to your voice?

OpenAI told me that audio clips stick around for up to 30 days so you can resume conversations, and the company says it does not use your actual audio or video clips to train its models unless you choose to share them separately.

The transcript is a different matter. If you have “Improve the model for everyone” turned on, OpenAI may use transcripts of your voice conversations for training, depending on your plan and settings.

Check yours here:

ChatGPT → profile icon → Settings → Data Controls

Turn off Improve the model for everyone if you do not want your conversation transcripts used for training.

If you leave that setting on, make sure “Include your audio recordings” and “Include your video recordings” remain off unless you deliberately want to share those clips.

OpenAI says deleting a chat also schedules its associated audio and video for deletion within 30 days, although the company may retain some data longer for legal, safety, or security reasons. Archiving a chat does not delete it.

I still won’t dictate a bank password, discuss a confidential work deal, or treat a pleasant-sounding chatbot like a licensed therapist (let alone my best friend or spouse). Remember, a voice can feel warm, attentive, and human long before the machine behind it deserves that level of trust.

CHEAT SHEET RECAP: HOW TO TURN ON CHATGPT VOICE

WORKS ON: iPhone, Android, and ChatGPT.com. Both free and paid accounts get it, though free accounts use the lighter GPT-Live mini model.

1. Open the ChatGPT app, or go to ChatGPT.com.

2. Look at the message box where you normally type. On the right, you will see a small Voice icon that looks like a sound wave. Tap it. It’s NOT the microphone icon (which would make sense, but it’s not that). The microphone only turns what you say into typed text. The Voice button opens an actual conversation.

3. The first time you use Voice, ChatGPT asks for microphone permission and lets you choose a voice.

4. When the Voice screen opens, look for the small Live label at the top. That is how you know you have the new GPT-Live upgrade. No Live label? Update the app and check again later. OpenAI is still rolling it out.

5. Start talking. There is no wake word and no button to hold down.

6. When you finish, tap the X to end the session. While that session remains open, the microphone remains live and may hear other people in the room. You start it with a tap, and you should be the one who ends it.

🎯 TRY THIS: PUT VOICE ONE PRESS AWAY

If you have an iPhone 15 Pro or newer, map the Action Button to ChatGPT Voice. Then, the second a thought hits — in the car, on a walk, elbow-deep in dinner — you can press one button and start talking instead of opening your phone, finding the app, and losing the thought somewhere along the way.

Go to:

Settings → Action Button → Controls → ChatGPT → Start Voice Conversation

No Action Button? Add the ChatGPT Voice control to your iPhone’s Control Center or Lock Screen instead.

That tiny change is what turns Voice from a feature you forget exists into something you might actually use.

This is tip No. 1 in my upcoming guide. I’m testing 15 more genius insider tips like these, with the exact words to say, the settings to change, and the tricks that sound smart but fall apart in real life.

Paid subscribers get the whole thing first. Stay tuned!

Now? The list includes laptops, apps, online accounts, family calendars and enough passwords to qualify as an AP course. The average family expects to spend nearly $600 per K–12 student this fall, and a good two-thirds of parents plan to hunt for sales and coupons.

That is where ChatGPT can earn a spot on the school-supply list. I have the paid Plus plan that runs $20 a month, but the free version or the Go plan ($8/month) works okay for this too.

Why ChatGPT specifically? It’s the one AI assistant that lets me tackle the entire back-to-school pileup—from shopping and family schedules to homework and school paperwork—in one place. Other popular AI models like Gemini, Claude, and Copilot can also handle many of these tasks one at a time, but ChatGPT bundled all of the photo uploads, voice conversations, and Temporary Chats for this particular test.

Safety First

Take 60 seconds to make ChatGPT more private before you hand it any personal documents.

• Open Settings, choose Data Controls, and turn off “Improve the model for everyone.”

• Use a Temporary Chat for family details; it stays out of your history and memory and is not used to train ChatGPT, although OpenAI may keep a copy for up to 30 days.

You should still remove names, addresses, ID numbers, account numbers, medical details, and anything else ChatGPT doesn’t need. No online service can promise zero risk, so the safest information is that stuff you never upload. OpenAI explains those controls here.

Once that’s done, it’s time to dig into a few genius AI life hacks and check out some of the gadgets it helped me find.

1. Price the whole list before you spend a dime:

Take a screenshot of the school-supply list. Open ChatGPT and paste this:

Here is my high school student’s [or elementary or college, etc] supply list. I am near ZIP code [ZIP]. Price it two ways: the lowest total across multiple stores and the lowest total from one nearby store offering pickup. Put every item in a table with prices and retailer links.

Snap a quick pic of whatever's left over from last year, because every family owns six glue sticks and five yellow highlighters, yet still buys more. Then upload that shot and ask ChatGPT to remove what you already have before it calculates the total.

OpenAI needs a few minutes to compare stores and build its answer. It was taking too long with the laptop on the list, so I took it out and ran it separately, which is now my go-to advice (see below). If you get an instant generic response, double-check that you selected it. Prices and inventory change quickly, and some retailers block automated access, so confirm both on the store’s site before you buy.

If the school hasn’t sent its back-to-school supply note yet, check TeacherLists.

The laptop deserves its own shopping plan:

One thing to do outside of that main list is the laptop. Most high schoolers and all college-age students are going to need their own device.

Shopping for one is different this year, though, because prices are skyrocketing. AI data centers are competing for some of the same memory and storage used in consumer computers. We’ve seen some popular laptops go up by $200 to $600 in one month, which makes comparison shopping more important than usual.

Two of the best laptops ChatGPT helped me uncover are from Lenovo and Asus, and so far they’re top picks in terms of value for the price:

The Lenovo IdeaPad Slim 3i was on sale for about $850 when I last checked. It has a 15.3-inch 2K touchscreen, 16GB of memory, 512GB of storage, and Intel’s new Core Series 3 processor. Lenovo says this line also passed MIL-STD-810H durability, which is the rough-and-tumble military-grade testing that proves it’s tough enough to outlive most teenagers’ backpacks.

The ASUS Zenbook A16 is another great pick that packs a 16-inch 3K OLED touchscreen, 48GB of memory, and 1TB of storage into a laptop that weighs less than three pounds — which is pretty amazing. It runs on a Snapdragon ARM processor, so check required programs, drivers, and accessories before buying. Most everyday Windows apps work, but Microsoft says specialized software and some hardware may still have compatibility limits on ARM-based PCs.

This one is more of the splurge at just under $1,550 right now, but it’s the last laptop you should ever have to buy a student, and others with similar features run nearly $1,000 more.

Check student discounts on both computers before you pay.

I’ll have more detailed reviews of these two specific laptops in my back-to-school buying guide, publishing by August 1st.

2. Build the phone runway before the fight starts.

Just about every child wants their own smartphone long before they’re ready for one. Outfitting kids with gadgets needs to happen in stages that progress and mature like they do. I’ve always called these “training wheel gadgets.” It’s good to have them because it teaches kids how to manage tech rather than binge on it the second they get out of your sight. But it’s all about striking a healthy balance.

Open ChatGPT’s voice mode with your child and say,

Our 9-year-old needs to reach us after school but does not need the whole internet. Ask us one question at a time, then build a step-by-step device plan with a review date

Let it interview everyone, then discuss the result (all via that voice mode) until it feels right and fits your family.

The plan might move from a home phone to a watch, then to a limited phone and eventually to a smartphone. That turns “no” into “not yet, and here is what comes next.”

Ask one final question:

What is the least-connected device that solves the problem?

I steered it toward a landline phone because, after reviewing three of them quite extensively over the past few months, I think it’s the best way to break kids into healthy first tech use, without social media, YouTube, or other stuff that’s too mature for their budding brains. I highly recommend that as the place to start.

Pinwheel Home is the best landline phone I’ve reviewed yet for overall cost and how easy it is to set up (start to finish took less than ten minutes). The devices start at $68 and use Wi-Fi, so they don’t need a phone jack or an installer. Parents approve contacts and calling hours from an app on their phone. Unapproved numbers never ring through. Calls between Pinwheel devices and 911 access are included, but calls to regular numbers require a monthly plan. Those run just under $7 for calls to five regular phone numbers or just under $10 for unlimited calling.

I love hearing kids actually having conversations with each other, their grandparents, and their favorite aunt (that’s me, 🤣).

3. Talk through the schedule instead of typing it.

Open voice mode again and tell it the schedule you’re trying to figure out. Here’s what I just tried:

My third grader starts at 8:40, my seventh grader starts at 7:55, and soccer begins at 5 on Tuesday and Thursday. Build our weekly schedule and make the mornings calmer than last year.

Push back when the plan gets ridiculous. Tell it, “She is never waking up at 6:15,” and ask it to revise the schedule. It can share this new plan right to any calendars you connect with it, like Google or Outlook.

Then put the finished calendar where everyone can see it. I LOVE (yes, I’m shouting because I’m excited) this new Greenlight Family Hub that combines Google, Outlook, and iCloud calendars and includes chores, photos, and shared lists without a subscription. Greenlight charges almost $10 a month for optional AI voice and family location features. The hubs come in two different sizes; the 10.1-inch version is just under $200, while the 15.6-inch model is just under $300. Let this live right inside the front door, on the kitchen counter, or wherever you need one dedicated family screen that keeps your life super organized.

I have a bunch more of these “back to school ChatGPT hacks” that I’ll continue to share with you 1-3 at a time. I also have more great back-to-school gadget picks coming this Friday.

Stay tuned, tech+life is heating up again!

What’s your favorite tech-life hack?

My producer friend Suzanne told me this morning that she uses ChatGPT to buy watermelon, and it works super well. She snaps a photo of the choices at her local grocery store, and it literally tells her which one to pick. That’s super genius because we’ve bought three really bad ones in a row that we basically had to throw out.

What’s yours?

I checked all of these prices and features on July 27, 2026, but they may change. Retailer links may be affiliate links, but the reporting and recommendations remain independent.

I got caught in the Pacific Northwest’s version of a torrential downpour yesterday: thunder, lightning, hail, raindrops the size of grapes. We don’t really do weather like that here, so by late afternoon, when Apple’s Weather app showed one harmless little cloud spooning a smiling sun, I totally believed it.

Rookie mistake.

After a full day cooped up inside, everyone on the farm needed out. So around 4 o’clock, I climbed on one horse, led a second, and set off with three dogs unleashing a day’s worth of zoomies.

Then came our “oh, hail no” moment. Like, for real, little icy rock chunks started coming in hot. Erm…cold. Not 30 seconds later, the sky opened like upstairs had kicked over an Olympic-size swimming pool. Rain splashed down, hail melted into my hair, and our whole rag-tag lot soaked in seconds. At one point, I swear, the horse I was leading (from the back of another horse) gave up on trotting and simply water-skied.

We made a break for my aunt’s barn, waited out the worst of it, and eventually squished home — soggy, ridiculous, and with a story you only get when technology fails you spectacularly.

Apple Weather: one little cloud, smiling sun, 100% chance of betrayal.

And yet here I am, doubling down on trusting even more unfinished Apple software: the iOS 27 public beta, living on my phone all week because apparently I enjoy turning perfectly functional technology into a trust exercise.

In today’s newsletter:

📱 Should you install the iOS 27 beta? My verdict in 60 seconds — what’s awesome, what’s wet paint.

🎒 My new back-to-school buying rules — why the smartest time to buy the laptop is right now, and the trap hiding at $300.

👀 As The Silicon Valley Turns: we now know what Apple and OpenAI are really fighting over — and it has eyes.

✨ And a TRY THIS: the $300 color analysis ChatGPT does for free from one selfie.

Let’s dive in. 👇

📱 iOS 27 IN 60 SECONDS

The beta’s been on my phone all week, so let me answer the question many of are asking:

Should you download it? Honestly — yes. This time, yes. I’ve spent years telling you to skip day-one betas, so that surprises even me. This is the first beta in years — maybe ever — that feels stable enough for me to recommend on day one. I said it on camera this week, and I’ll say it again and again. I’ve had it on my phone all week, I LOVE it, and I keep finding new things.

The grown-up caveats still apply: back up first (iCloud or a computer — no exceptions), expect a rogue bug or faster battery drain here and there, and know that going back to iOS 26 isn’t a tap — it’s a wipe-and-restore project. If your phone is truly mission-critical — health apps, work, a kid’s check-in — the polished version arrives free in September and waiting costs you nothing but Apple fan-girl/boy bragging rights.

One insider heads-up: the new Siri sits behind a waitlist after you install. Mine cleared in a minute or two — good luck with yours; your mileage may vary.

(And yes, the classic advice is “put betas on a spare phone.” Who actually has a spare phone — aside from a tech reviewer like me?! As long as you realize there’s a learning curve, this is the rare beta where you don’t need one.)

What you get once you’re in: your phone truly does get faster, not slower, even on older iPhones. Parents get the biggest Screen Time overhaul in years — one-tap controls, time allowances by app category, school-day schedules, and Ask to Browse, which extends approval to websites, not just apps. And Reframe lets Apple Intelligence fix a photo’s angle after you take it. Botched the shot at the summer T-ball game? Fix it by Tuesday.

The wet paint: the new Siri — impressive one minute, confidently lost the next. The beta is solid; Siri is the part still drying. Temper your waitlist excitement accordingly, and check back in September.

(The full feature-by-feature breakdown hits Techish this weekend.)

🎧 ON THE POD: SHOULD KIDS START WITH A WATCH INSTEAD OF A PHONE?

This week on Techish with Jennifer Jolly (that’s me), I sit down with Russell York, founder of kids’ smartwatch maker Cosmo, for the question so many parents ask: Is a smartwatch the right first device? We get into Australia’s and the UK’s under-16 social media crackdowns, what the research actually says about early smartphone adoption, why “training wheels” tech is having a moment — and yes, I push back, including on the surveillance question this whole category would rather not talk about.

Watch the full episode [HERE], listen wherever you get your podcasts, or right here → [WGN]

🎒 THE BIG STORY: BUY THE LAPTOP NOW — MY BACK-TO-SCHOOL RULES FOR THE YEAR TECH PRICES BROKE

Every July and August, I test a mountain of back-to-school tech and tell you what’s promising. This year, I’m doing something different: I’m telling you what to buy right now — because for the first time in my reviewing career, waiting for my full, in-depth reviews could cost you money.

Here’s why, and here are my rules.

Rule 1: If you need a laptop this school year, buy it now

AI companies are buying up all the memory chips. The same RAM that goes into your kid’s laptop? Data centers want it too, and they’ll pay whatever it takes. So the chipmakers are selling to them first, and everybody else is fighting over what’s left.

The result? Industry analysts say memory prices are on track to more than double this year. Laptop prices are already creeping up $100, $200 at a time. And this isn’t blowing over by Christmas — building new chip factories takes years, so we’re stuck with it for a while.

So the usual back-to-school advice — wait for Labor Day sales — is backward in 2026. The deal you see today may be the best deal you see all year.

Rule 2: 16GB of RAM is the magic number

Pay attention to RAM, because that’s where laptop makers are cutting corners to keep prices looking normal. Consumer Reports calls it shrinkflation, which is a much more PG term than what I call it. Same price tag, worse laptop.

A real-world example from my own desk: I’m testing the Lenovo IdeaPad Slim 3i, and as I write this, it’s $400 off — just under $900. My full review isn’t done, but the price is the story: 16GB of RAM and 512GB of storage at a price where other brands are starting to ship 8GB and hoping you won’t notice. Deals like this are what the memory crunch is killing off, which is why I’m flagging it now instead of waiting. (If the price has crept back up by the time you click, that’s RAMageddon working in real time.)

Rule 3: The $300 laptop is a trap this year — with one exception

I get this question every August, and this year the answer changed: Normally, a $300 laptop is a compromise. This year, it’s a trap. That’s the price where all the corner-cutting hides — 4GB of RAM, slow storage, dim screens — because the cheap memory that used to make those machines okay-ish is gone. A $300 Windows laptop in 2026 will feel slow the day you open the box and be even more painful by spring break.

The exception: Chromebooks. If your youngster lives in Google Classroom and Docs — and way most elementary and middle schoolers do — a $300 Chromebook is still totally fine, because ChromeOS doesn’t need much to run well. Just be honest with yourself about what you’re getting: a browser with a keyboard, not a computer that grows with them.

My rule this year: under $400, buy a Chromebook or don’t buy at all. For a real Windows machine, start where the 16GB models live.

Rule 4: Refurbished is smarter than ever — but read this first

Can’t stomach the new prices? The same AI boom that's pushing up laptop costs — Apple just raised Mac and iPad prices by as much as $200 — makes refurbished tech the smart move for many families. The short rules: buy certified refurbished (manufacturer stores or major retailers with warranties), not “used, trust me” from a stranger; check the battery health before you check out; and never buy without a return window — 90 days minimum. My full refurbished guide, with where to shop and the new catches this year, hits Techish this weekend.

Rule 5: If there’s a student in the house, never pay sticker price

Here’s the claw-back for those price hikes: education discounts. Almost every major tech brand runs an education store, and eligibility is broader than most families realize. Case in point: Apple’s back-to-school offer is live right now — education pricing plus an Apple Gift Card from $100 to $150 with an eligible Mac or iPad, through August 27, and it covers college students, their parents, faculty, staff, and homeschool teachers. (Education pricing also drops the new MacBook Neo to $599 — basically undoing the AI-price hike.) Verification takes two minutes with a school email.

Three things to know: the discounts usually stack on top of sale prices — check the education store after you find the deal, not instead of it. A gift card is not a discount — those promos only count as savings if you were going to buy the accessory anyway. And verify before you count on it — offers change midsummer, and two minutes on the terms page beats a surprise at checkout.

The verdicts: kids’ tech that earned my recommendation

Not everything on my desk is still “in testing.” Some of it has already survived the wringer.

Kids’ smartwatches — the training wheels of family tech. I’m in the final stretch of a months-long, head-to-head test of three of the highest rated ones — TickTalk, Cosmo, and Pinwheel. Here’s what I know for sure: Calling, texting, and GPS with zero social media are a great deal at a fraction of a smartphone’s cost, and each has real strengths and real quirks. The winners, by age group and with all my testing receipts, land in about a week.

Headphones with a permanent parent in the volume knob. The iClever Q950 ($70) caps the volume at safe-for-young-ears levels no matter how hard your kid mashes the button, and noise canceling kills the background racket so they never want to crank it in the first place. It’s certified under an actual hearing-protection standard — rare in the kids’ headphone aisle — with a 60-hour battery that survives the whole school week.

The dorm room glow-up (and the analog counterattack)

The college end of my testing table is trending gloriously backward in time. (Google’s search data says I’m not imagining it: “dorm designer” searches just spiked over 500%, and the top trending dorm aesthetic is “coastal.” Fairy lights, as ever, remain undefeated.

Tiny Vinyl is exactly what it sounds like: real 4-inch vinyl records ($15) from Olivia Rodrigo, Chappell Roan, BTS, and more, plus an adorably tiny record player ($50) that just launched exclusively at Target. Music with zero screens and zero algorithm.

IKEA’s viral VARMBLIXT “Donut” lamp ($100) just went smart, with colors you change by remote or app — study glow, movie mode, and the room’s whole personality in a single ring.

The $10 mic drop: IKEA’s KALLSUP Bluetooth speaker made a splash at CES for sounding way better than ten bucks should. Backpack-sized, dorm-priced, and cheap enough not to cry about when a roommate borrows it forever.

And because you know I’m a sucker for nostalgia: the Camp Snap 2 ($69.95) blends a vintage point-and-shoot’s simplicity with an 8MP sensor, room for 2,000 photos, and USB-C. Its new sibling, the CS8 Video Camera ($199), channels the iconic Super 8, with no playback screen at all, so kids (and by kids, I mean ME) can capture the moment instead of reviewing it for 10 minutes. Screen-free cameras for the screen-free childhood movement. The theme writes itself.

Still in the lab

The rest of the pile is mid-wringer: the ASUS Zenbook A16 (16 inches, a frankly absurd 48GB of RAM, making the “buy once, keep all four years” argument), the battle of the comeback landlines (I promised you a winner — it’s close), the new Greenlight Family Hub, and a MagSafe case with an E-Ink screen your kid can update from campus. AND SO MUCH MORE. Verdicts soon.

👀 AS THE SILICON VALLEY TURNS

Last week: Apple sued OpenAI over stolen hardware secrets. This week, a suspiciously well-timed leak revealed what those secrets allegedly helped build — a screen-free speaker with a camera, a battery so it can roam your house, and mechanical parts that move on their own so it feels alive. No screen, but it has eyes. (Um…how do you feel about that?)

Apple wants a court order blocking OpenAI from using its secrets — which, if granted, could stall OpenAI’s hardware plans while Apple builds its own kitchen-counter device. (OpenAI’s side says the speaker is nothing like anything Apple sells; nothing’s been proven.)

The punchline: OpenAI’s actual first hardware product went on sale this week, and it’s... a $230 keyboard accessory for programmers with a knob that makes the AI think harder. I wish I were kidding about the knob.

The craziest part in all of this, though, is that they’re still partners. ChatGPT is literally built into your iPhone. That tangled mess, and why it matters to you, is next week’s story, and it’s messier than you think.

Would you let a robot with a camera and a personality live in your kitchen? I will, but only if it actually does the dishes.

Leave a comment

✨ TRY THIS: THE $300 COLOR ANALYSIS — FOR FREE

Seasonal color analysis — the “which colors actually look good on me” service that took over TikTok — runs a few hundred dollars with a human expert. ChatGPT does a surprisingly good version in minutes. I know because I just did it, and I am apparently a Light Spring. More on that in a second.

Here’s how to do it (I have the paid Pro version of ChatGPT, so that’s what I used to test this one out):

1. Take one clear, well-lit photo of your face — natural light, no filter. I started with this one (which I took right after someone called Punky Brewster 🤣.

1. Attach it to ChatGPT and ask: “Do a seasonal color analysis. What season am I, and what’s a palette of my best colors?”

2. Then ask it to generate your palette as an image — you’ll get a color card to save on your phone for shopping. (Because all those words don’t mean anything to me, I need to see it all in action.)

3. Okay, now for the big payoff: Upload one more photo and ask, “Can you show my best and worst colors on me?” It generates side-by-sides of you wearing your winners and your losers.

My results: coral was “the winner,” turquoise apparently sharpens my eyes, and burgundy “pulls your face downward” — rude, but after seeing the side-by-side... its not wrong. It even called out the hoodie I was wearing in the photo as one of my colors, then told me to swap stark white for cream and black for a warm navy. My shopping list just rewrote itself.

The bonus round for the season: “Now tell me which colors to look for — and which to avoid — when I shop this fall.”

Then hit reply and tell me your season — I’m collecting the results, and I want to know if the robot stylist is onto something.

Leave a comment

Fair warning, as always: you’re handing an AI photos of your face. That’s the trade — decide how you feel about it first. And treat the verdict as a fun second opinion, not gospel — even ChatGPT admits the try-on images are directional, not a controlled draping test.

(This is tip #2 from my upcoming insider guide — the rest are coming to paid subscribers first. Stay tuned.)Fair warning, as always: you’re handing an AI a photo of your face. That’s the trade — decide how you feel about it first. And treat the verdict as a fun second opinion, not gospel.

(This is tip #2 from my upcoming insider guide — the rest are coming to paid subscribers first. Stay tuned.)

💬 THE ASK: BRING ME YOUR BACK-TO-SCHOOL DILEMMA

This week’s assignment: hit reply, comment, or message and tell me the back-to-school tech decision you’re stuck on. Give me three things: who it’s for (age/grade), your budget, and what they actually need it to do. For example: “9th grader, $500, needs it for essays and Canvas but will absolutely try to game on it and I want to block that, most of the time.”

I’ll pick the best dilemmas and answer them in next Friday’s issue — with your first name on it, unless you tell me otherwise. Your problem becomes everyone’s answer.

(And if you take me up on the iOS 27 beta, or are already using it: send me the best and most broken things you find. I’m collecting both.)

I read every reply. Yours often decides what I test next.

🎮 NEXT WEEK: I’m headed to the Games for Change Festival in NYC and can’t wait to spend time with the people using games to actually help kids rather than just harvest their attention. I’m moderating this panel discussion with some of the top minds in the world when it comes to raising healthy — and tech-connected — kids.

Plus: your smartwatch winners (with all my testing receipts), your back-to-school dilemmas answered, and the Apple–OpenAI mess finally untangled. It’s going to be a big one. Get your replies in. 👇

Leave a comment

More soon.

XO, Jenn

Techish stays free every week because readers who can chip in help keep it going. If this issue saved you a Google search, a bad purchase, or $200 on a laptop, it did its job.

Here’s the ironic part, though: Nearly three-quarters of the parents surveyed said their child already has a smartphone. And even more than that (77%) said they wish they had a simpler way to keep their kid connected without relying on the internet or social media.

I’ve heard that same thing over and over for the past decade. I’ve never once talked with a parent, or a teen, or a young adult, for that matter, who wishes their parents had given them a phone earlier. It’s always exactly the opposite.

Cosmo CEO Russell York calls that gap "techxiety."

I just know it as the knot in your stomach when your child needs to reach you, and the easiest answer is the device you’re not sure they’re ready for.

There are times when our kids need to reach us, and vice versa, especially as they get more independent. When I was growing up, that meant using a friend's parents' landline or finding a payphone. Now, it often means a hand-me-down smartphone and a wide-open door to YouTube, group chats, social media, strangers, algorithms, and every app adults are still trying to pry out of their own hands.

That’s the parenting problem at the center of my latest Techish conversation with Jennifer Jolly, featuring Russell York, founder and CEO of Cosmo, a company that makes connected watches and devices for kids.

Cosmo is part of this fast-growing “first device” category, so I went into the interview with the same questions I ask about any child-tech product: What does it actually do? What does it block? What can a child still get around? What does it cost? And most importantly, what problem does it really solve?

The larger question is the one a lot of parents are already wrestling with: What comes before the first smartphone?

Russell says we skipped a step. “Whenever I hear other generations — maybe Gen Xers, if we’ll just throw them under the bus here — saying, ‘I didn’t have a phone when I was a kid.’ But you know what you had? You had 2.5 million payphones around the U.S., and you probably had a quarter in your pocket, and you probably knew how to call collect to Mom and Dad, and let the phone ring once at the house so they knew you’d arrived. We have always connected kids. It’s always been something that society understood was a right that everybody had — to call for help, to connect with your parents.”

I tested Cosmo’s newest watch, and it’s better in many ways than all the others I’ve reviewed, with some caveats, of course, which are in the show. (Like subscriptons. We hate subscriptions, especially two-year commitments on new plans.) I also review the Pinwheel Watch and TickTalk and will link to that full review as soon as I finish it up! (These have come a really long way since I last reviewed them all just four years ago!)

There’s a lot more in the full conversation — why Russell says the big tech, trillion-dollar platforms could fix this tomorrow and won’t, whether Apple’s biggest-ever parental controls update will make a difference, and a whole lot more insight since he made the switch from intelligence and cybersecurity to protecting kids.

Listen to the full episode of Techish with Jennifer Jolly wherever you get your podcasts, or watch it here.

Also? The most powerful safety tool in your house isn’t on your child’s wrist — it’s you paying attention. You’re already on the right track.

I’m doing a happy dance because Meta already shut down the Muse Image feature that let strangers turn public Instagram photos into fake scenes with a simple @mention.

I tested it, the backlash hit, and Meta pulled the feature three days after launch.

I updated our big story below with what happened, why the company backed down, and why your Instagram sharing settings still matter. Read it → [HERE]

In today’s newsletter:

• ChatGPT launched a huge Voice upgrade that lets you talk to it. Here’s everything I’ve learned using it way more than I expected to.

• 🌎 OMG, As The Silicon Valley Turns: Apple’s just filed a juicy lawsuit against OpenAI — why this matters to you 👀

• And, the iPhone setting that could turn an old phone into the perfect first phone for your kid.

Every week, I think this can’t get any more packed with useful tips and gadgety gossip, and then it just … does. Let’s dive in. 👇

📸 UPDATE: META KILLED THE INSTAGRAM AI FEATURE I WARNED YOU ABOUT

Meta launched Muse Image on Tuesday as its newest AI image generator.

Because Erling Haaland — the internet’s favorite Viking golden retriever, who has already won this World Cup in every category except the actual trophy — has a public Instagram account, it took me about 60 seconds to manufacture a completely fake relationship with him.

Then I made a few more spectacularly fake photos with Messi, Oprah, Cristiano Ronaldo, and turned myself into a dog.

Only after I had thoroughly enjoyed myself did I show you where Meta buried the off switch. (C’mon, I take my consumer-protection responsibilities seriously, but I’m not made of stone.)

The problem was obvious: Meta automatically included eligible public Instagram accounts. Anyone could type an @mention and pull someone’s photos into a new AI scene without following them, asking permission, or telling them afterward.

My examples were mostly ridiculous and clearly labeled. The same tool could just as easily create a fake endorsement, manufacture evidence of an affair, humiliate a kid, or hand a scammer a convincing new prop.

Three days later, after backlash from SAG-AFTRA, CAA, creators, privacy advocates and reporters, Meta pulled the feature and admitted it had “missed the mark.”

Duh.

I can’t believe anyone over at Meta thought this was a good idea in the first place. Seriously. How could this feature pass through product meetings, privacy reviews, legal teams, and really smart people conversations — without anyone standing up and saying, “Wait. Are people really going to love discovering that strangers can turn their faces into AI props without asking?” Blech.

Muse Image still exists, but people can no longer summon photos directly from public Instagram accounts with an @mention. The images people created during those three days still exist. (Hello, I’m framing mine…) Meta cannot pull them back from phones, group chats, social feeds, or wherever else they landed.

I am also leaving the Instagram reuse permission turned off. The specific public-account feature is gone, but the Sharing and reuse controls still govern other ways people can reuse your posts and Reels.

To check yours, open Instagram and go to:

Profile → three lines → Sharing and reuse

If you still see Allow people to use your content on Instagram and with AI features at Meta, turn off both Posts and Reels.

Changing the setting cannot stop someone from taking a screenshot of a public photo and uploading it to another AI generator. A public account remains public. It does, however, give Meta a clear answer about whether you want your content fed directly into its own reuse tools.

Don’t forget, Meta has made this bet before. It used facial recognition to suggest photo tags. It used public posts to help train its AI. The company launches first, buries the controls, and counts on most people never noticing.

This week, we noticed. Celebrate that. Then remember it the next time Instagram rolls out a shiny new feature with the permission box already checked for you.

See every picture I made, the exact prompt I used, and the full update here → [FULL STORY]

🎙️ THE BIG STORY: CHATGPT VOICE GOT A NEW BRAIN — AND I’M USING IT. A LOT.

Quick show of hands: did you know you can talk to ChatGPT out loud? There’s a little voice icon next to the message box; it’s been there for a while, and most people have never tapped it once. Honestly? You weren’t missing much—until now.

OpenAI rebuilt ChatGPT Voice on new models called GPT-Live, and for the first time, you can talk with it instead of at it. Interrupt it mid-answer. Trail off, change your mind halfway through a sentence, go quiet to think — it waits. You can ramble.

Rambling is kind of the whole point, because talking is how people actually think. Typing was always the bottleneck. Now the twenty minutes in the car becomes the meeting with yourself you never schedule. The dishes become time to rehearse that sensitive conversation with a partner who has zero stakes and infinite patience. The afternoon trail run becomes where the newsletter gets outlined. (Ask me how I know.)

I did not plan on liking this. I have spent two decades watching “just talk to it” go sideways. Siri mangles simple requests. Alexa once recorded a family’s private conversation and sent it to someone in their contacts. Contractors reviewing voice-assistant recordings have heard things nobody intended to share.

Then I started using the new ChatGPT Voice — and I can’t believe how different and good it is.

How to find it and turn it on

Open the ChatGPT and look at the message box where you normally type. Tap the Voice icon on the right side. Do not tap the little microphone that records a message and turns it into text; tap the Voice button to open a live conversation.

The first time you use it, ChatGPT asks for permission to use your microphone and lets you choose a voice (I’m using “Vale”). After that, you just start talking. On a computer, go to ChatGPT.com and tap the same Voice icon beside the prompt box.

To make sure you have the new version, go to Settings → Voice → Live. You should also see a small Live label at the top of the conversation after Voice opens. If Live does not appear in your settings, update the ChatGPT app. OpenAI is rolling it out gradually, so it may not have reached your account yet.

The feature is still called ChatGPT Voice. GPT-Live is the technology running underneath it. Paid accounts use GPT-Live-1, while free accounts use GPT-Live-1 mini. You do not have to choose between those two models; ChatGPT picks the right one for your plan.

When you ask a harder question, GPT-Live can quietly hand the job to GPT-5.5 in the background. That means you can ask it to search the web, reason through a complicated problem or work on something more involved without ending the voice conversation and starting over in text.

Live works in the ChatGPT iPhone and Android apps and at ChatGPT.com. It does not support live video or screen sharing yet. For those, eligible subscribers still need to switch back to Advanced under Settings → Voice.

Is it always listening?

I asked OpenAI the question that I’ve been trained by a decade of creepy assistants to ask. The answer, on the record: no. There is no wake word. The microphone turns on when you tap the voice button, and not before. In their words, it is not an always-on model — if you never tap in, it is not listening to you.

Then I did the very meta thing and asked the microphone about the microphone. Mid-conversation, out loud, I grilled ChatGPT Voice about its own listening habits. It confirmed the tap-to-start part, and then got pretty honest about the rest:

• While a session is open, the mic stays live; it can hear anyone else in the room, and it might even respond to them.

• Its advice — its own, not mine — was to end the session before having a private conversation.

• When I brought up the long history of assistants listening when they shouldn’t, it told me that once you close a session, it’s “not supposed to” keep listening in the background.

I’ll let you decide how comforting “not supposed to” is. The practical rule is simple: you start it with a tap, and you should be the one who ends it. Don’t leave a session running while the rest of your household wanders through.

What happens to your voice?

OpenAI told me that audio clips stick around for up to 30 days so you can resume conversations, and the company says it does not use your actual audio or video clips to train its models unless you choose to share them separately.

The transcript is a different matter. If you have "Improve the model for everyone" turned on, OpenAI may use transcripts of your voice conversations for training, depending on your plan and settings.

Check yours here:

ChatGPT → profile icon → Settings → Data Controls

Turn off Improve the model for everyone if you do not want your conversation transcripts used for training.

If you leave that setting on, make sure “Include your audio recordings” and “Include your video recordings” remain off unless you deliberately want to share those clips.

OpenAI says deleting a chat also schedules its associated audio and video for deletion within 30 days, although the company may retain some data longer for legal, safety, or security reasons. Archiving a chat does not delete it.

I still won’t dictate a bank password, discuss a confidential work deal, or treat a pleasant-sounding chatbot like a licensed therapist (let alone my best friend or spouse). Remember, a voice can feel warm, attentive, and human long before the machine behind it deserves that level of trust.

CHEAT SHEET: HOW TO TURN ON CHATGPT VOICE

WORKS ON: iPhone, Android and ChatGPT.com. Both free and paid accounts get it, though free accounts use the lighter GPT-Live mini model.

1. Open the ChatGPT app, or go to ChatGPT.com.

2. Look at the message box where you normally type. On the right, you will see a small Voice icon that looks like a sound wave. Tap it.

   Do not confuse it with the microphone icon. The microphone only turns what you say into typed text. The Voice button opens an actual conversation.

3. The first time you use Voice, ChatGPT will ask for microphone permission and let you choose a voice. I use Vale.

4. When the Voice screen opens, look for the small Live label at the top. That is how you know you have the new GPT-Live upgrade. No Live label? Update the app and check again later. OpenAI is still rolling it out.

5. Start talking. There is no wake word and no button to hold down.

6. When you finish, tap the X to end the session. While that session remains open, the microphone remains live and may hear other people in the room. You start it with a tap, and you should be the one who ends it.

🎯 TRY THIS: PUT VOICE ONE PRESS AWAY

If you have an iPhone 15 Pro or newer, map the Action Button to ChatGPT Voice. Then, the second a thought hits — in the car, on a walk, elbow-deep in dinner — you can press one button and start talking instead of opening your phone, finding the app, and losing the thought somewhere along the way.

Go to:

Settings → Action Button → Controls → ChatGPT → Start Voice Conversation

No Action Button? Add the ChatGPT Voice control to your iPhone’s Control Center or Lock Screen instead.

That tiny change is what turns Voice from a feature you forget exists into something you might actually use.

This is tip No. 1 in my upcoming guide. I’m testing 15 more genius insider tips like these, with the exact words to say, the settings to change, and the tricks that sound smart but fall apart in real life. Paid subscribers get the whole thing first. Stay tuned!

🍎 APPLE SUED OPENAI — AND WOW, THIS ONE IS MESSY 👀

Apple sued OpenAI on Friday — its own partner — claiming the ChatGPT maker used stolen Apple secrets to jump-start its mystery gadget.

Apple claims OpenAI turned recruiting into a trade-secret shopping spree. One former engineer allegedly kept his Apple laptop, found a bug that let him back into the company network, and downloaded dozens of confidential hardware files.

Apple also says OpenAI’s hardware chief told job candidates to bring “actual Apple parts” to interviews for show-and-tell.

OpenAI says it has no interest in anyone else’s trade secrets, and none of this has been proven.

The case could delay OpenAI’s first device, damage one of the strangest partnerships in tech, and help decide who controls whatever comes after the smartphone.

And I, for one, am here for all of it. I love a good made-for-reality-TV Bravo-esque drama, especially when the couple is still sharing an iPhone.

📡 WORTH YOUR CLICKS

Meta fixed the spy-glasses hack after Joanna Stern exposed it

Last month, New Things rockstar jouralist, Joanna Stern, paid someone $100 to drill the recording light out of a pair of Ray-Ban Meta glasses. The modified glasses could film people without showing the white light meant to warn everyone nearby. Creepy, right?!

This week, Meta released a mandatory software update that disables the camera when the glasses detect tampering with the recording light. Meta told Joanna directly that the update addressed her reporting.

Facebook’s 0 for 2 this week, but doing a good job fixing what we’re mad about.

Read Joanna’s follow-up at The New Things.

Your child’s first phone may already be sitting in your junk drawer

WIRED’s Jeremy White found a brilliant use for an accessibility feature Apple barely promotes.

It is called Assistive Access, and Apple originally designed it for people with cognitive disabilities. It can also turn an old iPhone into a simple first phone for a child.

You decide exactly which apps appear. The icons become large and simple. You can allow Calls, Messages, Maps, Camera, Photos, and Music while leaving out Safari, social media, and games.

The smartest part: links sent through Messages do not open when web access is disabled. Your child keeps navigating and Find My tracking without gaining access through the back door.

Set it up here:

Settings → Accessibility → Assistive Access → Set Up Assistive Access

Choose the apps you want, pick a separate Assistive Access passcode and leave everything else off.

A few warnings: the mode can feel sluggish, voicemail does not work, and you have to exit Assistive Access before turning off the phone. Still, the old iPhone sitting in your drawer may do the job without another parental-control subscription.

Read Jeremy White’s full WIRED story.

Google Erling Haaland before 2 p.m. Pacific

And just in case you have no idea why I’m making such a big deal about Erling Haaland, come out from that rock you’ve been hiding under. He’s the breakout star of this World Cup —part striker, part Viking, and part meme machine.

Haaland’s led Norway into a quarterfinal against England while posting like the funniest person in your group chat. Google even created a special Viking Row animation tied to Norway’s celebration. Haaland told everyone to search his name, so naturally, we must.

Norway plays England Saturday at 2 p.m. Pacific, which gives you just enough time to Google him before kickoff.

Read Slate’s story about Haaland’s World Cup takeover.

💬 THE ASK

Which story made you laugh, swear, or immediately reach for your phone?

Did you check your Instagram settings? Try ChatGPT Voice? Dig an old iPhone out of a drawer for your kid? Or are you still stuck on the part where Apple is suing the company whose chatbot lives inside its own phone?

Hit reply and tell me the one thing from this issue that you are actually going to try, change, or send to someone else.

I read every reply, and yours often tells me what I need to test, explain, or investigate next.

🎒 WHAT’S NEXT?

My office has officially entered back-to-school testing season, which means my desk has disappeared under a small mountain of laptops, chargers, kids' gadgets, and products that claim to finally organize my family.

This year’s gear already points to a few bigger trends.

Laptop makers are stretching the definition of what a school computer should be, which is awesome because laptop prices are going up, up, up. I look at the best budget, splurge, and refurbished models.

Kids’ tech is moving away from handing every child a smartphone and hoping for the best. Several companies are bringing back the landline as a screen-free Wi-Fi phone. I’ve tested three of the brands so far, and there’s already one clear winner.

Then there are the gadgets trying to run the entire household. Every family-tech company wants to build the command center that finally gets everyone organized. I am testing the newest one to see if it reduces the chaos or gives my family a larger, more expensive place to ignore the grocery list.

Even phone cases have entered the era of customization. Wait till you see the latest MagSafe case with a battery-free E-Ink screen that lets you change the picture on the back of your phone in seconds.

Those products represent the first wave. Over the next few weeks, I will sort through which back-to-school trends deserve your money, which gadgets solve a real family problem, and which products should remain safely sealed inside the review box.

If a back-to-school gadget has caught your eye, hit reply and tell me what you want me to put through the wringer.

So much good tech, so little desk left.

More soon.

XO,
Jenn

Techish stays free every week because readers who can chip in help keep it going. If this issue saved you a Google search, a bad purchase, or one creepy AI image of yourself, it did its job.

After backlash from just about everyone — including SAG-AFTRA, talent agency CAA, privacy advocates, creators and reporters (hi 👋) — Meta said people could no longer reference public Instagram accounts by @mention and admitted the feature had “missed the mark.”

The AI images people created during those three days still exist. The Sharing and Reuse setting below also continues to control how Meta can use your content with its AI features more broadly, so checking it remains worth two minutes of your weekend.

Also? I couldn’t believe Meta launched Muse AI [holy tone-deaf Batman] in the first place. The company watched OpenAI get pummeled over Sora’s face-swapping rules and somehow decided the real problem was asking permission at all.

Why? Do we really matter so little that you can’t even ask?

Meta has something no other AI company can build overnight — nearly two decades of our faces, families, vacations, weddings, children, bad bangs, and perfection-maxxing filtered brunches. That photo archive gives it an enormous advantage, and the company bet that turning all of us into AI source material by default would be worth the outrage.

Meta has made this bet before. It did it with facial-recognition tagging. It did it when it used public posts to train its AI. Launch first, bury the controls, and count on most people never noticing.

This week, people noticed, and the company folded in three days. Celebrate that. Then remember it the next time a tech company rolls out a shiny new feature with the permission box already checked for you.

My original story continues below, as published.

I have a public Instagram account. It is part family album, part work portfolio, and part running record of every bad hairstyle since 2010.

Starting this week, Meta’s new AI image generator, Muse, can use photos from my public Instagram account to create brand-new pictures of me that never happened.

If your Instagram account is public, it can do the same thing with yours. So naturally, I had to try it.

First Up: Me and the Internet’s Favorite Viking

Yep, this is me standing shoulder-to-shoulder with soccer superstar Erling Haaland—the 6-foot-4 Viking golden retriever currently taking over the internet.

It’s 100% fake.

I started with a photo from Haaland’s public Instagram account, showing him with his real-life girlfriend. Then, I attached a reference photo of me and gave Muse one simple instruction:

“Swap me into this pose with @erling using the reference photo I attached of me in pink pants.”

That was it, and a few seconds later, Muse gave me the image above.

Sorry, Erling. Sorry, Erling’s girlfriend. I promise this fake relationship lasted only about 60 seconds.

Once I stopped sending the photo to my friends and family as if it were real (they all knew it wasn’t), I played around with Muse a little more.

As you can see below, it generated more World Cup moments. It turned me into a golden retriever hosting a podcast. It drew me as a Pixar character interviewing David Pogue and put me on a couch with Oprah Winfrey and Cristiano Ronaldo. It even showed a winning moment with Lionel Messi and me. And of course, it offered a red-card moment with a prompt of its own.

The results are fun, and very Sora-esque — except the Muse AI hands have all the right number of fingers.

What Meta’s New Muse Can Do

Meta says Muse can combine several images, follow conversational instructions, and bring people from public Instagram profiles into a new picture when someone types an @mention. That’s all true.

Here’s where it can all go sideways, really fast: The feature applies to eligible adult public Instagram accounts, and it’s turned on by default. Anyone using Muse can @mention one of those accounts and use its public photos as reference material for a new AI image. They don’t have to follow you, ask your permission, or notify you afterward.

The company also says the feature is meant for fun — birthday invitations, fantasy scenes, and creative projects. That’s true, too, and it’s mainly how I used it. But if life on the internet has taught us anything, it’s that not everyone will use it that way.

For every person making a birthday invitation, there is someone else who will use the same tool to embarrass, deceive, or impersonate another person.

In less time than it takes to read this paragraph, someone could create a picture of you attending an event you never went to, endorsing a product you have never used, standing beside someone you have never met, or wearing something you never owned.

Meta’s rules still apply to what people can create and share, but making the image now takes little more than a few words, an Instagram handle and about 60 seconds.

Twenty years ago, posting a public photo meant people could look at it. Today, someone can use it as the starting point for an entirely new picture. That’s a different bargain than the one any of us agreed to.

And parents, take note: Meta says accounts belonging to people under 18 are excluded from this feature, but its public explanation does not clearly address children who appear in photos posted by adults with public accounts. If you share pictures of your kids publicly, that's another reason to check this setting now.

How to Turn It Off

The simplest option is to make your Instagram account private, because Muse can’t use private accounts through this feature.

You can also keep your account public and turn off the permission:

1. Open Instagram.

2. Tap your profile picture in the lower-right corner.

3. Tap the three lines in the upper-right corner.

4. Scroll down and tap Sharing and reuse.

5. Find Allow people to reuse your content on Instagram and with AI features at Meta.

6. Turn off Posts.

7. Turn off Reels.

You need both toggles off. Instagram also lets you adjust reuse permissions for individual photos and videos, although changing the account-wide setting is much easier than checking years' worth of posts one at a time.

Before you relax, know the limits. Changing the setting stops future use through this Meta feature, but it won’t delete an AI image someone has already created. It also doesn’t make a public account private, and it can’t prevent someone from taking a screenshot of one of your photos and uploading it to another AI tool outside Meta.

I had fun with Muse. The problem is that Meta turned the feature on for eligible public accounts before many people knew it existed.

When a company suddenly launches a tool that lets strangers generate AI images using your face, it should ask first. Permission should not be buried in an Instagram setting you discover after launch.

Go check yours now, and let me know if you have trouble finding it.

Leave a comment

What are you doing to celebrate this weekend? Crabbing season kicks off around our little island tomorrow, so I hope to head out on the kayaks (I call it “crabayacking”) to drop a pot in and see what the sea-gods send our way. If it’s anything like the past few years, we’ll be cracking Dungeness, drinking a cool cocktail, and watching the pros handle the fireworks out over Puget Sound.

If you're watching fireworks this year, [your fireworks photos don't have to stink] — here's my latest article on how to nail them on your phone. READ HERE.

Okay, now onto some fun gadget-ish history:

Two and a half centuries ago, a bunch of guys in wool suits sat in a Philadelphia summer with no AC, no ceiling fan, and not so much as a cold drink, and signed their names to a document that started this whole experiment. I’ll try to remember that the next time I complain about my Wi-Fi.

Granted, in 1776, according to the OG weather nerd himself, Thomas Jefferson, it was a mild, breezy 76 degrees on that momentous day. Philly’s average Fourth is 87 now. The founders got lucky. We get a heat dome. (Not to mention a few decades of fossil-fuel-driven climate change... but I digress.)

Now, as a historic heat bubble parks itself over more than 200 million of you on the exact weekend of the big birthday, I went down a rabbit hole on how we’ve beaten the heat across all 250 years.

Beat-the-heat “high tech,” 1776 edition? Thomas Jefferson’s idea of climate control was an ice house — a pit he packed each winter with 37 wagonloads of river ice hauled up his little mountain, praying it’d last till July. Add a turret to pull the hot air up and out, and a shady porch, and that was cutting edge.

Fifty years ago, at the Bicentennial? I was five. I mostly remember fireworks, sparklers, and a lot of red, white, and blue sheet cake. Air conditioning existed — but only about 55% of American homes had it by 1980, up from just 12% in 1960. The window unit rattling in the den was the flex. Everybody else had a box fan and a wet washcloth.

And today? On what’s forecast to be the hottest Fourth of July in Washington, D.C.’s history — 103 degrees, beating a record from 1919 — I can clip a tiny air conditioner around my neck and drop my skin temperature by about 20 degrees. We wear the AC now.

Guess what the big chill hack going viral on TikTok and Reddit this week is? A bowl of ice in front of a fan. That "$0 air conditioner" everyone's sharing is the same thing Navy engineers used to cool a dying President Garfield in 1881 — and basically Jefferson's ice house with a USB fan taped to it.

Two hundred fifty years of innovation, and the internet just rediscovered the ice cube. God bless America. Let’s get into it. 👇

🥵 Beat the Heat: What Actually Works (and What’s Just Blowing Hot Air)

First, the serious part, because this heat is not messing around: More than 200 million of you are under extreme heat warnings, the grid is straining, and heat kills more Americans in an average year than hurricanes, tornadoes, and lightning combined.

Drink water, find shade or AC, never leave a kid or pet in a parked car, and please — check on the older people in your life today. Cooling centers are open in most big cities; a quick search for “[your city] cooling center“ pulls up the nearest one.

Okay. Now the gadgets:

Skip: most cheap neck fans. I know they’re everywhere right now. Here’s the problem — they just move the hot air that’s already around you. In real humidity, that’s a warm hairdryer aimed at the nape of your neck. That’s fine on a dry 80-degree day; but absolutely useless in a heat dome.

Worth it: anything with a cooling plate. The gadgets that actually work chill your skin rather than just fan it. The fancy TORRAS Coolify ($299) uses a thermoelectric plate that can drop your neck skin temperature by around 20 degrees on contact. (This is the “wearable AC” I keep raving about.) The Shark ChillPill ($130) also layers a fine mist on top of a cold plate.

The no-battery heroes: For pennies compared to the gadgets, a $15 soak-and-wring cooling towel does the trick outside, and a phase-change (PCM) cooling pad or blanket keeps you cool while you sleep. One heads-up on those viral “cooling blankets” — only the ones made with real phase-change material keep working. The cheap ones feel amazing for about a minute, then they’re just a blanket. On the label, look for “PCM” or a named version like Outlast, GlacioTex, or TitanCool. If it just says “cooling” or “ice silk,” that’s the one-minute kind.

And the hack blowing up your feed: the “$0 air conditioner” — a bowl of ice or a frozen water bottle set in front of a fan. It’s all over TikTok and Reddit right now, and… it works, sort of, while the ice lasts, then you’re back to a plain fan. Which is charming, because it’s the same physics we used 150 years ago. Some things you don’t improve — you just add a USB port.

Read more about gadgets that keep you cool HERE.

🔥 Your Cookout, Upgraded

A great cookout isn't about the fanciest grill. It's about not sweating the small stuff — the bugs, the food-safety guessing, the warm drinks. Here's what handles it for me.

1. The one that clears the deck: the Thermacell E55. Press one button and it puts up a roughly 20-foot bubble of mosquito protection — no spray, no lotion, nothing on your skin. Turn it on 15 minutes before people arrive, not after the first bite.

2. The one that saves the party: the MEATER Pro (or the cheaper ThermoPro TempSpike). Stick the wireless probe in the chicken, walk away, and your phone buzzes when it’s safe to eat. This is how you make sure the burgers aren’t the reason Uncle Rick spends the Fourth on the bathroom floor. Pull the meat a few degrees early — it keeps cooking while it rests.

3. The splurge that ended soggy sandwiches: the Anker Solix EverFrost 2, a battery-powered cooler with a fridge inside. No ice, ever. Ours rode around in our old Sprinter Van as the fridge, and it still comes to every backyard barbecue. It’ll even charge your phone.

4. The $80 crowd-pleaser: the JBL Clip 5. Waterproof, dustproof, and it clips onto a beach chair. People compliment mine constantly. Rinse the charging port with fresh water after the beach — sand kills more of these than water does.

5. The under-$30 thing everyone actually buys: a waterproof phone pouch with a floating strap (I like the Pelican Marine two-pack). Salt, sand, and sunscreen-slick hands all go straight for your phone. Drop it in the lake? It floats.

These are just five of 12. The full list — neck coolers, a beach canopy that runs on wind, and one genuinely wild red-light “therapy” sun shirt — is right HERE.

📸 And Before the Fireworks Tonight…

Every Fourth, millions of us point a phone at the sky and get back a blurry orange smudge. It doesn’t have to be that way. The two-minute version:

• Turn off the flash. It does nothing for something exploding 500 feet up, and it just washes out whoever’s standing next to you.

• Lock your focus and exposure. Tap and hold on the sky until it locks, then slide the brightness down a notch so the bursts pop instead of blowing out to a white blob.

• Hold still — or better, prop it up. Steady beats fancy every time. A cheap mini tripod or a nearby railing turns a smear into a crisp streak of light.

I walk through all five setup steps — including the long-exposure trick that gets those gorgeous light trails — right HERE. Do it before the first boom, not during it.

🚨 Scam Season Has a Long Weekend, Too

Holidays are prime time for scammers. Everyone’s distracted, traveling, and helping older relatives with their phones — which is exactly when the bad texts land. Here are the three I’d warn your family about this weekend, plus a free tool that settles “is this real?” in seconds.

1. “Switch your texting app NOW” texts (Samsung people, this is for you). Samsung is retiring Samsung Messages in the U.S. this month and moving Galaxy users to Google Messages. The switch itself is fine — but scammers love a real transition, because it’s easy to fake an urgent “act now or lose your texts” message. Don’t tap links in a text about this. I put the safe, do-it-yourself steps in the Tip of the Day below. 👇

2. Fake CAPTCHA pop-ups. The FTC is warning about a nasty one: a website throws up a “prove you’re human” box that tells you to copy and paste something or install a little “verification” tool — and it quietly installs malware instead. New rule for the whole family: a real CAPTCHA never asks you to copy/paste a command, run anything, or install software. If a pop-up gives you instructions, close the tab.

3. The “but the grammar was perfect” problem. An AP/FRONTLINE investigation found overseas scam operations are running on AI tools and U.S.-linked internet infrastructure. Translation: that’s why Mom’s fake “bank text” suddenly reads flawlessly. The old advice — “watch for typos” — is dead. The new tell isn’t spelling. It’s urgency + a link + a request for money or a code.

The free tool I keep recommending: Scamwise.com. Get a weird toll bill, “package problem,” or bank alert? Paste it in, and it tells you in seconds if it’s a known scam. Bookmark it, then text the link to the person in your family who forwards you everything asking “is this real??” It is, and it’s awesome. I even talked about it on the Today Show recently.

📲 Techish Tip

Your Samsung Texting App May Be Going Away. Fix It Before Your Messages Get Weird.

Who this is for: Samsung Galaxy owners still using Samsung Messages — especially on older Galaxy phones.

Why now: Samsung is ending Samsung Messages in the U.S. this month, and Google Messages is the replacement. Doing this yourself (instead of tapping a text) means no scammer gets to “help.”

TL;DR: Install Google Messages, make it your default, turn on spam protection. Five minutes, done.

Samsung / Android steps:

1. Open the Play Store.

2. Search Google Messages → Install or Update.

3. Open it and tap Set default SMS app.

4. Confirm your old conversations show up.

5. Turn on spam protection: Google Messages → your profile photo → Messages settings → Spam protection.

On an iPhone? Nothing to switch — but it’s a good moment to turn on: Settings → Messages → Filter Unknown Senders.

BONUS: 📲 Road-tripping this weekend? If reading your phone in the car makes you queasy, there's an iPhone setting that helps — [here's how to turn it on]. (Android folks, I've got you covered too.)

That’s a wrap for the Fourth. Go be with your people. 🎆

Not a paid subscriber yet? That’s where the deep dives, the full gadget testing, and the podcast extras live. Subscribe HERE.

Have a safe, happy, sunburn-free Fourth.

More soon — XO, Jenn

Here’s where to switch it on:

Settings → Accessibility → Motion → Vehicle Motion Cues → Automatic

Once it’s on, a scatter of animated dots appears along the edges of your screen and drifts with the motion of the car — leaning when you turn, sliding when you brake. That’s the whole trick. Motion sickness kicks in when your eyes (locked on a still screen) and your inner ear (feeling every curve and bump) disagree about whether you’re moving. The dots close that gap, giving your eyes the movement your body already feels. Set to Automatic, it turns on by itself when your phone senses you’re in a moving vehicle, then disappears once you stop.

It works on iPad, too. And if you’d rather flip it on and off yourself, you can add a one-tap toggle to your Control Center.

On Android? There’s no built-in version yet. Google is reportedly building one — codenamed Motion Assist, likely arriving with Android 17 — but nothing’s official until it ships. In the meantime, a free app called KineStop does the same dots-on-the-edges trick as a screen overlay. It’s the closest Android equivalent, and it’s quietly been doing this since 2018, years before Apple’s version showed up.

One obvious rule: this is for passengers, not drivers. Never use it behind the wheel. But if your kid goes green five minutes into a trip and still won’t put the screen down, it’s absolutely worth a shot before the next “I think I’m gonna throw up” from the back seat.

Got a question or a tip? Be sure to reach out!

Leave a comment

Here's a list of fixes — 12 gadgets that tackle the biggest common problems of summer one at a time. The best part? I’ve tried every single one of these myself, and they all work exactly as described.

***THIS IS NOT a paid post or partner endorsement. We independently review everything we recommend. When you buy through our links, we may earn a small commission. Please consider becoming a paid subscriber to support this work.***

1. Thermacell E55 Rechargeable

Mosquitos LOVE me. I’m a blood-sucking magnet, and there are few things I hate more in life than leaving a fun outing covered in red welts. This little gadget saves me all the time. Press the button, and it mists out a roughly 20-foot bubble of mosquito protection — no spray, lotion, scent — leaving nothing on your skin (including bites). One charge covers about five and a half hours, enough for dinner and drinks.

The catch: it takes 15 to 20 minutes to build the zone, so switch it on before people show up, not after the first bite. And it’s mosquitoes only. Flies and gnats don’t care.

Price: $43.98

BUY NOW

2. LightSmart Sun Shirt

This UPF 50+ sun-layer uses a fiber that claims to convert UV light into red and near-infrared light — wearable red-light therapy, basically, powered by you standing in the sun.

LightSmart cites general research on red-light therapy for recovery, skin, and hair, and says the fabric runs cooler than a normal shirt because it converts UV rather than just blocking it.

The caveat? Keep a clear head about what you’re buying. The UPF 50+ protection is real and simple. The red-light “cellular energy” benefits are the company’s own claims, and that science is newer and less proven, so buy it for the sun protection and the cooling, and treat the therapy angle as an interesting maybe.

Price: $98

BUY NOW

3. Torras Coolify Neck Cooler

This is a little AC unit you wear around your neck. Cooling plates sit against your skin and knock the temperature down about 10°F almost instantly, with a fan handling the rest of you. The top-end Coolify runs just under $300; the simpler ones start around $30.

Not ready to spend? A $15 soak-and-wring cooling towel does a rougher version of the same trick with no battery to charge.

Price: $299

BUY NOW

4. Dyson HushJet Mini Cool

I love Dyson’s first handheld fan, with the bladeless airflow tech shrunk down to something you can hold, wear, or stand on a desk. It sports five speeds, a Boost mode, and lasts up to six hours on a single charge.

Most handheld fans barely move your hair. This one actually cools you — at an outdoor wedding, in a security line, or in the middle of a hot flash. Genius.

Price: $99.95

BUY NOW

5. H2O Audio RIPT-Lite Earbuds

These are feather-light — about 3.7 grams each — so they disappear in your ears and stay put through a run, mow, or long walk. You get around five hours per charge, plus thirty more in the case, and the case shows an actual battery percentage instead of making you decode a blinking light.

One thing to be straight about: these are the sweatproof RIPT-Lites ($80) (IPX5), not the swim-in-the-ocean ones. Sweat and rain, no problem. Dunking them, no.

If you actually want to swim with music, that’s a different H2O product — the Sonar 2 PRO ($150) is the swimmer’s pick, and the Tri 2 ($150) does double duty for swim, bike, and run. Both are IPX8, good to 12 feet down, and open-ear bone conduction, so they ride on your cheekbones instead of plugging your ears.

Price: From $79.98

BUY NOW

6. Anker Solix EverFrost 2

When we still had our Sprinter Van, this was what we used in place of a mini fridge. When we sold our van, this came home with us, and we use it all the time for camping and backyard barbecues. It never needs ice, drops to 32°F in about 15 minutes, holds as cold as -4°F, and one battery keeps it running for up to two days. It also makes life on the go easier with wheels, a handle, and a battery that’ll charge your phone too.

It’s not cheap, so it earns its keep only if you’re on the boat or tailgating every weekend. But you’ll never fish a soggy sandwich out of melted ice again.

Price: $749.99

BUY NOW

7. JBL Clip 5

Waterproof, dustproof, and it clips onto a backpack, a beach chair, whatever’s nearby. Built for pool days, campsites, and sand. I have the blue and red-topped one, and people compliment me on it ALL THE TIME.

The thing that actually kills these speakers isn’t water — it’s sand in the charging port. Rinse it with fresh water after the beach and let the port dry all the way before you plug it in.

Price: $79.95

BUY NOW

8. Shibumi Shade

This is a four-pound beach canopy that runs on wind. Instead of wrestling an umbrella every time the breeze picks up, the fabric catches that same breeze and floats overhead, throwing up to 150 square feet of UPF 50+ shade.

It covers a whole family without the weight or the fifteen-minute tent battle. The newest version is quieter in the wind and adds a Wind Assist accessory for calm days. Check your beach’s rules first — some don’t allow big canopies.

Price: $305

BUY NOW

9. Twelve South PowerClip

Nothing wears on your nerves as much as a dead battery in the heat of the day…or night. This 2,000mAh emergency battery is about the size of a floss container that clips right onto your keys or bag. It won’t fully charge your phone, and it’s not trying to — this is built for the 1% moment. Figure 25 to 30% back on an iPhone, or a full top-up for your AirPods. Enough to call a ride, pull up a boarding pass, or find your way home.

The smart part is that both USB-C cables are built in, so there’s nothing to dig for and nothing to leave behind. It refills itself in under two hours, and the zinc carabiner means it actually lives with you instead of dying in a drawer at home.

Price: $40

BUY NOW

10. Wireless Meat Thermometer

Stick the probe into the chicken, walk away, and your phone will buzz when it hits temp. The MEATER Pro ($99.95) is the slick pick; the ThermoPro TempSpike does the same job closer to $30. Either way, you stay in the conversation instead of hovering over the grill.

One tip that saves dinner: pull the meat a few degrees early. It keeps cooking after it comes off the heat — another 5 to 10°F while it rests.

Price: $99.95

MEATER Pro BUY NOW

Price: $45.99

ThermoPro TempSpike BUY NOW

11. Wearable UV Tracker

A clip that reads the actual UV hitting you and pings your phone when it’s time to reapply. SunAWear is the name people know. Log your sunscreen in the app, clip the sensor to a shirt, cap, or wristband, and let it track against your skin type.

Cheaper route: SpotMyUV stickers, about $15 for a dozen. Stick one on, sunscreen over it, and it changes color when your coverage starts to fade.

Price: $16.48

BUY NOW

12. Waterproof Phone Pouch + Floating Wrist Strap

The Pelican Two-Pack Marine protects your gear from salt spray, sand, and sunscreen-slick hands, with a pouch that keeps your smartphone sealed yet still usable. The floating strap does exactly what it sounds like — drop the phone, and it floats instead of sinking.

Add a sand-free mat to round things out. Blankets hold sand; mats let it fall through.

Price: $27.19 for the two-pack

BUY NOW

***All prices are correct at the time of publication, but might change over time.***

Leave a comment

You know the ones — fuzzy orange smears that look like flashbacks from that time you ate those weird mushrooms in college. And there are forty-seven of them.

Here’s the thing: Your phone can capture stunning fireworks shots — that actually look like fireworks — and give people a sense of the same awe you have watching them live. That camera in your pocket is better at night photography than the $2,000 DSLR you almost bought in 2015. You just have to stop letting it think for itself.

Do these five things before the show starts Saturday night. The whole setup takes two minutes.

1. Wipe your lens:

This is the least sexy tip on the list, but potentially pays off the most. Your camera lens has been marinating in sunscreen, pocket lint, and potato salad fingers all day. Fireworks are bright points of light against a dark sky, and a smudged lens turns every burst into a hazy glow-blob.

Grab that microfiber cloth that came with your last pair of reading glasses. Don’t have that? A soft cotton shirt will do. Wipe the lens. You just leveled up in two seconds. #yourewelcome

2. Turn off your flash:

Your flash reaches about eight feet. The fireworks are a quarter mile away. All your flash does is light up the back of some guy’s head and ruin the shot.

Open your camera, tap the lightning bolt, tap Off. While you’re in there, don’t pinch to zoom. Pinching doesn’t get you closer; it just crops and mushes the image. Shoot wide at 1x and crop later. A sharp crop beats a zoomed blur every time, and the crowd silhouettes at the bottom of the frame usually make the shot better anyway.

There’s one exception: the zoom buttons. If you have an iPhone Pro, a Pixel 10, or a Galaxy S25, tapping 3x, 4x, or 5x switches to a real telephoto lens. That’s actual glass, so use it. Tapping is fine. Pinching to some random 2.7x is how good photos go bad.

3. DO THIS! Live Photo, then Long Exposure:

This is the one that makes your friends and family think you’re an absolute phonographic genius (get it, phone-ographic…).

Make sure Live Photos is on (the circles icon at the top of your camera). Don’t chase rockets around the sky. After a few bursts, you’ll see where they keep exploding. Frame that patch and let the fireworks come to you. Snap right as one opens up. Then go into your Photos app, tap the Live badge on that photo, and choose Long Exposure.

Your phone blends that little three-second clip into one frame with the silky light trails you see in pro shots. The photo was already sitting in your camera roll. Most people never find the button.

Did you tap too late? A Live Photo starts recording 1.5 seconds BEFORE you press the shutter. Open the photo, tap Edit, slide back to the frame you missed, and set it as the Key Photo. Your phone already got the shot. Go find it.

While you’re in there, try Loop too. It turns the burst into a three-second replay of what you actually saw, and it usually beats the still snap.

Remember to turn Live Photos off after the show. Every Live Photo is a tiny video, and forty-seven tiny videos eat real storage.

There’s one catch: Live Photos and Night mode won’t run at the same time. For fireworks, pick Live Photos. Night mode solves a lighting problem, and fireworks give you a timing problem. The sky is dark, but the fireworks are one of the brightest things your camera will see all year.

4. Android has the same magic. It’s just hiding:

Pixel: Open your camera and swipe to Long Exposure (it lives under Motion mode on most models). Point at the sky, tap when the firework blooms, and the phone does the light-trail math for you. Pixels crush this.

Samsung Galaxy: Fire off Single Take during a big burst and the phone hands you a whole menu of shots and clips to pick from. If you want more control, open Pro mode, set the ISO around 100, stretch the shutter to a few seconds, and prop the phone on something solid.

Any Android: Tap and hold on the sky to lock your focus, then drag the brightness slider DOWN a touch. Phones panic in the dark and brighten everything, which washes out the colors. A darker sky gives you punchier fireworks.

That focus-lock move works on iPhone too. Press and hold until you see AE/AF Lock, then nudge the little sun slider down.

5. When in doubt, shoot video and steal the photos later:

Fireworks move faster than your thumbs, so hedge your bets a bit.

Roll 4K video of the finale. On iPhone, tap the white shutter button WHILE recording to grab stills at the same time. Most Androids can do the same, or you can screenshot the best frames from the video afterward. This means one take, potentially dozens of usable shots.

If you have an iPhone 17 Pro, try 4K slow motion on a single big burst. Watching a firework bloom at quarter speed is that one clip that makes for a more spectacular social media post — and more brag-worthy share in your group chat.

Bonus Pro Tip: Steady the phone:

You don't need a tripod. You need anything that isn't your hands: a cooler, a fence post, a car roof. Lean your elbows on something and hold your breath when you tap. At night, your camera holds the shutter open longer, and even your breathing shows up as blur.

One more thing:

Get your shots in the first ten minutes, then put the phone down. The finale looks better with your actual eyes, and you already have proof it happened.

Happy 250th, America. Go make your camera roll earn it.

The 30-second cheat sheet:

• Wipe the lens.

• Flash off. No zoom.

• Live Photos on (iPhone) or Long Exposure mode (Pixel).

• Lock focus on the sky and slide exposure down.

• Shoot video of the finale and pull stills later.

• Prop your phone on something solid.

• Watch the fireworks, not the screen.

Do you have some awesome fireworks photos? SHARE THEM WITH US and tell us exactly how you captured them! And, as always, reach out with questions any time!

Leave a comment

I’ve covered, tracked, and talked about this shopping “holiday” for more than a decade — this was year 11 — and for the first time, I didn’t rush to buy anything.

That makes me the outlier.

Americans spent $26.4 billion online during the four-day Prime Day shopping window, according to Adobe Analytics. That’s a record. It also means a lot of people bought things fast, late at night, on their phones, while staring at countdown clocks and “limited-time” banners.

If you stocked up on detergent, diapers, toothpaste, sunscreen, chargers, backpacks, luggage, trash bags, or all the other boring things that somehow cost $400 every time you turn around, fantastic.

If you finally bought the noise-canceling headphones, Kindle, robot vacuum, smartwatch, tablet or portable charger you had been watching for months, also great.

But if a box shows up this week and your first thought is, why did I order this? — do not shove it into the closet and pretend it never happened. Return it. A lot of people will.

About 1 in 5 retailers said people return more than half of their Prime Day sales. Some of that is just how online shopping works now. The sale creates urgency. The box creates reality.

If you have buyers remorse, here’s how to take action:

Open Amazon and go to Your Orders.

Look at every Prime Day purchase and ask: Would I buy this again today?

If the answer is no, start the return.

If the answer is maybe, check whether it was actually a deal.

1. Check the real price history

On some Amazon product pages, you’ll see a Price History link near the price. Tap it.

If you do not see it, look for the Alexa for Shopping section on the product page, tap Ask Something Else, and ask:

What were the two lowest prices for this item over the past six months, and how does today’s price compare?

If Amazon does not show enough price history, try CamelCamelCamel or Keepa. Both track Amazon prices over time.

Ignore the dramatic “list price.” What matters is whether the item usually sells for less.

A $79 gadget marked down from $129 is not automatically a deal. If it normally sells for $79, you bought it at the regular price with a bait-and-switch sign on it.

2. Start the return before you talk yourself out of it

Go to Your Orders → Return or Replace Items.

Pick the item. Choose the reason. Then read the return instructions for that specific order.

Amazon says you can return most items within 30 days of delivery, but “most” is not “all.” Some items have shorter windows. Apple-brand products sold in new condition often have a 15-day return window. Some Amazon Renewed products give you 90 days. Some Renewed Premium products give you up to a year. Baby registry and wedding registry items can have even longer windows.

The only date that matters is the one Amazon shows you for your order.

Some returns are easy: no box, no label, just a QR code and a drop-off at Whole Foods, The UPS Store, Kohl’s, Staples, FedEx Office, or another location Amazon offers you.

Other returns need packaging. Some need the original box. Some need a printed label. Some third-party sellers have different rules. Electronics and expensive items can come with extra steps. Some things are final sale.

Read the return screen before you throw away the packaging.

3. Cancel anything that hasn’t shipped

If the item has not shipped yet, cancel it instead.

Go to Your Orders, tap the order, and look for Cancel Items.

That saves you the return errand and one more box on the porch.

4. If you can’t return it, get it out of your house

Sell it. Regift it. Donate it. Put it in a Buy Nothing group. Give it to a teacher, a college kid, a neighbor, a shelter, a family member, or someone who will actually use it.

Don’t let one bad click become permanent clutter.

5. Check for sneaky charges

Go to Account → Memberships & Subscriptions and look for Prime Video add-ons, Kindle Unlimited, Music Unlimited, Audible, app subscriptions, protection plans, warranties, Subscribe & Save orders, and free trials you clicked because they looked harmless at the time.

Cancel what you do not want now.

Many trials let you cancel right away and keep access until the trial ends. You will not remember in 29 days. None of us will.

My Prime Day rule: keep the things you needed before the sale, keep the things you researched before the sale, and return anything you bought because a yellow badge made it feel urgent.

A deal only counts if you still want it after the clock disappears.

Amazon return policy: Amazon returns
No-box, no-label returns: Amazon return drop-off options

What did you buy this year that you’re happy about? What’s the one thing you bought that you still regret (on Amazon or anywhere)??? What’s the most helpful thing you’ve ever picked up on sale?

Leave a comment