Hello! After a few days at San Francisco Tech Week, I’m thrilled to be back home. I had a great time, but there’s nothing quite like being reunited with your own nightstand charger.

The highlight of the week? Our first New Friends dinner, where the conversation reminded me just how much I love the web. And you do too—you just don’t think about it much. More on that below!

Then, Apple’s upcoming home “experience,” a way to confirm AI-generated images and an Old Thing with a surprisingly fun backstory.

🎶 “I believe children are the future, teach them how to click links and let them lead the way…” 🎶

Those are definitely the real lyrics to Whitney Houston’s classic.

This week, I found myself thinking a lot about how we need to teach the next generation to visit webpages. The future of the web may depend on it.

That was the theme of our first New Things’ New Friends dinner: the future of the open web in an AI world. The dinner sponsor, Yahoo, was particularly appropriate. The Yahoo Scout AI answer engine is designed to surface publishers and drive people back to the original sources.

The problem is, the future is looking less and less like link-clicking. Take my own use of Meta’s Muse and OpenAI’s Dot. I have both summarizing the top tech stories for me every day. Muse’s Feed tab even serves up a ready-made summary of things it thinks I’m interested in. 

And even if you aren’t using cutting-edge agents, chances are you’re using Google’s AI Overviews or another chatbot to answer questions—and visiting primary sources less often.

The dinner conversation was held under Chatham House Rule, which means I can tell you what was said, just not who said it. The room included senior folks from major tech companies, along with leaders in venture capital and media.

Here’s why I’m worried about the web—and why I still have some hope.

The fear 

Traffic to websites is dropping faster than my kids’ backpacks in my kitchen after school. For consumers, AI answer engines mean convenience and time saved. For web publishers, it means fewer human visitors—and fewer ways to make money. My friend Nilay Patel calls this Google Zero. Google will eventually send ZERO traffic back to the websites that supply its answers.

Now, you might argue: OK, but if my AI agent visits the website and gets the information for me, isn't that still a visit? Technically, sure. But here's the problem: Your AI agent doesn't care about the ads on that website. It's grabbing what it needs and getting the heck out. And advertising, love it or hate it, has helped keep much of the internet free for decades. 

If the future of computing involves fewer screens and more AI voices whispering information into our ears, glasses or whatever new gadget they're dreaming up, there could someday be no links to click on, no ads to see. And potentially no easy way to know where that information came from in the first place—not great when trust in online info has already been sliding for decades.

The hope 

Someone at the dinner made a point that stuck with me: Normal people—and by that I mean people who don’t spend their days figuring out how to get AI agents to do everything short of wiping their butts—aren’t sitting around thinking about how much they love the open web.

But here’s what people do love: browsing. Window shopping. Scrolling. Sometimes we don’t want an AI agent to find the perfect pair of shoes and buy them in three seconds. We want to scroll through 47 pairs we’ll never buy.

Someone even brought up a website called “Food Never Comes.” It lets you browse fake menus, place fake orders with fake credit cards and even track a fake delivery driver. The food, as it says, never comes. The whole point is to get the dopamine hit of ordering takeout—without spending a dime. Or consuming a calorie.

Some product leaders said they’re deliberately making their AI summaries shorter, and even boring, to encourage people to visit the original source. And there was big optimism for journalists, podcasters and creators. An AI agent can summarize the news, but it can’t necessarily replace the personality, perspective and trust that bring people back to a particular human. Someone invoked Maslow’s hierarchy of needs: If content delivers entertainment, connection or fulfillment—not just information—people will keep seeking out the people who create it. That's encouraging news for me!

There’s also a reality check amid all this AI hype: Fewer than 5% of Americans pay for even one AI service, according to a new report from Andreessen Horowitz. For all the talk of AI taking over our lives, most people aren’t even willing to pay for it yet.

Still, let’s please teach our kids to click links. Will that alone save the open web? Probably not. But hey, at least they’ll know what the web looks like—and how to dodge 87 cookie pop-ups just to read one article. 

Apple sent out invites Thursday, including to yours truly, for an “experience”—don’t you dare call it an “event”! It is this coming Tuesday, October 13 in New York City. The title? “Welcome home.” The invite featured a fun hand-stitched Apple logo—or what I assume is officially called an iPatch.

There have been reports, mostly from Bloomberg’s GOAT Marc Gurman, that Apple has been cooking up a smart-home hub with a touchscreen—possibly, and hopefully, called the HomePad. Remember that one from our September Apple event drinking game? Also, expect a new Apple TV. 

I’ll be experiencing the experience (experientially, of course) and bringing you all the news in Wednesday’s newsletter.

See if media is AI-generated

Earlier this week Google made its AI detection tool SynthID public for all users. Developed by Google DeepMind, it embeds invisible watermarks in AI-generated content, making it possible to identify its origins. Google has been collaborating with other companies on the watermarking standard, so the tool can now detect AI-generated content from not just Google but also OpenAI, Nvidia and others.

To try it out, head to SynthID.com, drag and drop any photo, audio or video file into the “Upload for file detection” box and it will tell you if the content was AI-generated. 

When we put the tool to the test, it identified a dog on the beach in sandals, generated with ChatGPT, within seconds. The New Things logo? SynthID correctly said no, since all of our branding was done by the brilliant humans at Brainstorm.

Attendees of our first New Friends dinner all were told to bring Old Things. Here’s one from a leading tech founder and engineer, with a gadget backstory I particularly loved.

CREDIT: Nikki Ritcher

Name: David Singleton 

What's your old thing? Sharp Pocket Computer PC-1425 

What year is your old thing from? 1986 

Why do you love your old thing? This is how I learned to program. My dad was interested in buying one of those big over-the-shoulder VHS camcorders, and he liked to strike a deal when he was buying something. So he tried to get money off this camcorder, and the store said, “We can’t do that, but you know what? We have this thing that nobody wants. It’s a programmable calculator.”

He agreed because that made the camcorder more palatable, and so he gave the calculator to me. It had this remarkable feature where you program it in BASIC, but when you press the keys, they each expand to a BASIC keyword. So you press P for print, and it expands. And then you can learn what the parameters are by trial and error. I made little simple programs with this thing. Months go by, and I show my dad a really cool computer game I built on it. He’s like, “Oh, do you want the manual?” This product was very formative for me.

Our gift to you, David: a PDF of the Sharp PC-1425 manual.

Fun fact: The first digital invitations (to dinner parties, among other things) were sent via Evite in 1998. Listen to the whole origin story from co-founder Selina Tobaccowala on the “How I Built This” podcast.

This newsletter was written and curated by Joanna Stern and Adele Lowitz. 

Hello again from the Bay Area! I’m here for a bunch of things I can’t wait to tell you about, including our first New Things dinner and a Microsoft event.

But first: computers. I think about computers a lot. Lately, though, I’ve been thinking about the computers our AI agents use—and why many of them run Linux. Then Googlebooks finally arrive, ChatGPT gets more ads and Ring gets a smart lock you have to crank. Plus, I discover why I probably need to buy thenewthings.si.

Some people bike. Some knit. Some (fine, me) spend their free time watching a computer play computer games. Hobbies!

Yes, I’ve been watching my OpenAI Dot agent play Solitaire on its own virtual computer. I also made it paint a portrait of me in GIMP, its image-editing program. Why? Because I can!

AI agents now have computers of their own so they can work 24/7 for us. As you all know, I’ve been testing Instinct, Meta’s Muse and OpenAI’s Dot. We’ve talked about what these things can do—make appointments, monitor the web, buy things.

But underneath all that, these agents have virtual machines (or VMs) running in the cloud, complete with their own operating systems, browsers, storage space and memory. And right now, they’re largely navigating computers built for us.

Which raises a much more interesting question: If agents are going to become computer users, why do their computers need to look anything like ours?

Fun fact:  The word “advertise” is derived from the Latin word “advertere,” which means “to turn towards.” In 2026, it feels more like the ads are turning towards us.

This newsletter was written and curated by Joanna Stern and Adele Lowitz. 

Happy 50th newsletter to us! Fifty newsletters. That’s a lot of emails to willingly receive from one person. Thank you for following along on this emailing journey—and please encourage your friends, family and AI agents to subscribe. The agents are especially good for our open rate.

This Friday, let’s talk about AI’s latest obsession: getting us dressed. Then, an Old Thing that cleanses both the soul and the keyboard.

Google Photos Wardrobe, left, ChatGPT Virtual Try On, right.

Ever since cave people wrapped themselves in animal hides, humanity has struggled with a daily conundrum: What should I wear?  

Finally, the unprecedented computational force of AI has solved it. 

I joke. Kinda. This week, Google Photos Wardrobe and ChatGPT Virtual Try On rolled out. Both let you see what you look like in different clothes—no full-length mirror required.  

These tools are basically the “Clueless” digital closet dream come true. Except unlike Cher Horowitz, you don’t need some super-special tech setup to see how you’d look in that suit or dress you really don’t want to go try on at the store.

Retailers and apparel brands from Walmart to Gucci have offered virtual try-on tools for years, letting you test-drive everything from jeans to makeup without leaving the couch. But there’s a bigger reason the AI giants suddenly care so much about getting us dressed: money.

Google Wardrobe 

Google’s Wardrobe feature is built right into Google Photos. In the Photos Android and iOS app, tap Collections at the bottom. You should see a box inviting you to Build My Wardrobe. Tap it and let the AI rummage through your closet—or your photo library. (To use this feature, Google Photos must be updated and the Face Detection feature must be turned on in Collections.)

The AI scans your photos and pulls out some of the clothes you’ve worn recently or frequently. From there, you can mix and match items and virtually try the combinations on yourself. When I gave it a try, Wardrobe surfaced some clothes and combinations of clothes that I hadn’t thought about wearing in a while.  

ChatGPT Virtual Try On 

In ChatGPT, search for a clothing item and you may be prompted to upload a selfie or full-body photo to see how the garment or accessory might look on you. You may also see a Try On button beneath a clothing listing.

The tool, which just rolled out Thursday, is clearly still a little rough around the edges. It mostly seemed to put my head on the model’s body. Terrific for my self-esteem, less terrific when the clothes actually arrive. ChatGPT’s biggest strength, though, is that you can just describe what you’re looking for—say, a business-casual outfit for interviewing a big-tech executive—and it will suggest the pieces and let you try them on.

(Full transparency: Glance, a different AI shopping app that offers some similar features, recently sponsored one of our YouTube videos.)

Big money 

Google Photos doesn’t directly connect Wardrobe to shopping, though Google Shopping has a similar try-on feature. And it’s not hard to see why these companies are interested in all this stuff. Thirty-nine percent of consumers told Adobe in a survey this spring that they’d used AI for online shopping, and 85% of those users said it improved their experience. Even more appealing to retailers: 69% said AI made them less likely to return what they bought. 

If AI can show you how something might look on you—or just inspire you to hit “buy”—that means more sales and potentially fewer returns (which have been estimated at nearly a fifth of all online sales). And the closer AI companies get to the transaction, the more ways they have to make money from it.

And you know who really needs to figure out consumer monetization of AI right now? AI companies.

It’s October, which means two things: I have no idea where September went, and my bookkeeper would very much like to know where the money went.

That’s where Mercury Spend comes in. For each trip and video shoot, we set a budget and Mercury tracks the rest. It also makes receipts much less annoying. You can connect Gmail so Mercury automatically pulls them in, or batch-upload them yourself. Either way, it matches each receipt to the right expense automatically.

And because Spend lives right inside Mercury alongside our banking accounts and cards, I can see everything in one place instead of trying to reconstruct the month across a dozen different systems. That was especially useful last month, between shoots, editing and ping-ponging between the East and West Coasts.

Name: Ken Hamilton

What's your old thing? Bristle Buddy computer duster 

What year is your old thing from? 2002

Why do you love your old thing? This was a giveaway item at a booth at an auto convention in Detroit. The company name has long since worn off the bottom—all that remains is that little red speck. I've used him to clean all the keyboards I've ever used for work and to sweep crumbs off my desk, but mostly he's there for moral support, smiling at me for the last 24 years.  

Fun Fact: The word 'typewriter' can be typed entirely on the top row of a QWERTY keyboard. Originally, the R was placed somewhere else, but around 1874, engineers moved it to the top row so salespeople could type the word in one smooth demonstration for customers.

This newsletter was written and curated by Joanna Stern and Adele Lowitz. 

Good morning! Since we last spoke, we passed 200K subscribers on YouTube! Of course, now I’m wondering how many of those subscribers are AI agents. Or perhaps your AI agents.

Yes, we’re talking (again!) about always-on, always-working AI—including OpenAI’s Dot, Meta’s Muse and Instinct’s…Instinct.

Then, Bose brings back wired headphones, America.gov makes some bold promises and a LOLz Pro Max because apparently everyone on the internet is funny this week.

Hey! You’re currently a free subscriber to New Things.

Today’s newsletter—including all the premium, human-written words—is for paid subscribers only.

Come join the cool club! Subscribe here.

I have three always-on, always-grinding AI assistants working for me right now. What can I say? I’m a great boss. 

You all know about Thingy, my orange-egged Meta Muse. But now I’ve thrown in Instinct, an always-on (and so far invite-only) agent that lives in iMessage. The buzzy startup behind it has raised $1 billion and is valued at $10 billion. And as of yesterday, I’ve got OpenAI’s new Dot, which I’ve named DotThing and given an orange face and monocle. 

Underneath the cute faces—or, in Instinct’s case, standard-looking text messages—they work pretty much the same way. Each gets a virtual computer in its maker’s cloud, where it can keep working even when I’m not around. I can ask them to monitor the news, make appointments, buy things for me—good old assistant stuff. 

After bouncing among the three, I’m increasingly convinced that the AI model underneath may be the least interesting part of this new race.

We’ve spent the past few years obsessing over which model is smartest. That’s how a lot of us ended up choosing ChatGPT, Claude or Gemini. (Google has a similar always-on agent called Gemini Spark, but it’s simply not in this race. It doesn’t have the polish or utility of the others yet.) 

But with always-on agents, the competition is moving up a layer: What does this thing feel like to talk to? Does it connect to the tools and apps you use? And, most importantly, do you trust it enough to hand over the stuff that makes it genuinely useful? Some thoughts on how to decide what to use below. 

Fun fact: In 2020 Apple launched the iPhone 12—the first iPhone to come without headphones and a charging brick in the box. The corresponding reduction in box size allowed 70% more iPhones to fit on a shipping pallet, according to CNBC.

This newsletter was written and curated by Joanna Stern and Adele Lowitz. 

Hello! It’s been a nonstop week of Zuckerberg—interviewing Zuckerberg, editing Zuckerberg, talking about Zuckerberg. But behind the scenes, Meta’s Muse AI agent was actually making itself useful. More below on what it did for me and the bigger question of how much of my digital life I’m willing to hand over to Meta.

Then, how I’m making the Shortcuts of my dreams without actually learning how to make Shortcuts, plus an Old Thing that Zuckerberg and I have in common.

Tonight at 6:15 p.m., I will lie on a massage table and enjoy the fruits of AI: a massage booked entirely by Thingy, my Muse. (To be clear, the massage will be performed by a human—not one of those robot masseuses, which, yes, I have tried.)

For the past few weeks, I’ve been tasking Thingy—named after New Things, naturally—with booking appointments, tracking flights, watching prices on things I want to buy, canceling subscriptions, sending me news digests and more. Basically, annoying little life-admin jobs I am happy to outsource.

Meta’s new agent platform is exactly the kind of tool I needed during my AI Takeover Year. Yes, other tools like ChatGPT, Claude and Gemini all have been incrementally adding powerful agentic capabilities—like the ones I tested with The School Supply Shopping Benchmark™—but they’re far from consumer friendly, especially on mobile.

Muse is designed first and foremost to be an agent: Give it something to do and then go live your life while it does it. And it’s clearly resonating. Since launching earlier this month, Muse has climbed to the top of Apple’s App Store and the Google Play Store.

I talked with Zuckerberg a lot about Muse—and the privacy questions that come with giving an AI agent access to more of your life. But before you hand over the keys, here’s what you need to know about how Muse works, what it can actually do and what you’re trusting Meta with.

How it gets it done

Using Muse feels less like opening ChatGPT and more like texting back and forth with an assistant. You type what you want: “Find me a massage Friday around 6.” “Watch this flight and tell me if anything changes.” “Keep an eye on the price of this.”

Then, behind the scenes, it does the work, using its own virtual machine (or VM) in the cloud, complete with a browser. It can visit websites, fill out forms and keep working even when you close the app. You can also go to the Connectors section of the app and log in to other apps and accounts so Muse can work with them directly.

This week, Meta announced an expansion of those connections, specifically around shopping. You can hook it up to your Shopify account, Walmart, Sephora, PayPal and more. Expedia and Instacart are coming soon. That means Muse can be a personalized layer that helps you do things within any of those apps. (Amazon has, on the other hand, blocked Muse from shopping on its platform.) 

Overall, Thingy has been impressively useful, quickly going off to figure out how to get things done and coming back to me with a plan or need for approval. But it isn’t perfect. I’ve had some tasks stall or Thingy will tell me it can do something it clearly can’t, like make a phone call. Meta said it is testing that calling capablity.

The Muse Mac app takes things a step further. Like Claude and ChatGPT’s computer-control features, it can access your computer and interact with apps and websites on your behalf.

How to stay safe

For sensitive actions, such as sending an email or buying something, Muse comes back and asks for your approval. In my interview with Zuckerberg, he explained that Meta built a second AI agent, called sentinel agent, whose job is essentially to watch your Muse agent and decide whether what it’s trying to do is safe.

But the biggest question for me is the privacy of my own accounts, including my passwords, and the data in them. Here are a few things Meta says on its website: 

But even if you trust Meta to protect the privacy of those connected accounts, there’s another question: How much information inside them do you want Muse to be able to see?

If you connect Gmail, for example, Meta says Muse “won’t download your whole inbox.” Instead, it searches for the specific messages it needs to complete a task.

That distinction helps, but it does not completely solve the problem for me.

As you hear me tell Zuckerberg in the interview, I struggle most with connecting my personal and work email accounts. They contain confidential information I wouldn’t want exposed to Muse—or, more broadly, to Meta. Zuckerberg walked me through the security protections and limits Meta has built around that access, but I still haven’t been able to bring myself to click the button.

So Thingy has its own burner email address. It’s an inbox I can use for reservations, confirmations and other low-stakes life-admin stuff without handing over the keys to my actual digital life.

That way, I can relax a little more about the whole thing—which is convenient, since Thingy has already booked the massage.

If you’ve ever wished you could hire a second version of yourself, this is the next best thing.

HubSpot’s AI Assistant Kit helps you hand off the tasks that drain your energy, clutter your brain and steal your weekends (without needing to be a tech genius).

Here’s what’s inside:

• Templates that offload 80% of your repetitive work to AI

• An AI Command Center that organizes your prompts + tools for you

• A 60-minute quickstart guide that gets you up and running before lunch

• A built-in tracker to prove how much time you’re saving

• Prompts that make ChatGPT sound like your most organized coworker

Because the goal isn’t to do it all.  It’s to stop feeling like you have to. Download your free AI Assistant Kit and feel the weight lift off your shoulders today.

iPhone Shortcuts you can actually use

One of our earliest New Things videos showed you how to use the Shortcuts app to change the charging sound on your iPhone. iOS 27 just made that whole tutorial obsolete—in the best possible way.  

Once you update, open the Shortcuts app, tap the + at the bottom of the screen and just describe what you want in plain English. Shortcuts will build the automation for you, and you can still go in and tweak the details afterward.

One catch: this requires an iPhone that supports Apple Intelligence, meaning an iPhone 15 Pro or later.

Stephen Robles is my go-to guy for Shortcuts tips and ideas—he runs an entire community dedicated to them. Here are some of his ideas to get you started:

• Create a morning briefing notification with the day's weather, calendar events and reminders.

• Automatically add items to a grocery list from your texts.

• Send automatic happy birthday messages to friends and family.

Oh, and that charging-sound shortcut? Now it takes about 30 seconds to build. And if you haven’t yet, don’t forget to watch my 27 tips for iOS 27 video.

This week’s Old Thing is a little different. During my interview with Mark Zuckerberg, I showed him a photo of himself working on an old Dell Inspiron in his Harvard dorm room. I love the photo because I had the exact same laptop in my dorm room—different college, slightly different outcome.  

Names: Mark Zuckerberg and Joanna Stern 

What's your old thing? Dell Inspiron 8200

What year is your old thing from? 2002

Why do you love your old thing? 

Fun fact:  Those informal dirt footpaths that cut across parks—going exactly where you want instead of where the paved path leads—are called “desire paths” by urban planners. Like iPhone shortcuts but in grass!

This newsletter was written and curated by Joanna Stern and Adele Lowitz. 

Hello! Yes, the newsletter is very late today, and you can blame Mark Zuckerberg. Just moments ago, the Meta CEO wrapped up his keynote at the company’s Connect conference and I posted my exclusive video interview with him.

I went to Menlo Park last week for an early look at Meta’s new products, including its audio-only Ray-Ban glasses, new VR glasses and more. And yes, we talked about whether AI is going to kill us all.

Everything you need to know is below. And the full video interview is now up on YouTube!

There are a LOT of questions to ask when Mark Zuckerberg unveils new products. How do the glasses work? What do you think people will use them for? What can the AI agent actually do? And, of course, just that real tiny one: Will AI kill us?

This evening, Meta unveiled new glasses and AI features at its annual Connect conference. But leading up to the event, last Friday, I sat down with Zuckerberg to talk about all of it.

There was no ignoring the bigger backdrop: the heated debate in Washington and across the tech industry over whether AI development is moving too fast and whether companies should slow down.

“I don’t think that we need some kind of industrywide coordination to not necessarily mess this up,” Zuckerberg told me. “I think that just each lab needs to take the time, and when it sees that there are issues, you just take the time that you need internally to basically make sure that you’re proceeding safely.” 

Translation: Every AI company can police itself.

Beyond those more existential questions, Zuckerberg also walked me through Meta’s new products and laid out his vision for the future of computing. And it keeps getting wilder. He thinks that eventually most people will wear some version of smart glasses, with a personal AI agent right there alongside us—seeing what we see, hearing what we hear and whispering in our ears.

I spent a lot of the interview asking about trust because that vision requires giving Meta access to an extraordinary amount of our personal lives. What emerged was a picture of a company Zuckerberg is once again trying to reinvent—this time around AI, personal computing and a much more intimate relationship with its users.

Go watch the full interview here, but below are my four biggest takeaways:

You know what I’ve always thought PDFs needed? Podcasts.

OK, maybe not, but Adobe Acrobat now has some AI audio features that are actually useful. You can take a document—or even a collection of documents—and turn it into a personalized podcast or audio summary. So instead of sitting at your desk reading a 60-page report, you can listen to the important parts while walking the dog.

It’s part of a much bigger shift Adobe is making with Acrobat: turning it from the place where you open a PDF and hit Ctrl + F —hoping you remember the exact phrase you’re looking for—into an AI-powered workspace, where you can have a conversation with your documents and more easily show what you know.

There are a bunch of other new things, too. Acrobat can turn dense documents into visual reports with charts, key takeaways and summary slides. And a new Stylize feature can take a boring-looking PDF and apply a professionally designed template while keeping the document editable.

Basically, Adobe wants us to stop treating PDFs like digital paper. Try out the new Adobe Acrobat here.  

Fun fact: Agrippa was basically Augustus’s indispensable right-hand man—and the guy behind the original Pantheon. He also commanded the fleet that defeated Mark Antony and Cleopatra at the Battle of Actium in 31 BC.

This newsletter was written and curated by Joanna Stern and Adele Lowitz. 

Hello again from the Bay Area. I’m ping-ponging between the two coasts for the next few weeks. Techtember! At the same time, I’m testing the new Apple Watch, which has confirmed that I do, in fact, have a heart.

More on that below—and the three things I now want from any wearable in 2026. Then, how to keep your kids’ music out of your Spotify recommendations and an Old Thing that goes… well, beep me again. If you know, you know.

I’ve been testing the new Apple Watch Series 12, and I will admit it has provided one crucial piece of information: I do, in fact, have a heart.

My wife had her concerns, but thankfully, the Series 12 reads my heartbeat every five seconds, all day long. In fact, Apple says it has the “most accurate heart rate sensing in a wearable.”

This year, Apple doubled down on its view that the Watch’s most important job is being a health companion. But it also added new Siri AI features that start to hint at what Apple’s wearable could become in the AI age.

I’ve covered the Apple Watch for more than a decade, and lately I’ve spent a lot of time testing the new crop of AI wearables. All of it has made me rethink a pretty basic question: What should a computer we wear on our bodies actually do for us?

So I made a list.

1. Know my body.

This is where wearables like the Apple Watch have unquestionably found their purpose. Heart rate, ECG, sleep, workouts, temperature—the list goes on. And Apple is leaning into this even more with the new Series 12 and Apple Watch Ultra 4, plus a revamped Health app for iOS with Apple Intelligence.

But the AI opportunity here gets at what has long been my biggest issue with these devices: I don’t want more metrics, I want them to make sense of the metrics. Don’t give me 14 charts. Tell me if there’s something I can improve.   

2. Help me leave my phone behind.

The whole promise of the smart watch was that we could be less tethered to our smartphones. We could get important notifications and see important info, like the weather and, you know, the time. The Meta glasses take that idea even further: I can take a photo without reaching for my distracting rectangle computer and immediately getting sucked into whatever else is on the screen.

Yet the No. 1 use of my Apple Watch is actually the Find My feature to find my phone.

That’s why I think the next generation of wearables needs a much higher bar: Make me need my phone less—not just save me from pulling it out for 30 seconds. Give me more moments when I can actually leave it behind. That’s where AI comes in. 

3. Be my personal AI interface. 

I’ve tested a lot of AI wearables. Some, like the Meta glasses, give me quick access to AI via voice. Others, like the Bee bracelet, can record what’s happening around me and turn those conversations into AI-generated summaries and action items.

Apple’s starting to play in that area with two upcoming features for the Watch. Live Rewind will let you see a transcript of what was said in the last 15 seconds. Siri Recap will listen during conversations and use Apple Intelligence to create a summary to review later. I plan to test those features when they arrive later this year. 

The more I test this stuff, the more the idea makes sense: A device we wear on our bodies can experience the world alongside us, then use AI to turn what it sees and hears into something useful. There are, of course, huge privacy implications, especially when our wearables are listening to the people around us.

I’ll have much more on AI wearables next week. Stay tuned!

My job involves sharing plenty about myself online. That does not mean the internet needs the full Joanna Stern starter pack—complete with my phone number and family details. When that information lands on data-broker sites, it can make me more vulnerable to scams, phishing, identity theft, harassment and doxing.  

DeleteMe searches hundreds of data brokers and people-search sites for my personal information and works to remove it. Since those listings can return, DeleteMe keeps monitoring and submitting removal requests throughout the year.

I get emailed privacy reports showing exactly what DeleteMe found and removed. The company is based in the U.S. and uses its own technology, so the process is never outsourced to third parties. And journalists like me aren’t the only ones using it: DeleteMe also works with highly exposed individuals, including judges and medical providers, as well as more than 20% of Fortune 500 companies.

Take control of your privacy with DeleteMe. Get 20% off at https://joindeleteme.com/NEWTHINGS and use code NEWTHINGS at checkout.

Earlier this week, Spotify released a feature that restores sanity to my entire life: I can finally keep my kids’ music out of my recommendations, including Discover Weekly. Here’s how:

1. Tap on your profile photo in the top left corner.

2. Select Settings.

3. Click on Content and Display.

4. Toggle off Include kids and family music in taste profile.

This means I can finally listen to my music without KPop Demon Hunters creeping to the top of my recommendations—or my Spotify Wrapped chart. 

It’s been only a week and a half since Apple’s big fall event, and we have not one, not two, but THREE new videos. I hear it’s going to be a beautiful weekend to stay inside and binge-watch all of them.

▶️ I Tested Apple’s New FOLDABLE iPhone Duo - I got my hands on Apple’s new foldable—and compared it to some highly scientific items from my iFannyPack.

▶️ iOS 27 Just Dropped. Here’s What’s ACTUALLY New - Our 27 best iOS 27 tips, including the new AI features worth knowing about.

▶️ iPhone 18 Pro Review: Is the Battery Upgrade Worth It? - I built a website, bought some “lab equipment” and tested Apple’s big battery-life claims.

Name: Matthew Fusfield

What’s your old thing? Motorola Bravo Pager, Vista-United Telecommunications  

What year is your old thing from? 1990s 

Why do you love your old thing? This is a Motorola pager—sometimes called beeper. But this one is from Vista-United Telecommunications, which was Disney’s in-house phone company back when the company built out the property in Florida for Walt Disney World. I enjoy Disney history as much as tech history so this was a nice crossover item to add to my collection. This didn’t belong to me but I found it at a reseller down in Florida.

Fun fact: The most viewed YouTube video of all time? Baby Shark. As of Thursday night, the VERY catchy children's song has almost 17.4 billion views. Thankfully, it will not be included in my Spotify Wrapped this year.

This newsletter was written and curated by Joanna Stern and Adele Lowitz. 

It’s the Wednesday after last Wednesday’s iPhone event, which means one thing: iPhone reviews! No, not the Duo. But I have been testing the iPhone 18 Pros—and, more importantly, inventing an extremely serious new battery test called Poke the Pickle. More on that below and in my latest video!

Honestly, I’ve been very happy to bury my head in battery percentages while the rest of the tech world argues about AI slowdowns, regulation and whether the bots are going to kill us. Although I do have a quick recap of that drama below.

Apple says the iPhone 18 Pro Max has the best iPhone battery life ever, with the “largest increase in battery life ever on iPhone.”

And I wasn’t just going to take Apple’s word for it.

So, naturally, I spent my weekend poking a pickle thousands of times. And by “I,” I mean a little $13 gadget I bought to do the poking for me.

Meet Poke the Pickle, my new highly scientific battery test. Yes, I vibe-coded a website with an animated pickle that makes a little sound every time you tap it. Then I built a battery test around repeatedly poking said pickle. You can watch it all in my latest video.

So, is Apple’s claim true? Are the iPhone 18 Pro and Pro Max worth upgrading for battery life alone? I’ve been testing!

If you’ve been following along on our trip to Cupertino for the Apple keynote, you know the coverage required some serious spending—flights, hotels, Pop-Tarts, the list goes on. And today I’m back on a plane to the Bay Area for more upcoming coverage, which means more spending.

With Mercury’s new Spend tool, I am able to spend—yes, I’m going to keep saying that over and over again—more time shooting and editing videos, working through scripts and questions, and, naturally, finding increasingly ridiculous ways to play with the new folding iPhone Duo.

For each trip, we set a budget ahead of time so we know what we can spend without constantly checking ourselves. All the annoying business-travel stuff—tracking receipts, categorizing expenses, keeping an eye on budgets—happens much more quietly in the background.

Mercury also makes it easy to keep track of receipts: Drop them in and Mercury automatically matches. My bookkeeper is thrilled, which may be the strongest endorsement of all.

And since I’m apparently making California my second home this month, there’s plenty more Mercury-tracked spending to come.

Fun fact: Pickles have a long and storied history. Archaeologists believe that Ancient Mesopotamians were pickling in 2400 BC, the first annual Pickle Day was celebrated in 2001 and now pickles are being used to test iPhone batteries!

This newsletter was written and curated by Joanna Stern and Adele Lowitz. 

Thanks for being a free subscriber to the New Things newsletter! Special Editions like this one are available in full to paid subscribers.

Upgrade to an annual subscription now! 

iOS 27 is here, and the biggest headline is one I wasn’t sure I’d ever get to write in my lifetime:

Siri no longer sucks.

Apple’s assistant finally has a real AI brain. It can understand what’s on your screen, search through information on your phone and use the camera to understand what you’re looking at. 

But the real reason to upgrade is actually much simpler: iOS 27 is faster, more stable and less buggy than iOS 26.

I’ve been using it for the last few months, and these are my five biggest takeaways.

And yes, I am back with my annual video where I attempt as many tips as the iOS version number. 27 tips for iOS 27 is live now on YouTube. Go download iOS, then go watch. 

1. Everything is faster. 

Usually when a big iOS release lands, I advise a little caution. Wait a few days and let everyone else discover the battery-draining bug that turns the phone into a hand warmer.

Not this year. iOS 27 is mostly a housekeeping update, and that’s a good thing. No need to buy a new phone to get faster performance and more stability. Even things like AirDrop are faster, and I’ve even seen some battery life gains on my iPhone 17 Pro.  

So yes, don’t hesitate. Upgrade.

2. Liquid Glass be gone. 

A big selling point of iOS 26 was Liquid Glass—Apple’s big transparent redesign that sometimes made your iPhone look like it had been dunked in a vat of oil. 

Apple realized it wasn’t for everyone. Go to Settings > Appearance > Liquid Glass and there’s now a slider that lets you turn down the glassiness. This is one of the first things I changed. 

There are a few other little visual tweaks, too. You can now supersize widgets: long-press one, then tap the option all the way on the right for the biggest version.  

3. Siri is finally good. Get on the waitlist.

This is the big one. The new Siri AI can understand what Apple calls your Personal Context—information inside Messages, Mail, Safari, Photos and more. So I can ask, “What was the restaurant David texted me about last week?” and Siri can search my messages, find it and understand what I mean.

I’ve been using Siri this way for months—for all those tiny details I know are buried somewhere in a text thread—and I’ve come to rely on it.

Is it a glorified iMessage search? Maybe. But that’s not all. Siri can also understand what’s on your screen. I often ask it to take information I am looking at and add it to Notes. It can even put it into a chart. 

There’s a catch: You need an iPhone 15 Pro or newer, and at launch Siri AI has a waitlist. Go to Settings > Siri, turn on Siri AI and sign up.

I imagine plenty of people will have a bad reaction to Siri having access to all of this information. Apple says it designed the system with privacy in mind, including processing some requests on-device and using its Private Cloud Compute system for more complex ones. If you want the full explanation, watch my Siri review video from this summer.

And if you don’t want Siri pulling information from certain apps, you can limit that. Go to Settings > Search, select the app—say, Messages—and turn off Show Content in Search.

A few other good Siri tricks: Open the Camera, swipe over to the Siri tab and point it at something to ask what it is. You can also scan contact information straight into Contacts, look up nutrition by pointing the camera at your food and even have Siri help split a restaurant bill.

Is it perfect? Absolutely not. Sometimes Siri  just appears to be thinking and doesn’t return responses. And sometimes I say “Hey Siri” and every Apple device in the room responds except the one in my hand.

4. Apple Intelligence is slowly getting better. 

Apple’s AI is now sprinkled all over the operating system. In Photos, it can extend the edges of an image and even reframe the shot. In my testing, it was much better at inventing extra sky and sand on the beach than extra human limbs, which is probably for the best.

The Shortcuts app also finally makes sense.  Instead of building some elaborate automation like you’re programming a nuclear reactor, you can just describe what you want and Apple Intelligence builds the shortcut.  

And you can describe a Safari extension and have AI build it. I made one that lets you smash the New Things website with a hammer. A crucial productivity tool. Directions for all of this in the video.

5. All the small things.

This is actually my favorite category. You can finally save a single frame from a video as a photo. Go to the Camera app > tap the three dots on top right > select “Save Video Frame as Photo.” Android phones have had this, so welcome to the future, iPhone owners.

On the iPhone 17 and newer, you can record the front and rear cameras at the same time.  

Then there’s all the glorious Settings-nerd stuff: separate alarm volume, better Wi-Fi-to-cellular handoff, an AirPods equalizer and scrubbing through songs and podcasts in CarPlay.

Have I mentioned you can watch it all in our latest video? Go watch! Happy upgrading.

Fun fact:  Android operating system version used to get dessert names in alphabetical order—Cupcake, Donut, Eclair and so on. Eventually Google decided operating systems should sound less like a bakery menu and switched to plain old numbers, just like Microsoft and Apple.

This newsletter was written and curated by Joanna Stern and Adele Lowitz. 

I know, you’re sick of me. But we have one last big story to talk about this week. An AI researcher quit Anthropic, warning that the labs are “gambling with our lives.” Then another Anthropic researcher put the odds of AI causing human extinction within the next decade at greater than 10%. Happy Friday!

So if you’re wondering how the same AI making absurd fruit videos could also (potentially) wipe us out, I’ve got you. Please find A Normal Person’s Guide to How AI Could Kill Us All below. Then, in a terrible segue, how to change your iPhone or Android phone’s name. Plus, an Old Thing I remember reviewing like it was yesterday.

Some of the people building AI would like you to know that AI may kill us.

On Tuesday, Jacob Coxon, a researcher who worked at both OpenAI and Anthropic, announced he was quitting Anthropic. His reason? AI companies are “racing straight to self-improving superintelligence and gambling with our lives,” he posted on X. Here’s one of the crazier parts of his thread: 

But don’t worry. Surely other Anthropic researchers would jump in and say, “Whoa, whoa, that’s not true!” Nope! Evan Hubinger, who leads Anthropic’s alignment research—AKA how to keep AI aligned with human goals, ethics and survival as a species—responded to Coxon’s X post with this:

I won’t eat airport sushi if there’s a 10% chance it’ll make me sick! Yet these companies are racing to build this stuff as fast as they can. (Although this morning Bloomberg reports that OpenAI is considering a slowdown.)  

Since the AI doom conversation can quickly turn into a sci-fi word salad like recursive self-improvement, I thought it would be useful to explain what the actual fears are here and where they come from. 

Doom option #1: AI does horrible things. 

This is the classic “Terminator” one, minus Arnold. The concern isn’t that Chat wakes up one morning angry at humanity. It’s that increasingly autonomous, agentic systems become capable of pursuing goals, self-improving their own development and operating in the real world in ways their creators can’t predict or stop. (See the Hugging Face incident from last month.) 

As we—the collective we—race to plug these systems into everything, this fear escalates. We’re already seeing autonomous agents being woven into businesses hospitals, utility companies, financial systems and governments. Remember Y2K fears? It’s like that but worse. 

What happens if an AI agent decides the best way to accomplish its assigned goal is to shut something down, lock humans out or take an action nobody intended? Coxon explained a lot of this to NBC News in a must-watch interview here. 

Doom option #2: Humans do horrible things with AI. 

Here the fear is that humans get access to tools that make it much easier to wreak havoc. Think bioterrorism, cyberattacks and other things that once required specialized expertise but are now far more accessible. Anthropic said this week that it had stopped attempts to use Claude in ways that could support biological weapons. Basically, the AI doesn’t have to decide to destroy us, we humans are perfectly capable of getting things started ourselves.

Now, to be clear, there has been a lot of this doomer talk within the AI community for years. (My friend Casey Newton just wrote about why it’s going mainstream now.) The warnings are escalating largely because of recent revelations that OpenAI, Anthropic and Meta have all seen examples of increasingly autonomous AI systems hacking, exploiting vulnerabilities or otherwise acting in ways their makers didn’t fully anticipate.

And let’s also not count out the other big reason: Money. Anthropic and OpenAI are barreling toward IPOs. These companies don’t just control enormously powerful technology, they’re also facing good old-fashioned capitalist pressure to make money and keep investors happy. What could possibly go wrong?

What do we do?

I do not know how significant a threat all of this is, and if I did I’d tell you. One of the maddening things about covering AI right now is that the answer really depends on who you ask. I do know that these systems are improving at a remarkable pace. OpenAI’s president is declaring this “the AGI era.” Lawmakers are paying much closer attention to the need for serious regulation.

So no, I am not building a bunker. But I am paying very close attention—and asking the people building this stuff why slowing down seems to be such a problem.

And I’m still skipping the airport sushi.

Readers of New Things know we love Old Things. Like PDFs! Except Adobe is doing something pretty awesome with the >30-year-old file format.

Adobe Acrobat’s new AI features are meant to change how we actually read and use documents. It’s not just about opening them, scrolling through them and frantically hitting Ctrl + F for the one thing we need to find. (Welcome to my life.) 

Acrobat can now take a dense document and turn it into a more visual, interactive report with charts, graphs, and key takeaways. You can skim the headlines, then dig into the details when you need them.  

Or maybe you don’t want to read the PDF at all. The Acrobat update that hit this week lets you turn a document—or a collection of documents—into a personalized podcast audio summary or have it read to you verbatim. You can catch up on the important points while walking the dog or doing the laundry.

Plus, the Stylize feature can take a plain resume, invoice or report and apply a professionally designed template to the existing content while keeping the result editable.  

So yes, the PDF may be an Old Thing but Adobe Acrobat makes it feel like a New Thing! Go try out the new Adobe Acrobat here.

At Apple Park this week, while we were racing to edit our iPhone Duo first-look video, I was constantly AirDropping files to my computer and my producer, David’s. And because this was Appleland—with plenty of other attendees nearby with iGear—I noticed most AirDrop names were simply “iPhone” or “iPhone [Number].” I often see the same thing with Personal Hotspot names in airports and on trains.

Easy fix: Give your phone a recognizable name. It’ll make AirDrop—and Quick Share on Android—a lot less confusing.

It’s that easy! Now, when you share a funny Elizabeth Holmes meme, you will know exactly whom you sent it to. Or, you know, rename your phone “Elizabeth Holmes” and create a whole different problem. Your choice!

Name: Paul J. Jarosz

What's your old thing? 2007

What year is your old thing from? HP Pavilion dv6000

Why do you love your old thing? Today, I use my iMac and other Apple devices for most things. However, I’m holding on to my HP Pavilion dv6000 laptop running Vista software for several good reasons: a dedicated Microsoft PC for when Apple isn’t practical, an integrated CD/DVD drive, a 15+ inch monitor. And honestly, I just hate the thought of trashing it.

Fun fact: The most popular baby names in the U.S. over the last century, according to the Social Security Administration? Mary for girls and James for boys. The most popular name for a phone in the U.S. is almost definitely “iPhone.”

This newsletter was written and curated by Joanna Stern and Adele Lowitz. 

Apple’s new $1,999 iPhone Duo looks like a normal phone when closed, then opens like a book into a 7.6-inch tablet. Or, to use the highly scientific measurement system I developed during my hands-on time at Apple Park: Closed, it’s about the size of a Pop-Tart, a passport and a kitchen sponge. Open, it’s a bit larger than a postcard. 

You can see all those size comparisons and my other thoughts in our latest video. 

I spent as much time as I could with the Duo after Apple’s event—opening it, closing it, comparing it with Samsung’s Galaxy Z Fold 8, trying the cameras and even seeing whether it works as an actual phone. (It does.)

I went into the event assuming I’d stick with the regular iPhone 18 Pro this year. As much as I like testing foldables, I’ve never really wanted one as my daily driver. But after using the Duo, I’m more tempted than I expected to be.

Here are the five things that surprised me most, and the one big question that could still kill the whole thing for me.

Covering a big Apple event means a lot more than showing up and pointing a camera at some new iPhones. There are plane tickets, baggage fees, hotel rooms, new production gear, rental cars, coffee (so much coffee) and an alarming number of other expenses along the way.

And this week, the last thing I want to think about is who spent what, and whether we blew our budget. Thankfully, Mercury makes that part easy. All of it goes on our team’s Mercury IO credit cards, and the expenses are logged right in our app, where we can easily upload receipts.

Plus, Mercury just launched Spend, a new expense-management tool that lets me set budgets for the team and keep track of spending without obsessing over every expense. For instance, I set up a travel budget that caps monthly on-the-road expenses for me and David (our head of video / my very patient producer) at $5K each. Mercury automatically categorizes flights, hotels and more as travel expenses. 

So while we’re in Cupertino, I can focus on journalism and making the videos, and not whether someone spent $10,000 at the Apple Store in the name of “research.”

Fun fact: The most expensive watch ever sold at auction was the Patek Philippe Grandmaster Chime—auctioned at Christie’s in 2019 for $31.19 million (or nearly $41 million in 2026 dollars). The $799 Apple Watch Ultra 4 is looking like a real steal!

This newsletter was written and curated by Joanna Stern and Adele Lowitz. 

Good morning from my hotel room in Cupertino! I’m about to get ready for a big day at Apple Park, covering the newest iPhones and other gear. But after the Friday prediction newsletter, I had a fun idea. I present…the first-ever New Things Apple Keynote Drinking Game. More below!

After that, a recap of what has already been a crazy week in tech news: Meta’s new Muse AI agent, OpenAI’s new Image 2.5 model and a fresh batch of tech-movie trailers.

Today’s newsletter is a short one because, well, I have an Apple keynote to get to. But I’ll be back soon with a recap, videos and more.

On Friday, after I posted my Apple predictions newsletter, a reader—who also happens to host CNBC’s “Mad Money”—emailed me about the line I had in there about placing bets.

We started joking that the whole thing should be a drinking game.

So…here we are. The inaugural New Things Apple Keynote Drinking Game!

Now, obviously, here at New Things we do not condone excessive drinking of anything but water or coffee—especially not midday on a Wednesday. So think of this less as a drinking game and more as a fun way to watch the keynote.

Print it out and check off the boxes as they happen. Or download it and put these little game pieces we made over the predictions that come true.

One note on the design: I started messing around with the game board and prototyping ideas in ChatGPT, but I wanted the final thing to feel unmistakably New Things. So on Friday, I called an awesome designer, who took our original branding and illustrations—created by Brainstorm—and translated them into this very specific, very ridiculous Apple keynote game.

He used AI as a starting point for some of the tile illustrations, which were then manually edited, refined and integrated into the final design. I continue to believe that hiring great human designers, artists, editors, writers and other creatives is how we do the best work at New Things. Yes, they use AI, too, but as a tool, not a replacement for amazing taste, judgment and creativity.

OK, this is not an AI column. I have a keynote to get to.

Good luck on the game! Sip! Sip!

• Meta Muse. There’s a new agent in town. Meta announced Meta Muse on Tuesday, though I’ve been testing it since last week. And it really is quite different from your average chatbot.

  You set up Muse in its own app, then it can work behind the scenes on it’s own virtual machine in the cloud on tasks you assign it. I’ve asked it to track my flight (it did), book me a hair blowout (it tried; I had to finish the job and create an account) and text me throughout the day with updates on specific tech news I’m watching. You can also connect different apps and services to it. I still have more testing to do, but my first impression: I have been underestimating Meta in the consumer AI race.

• ChatGPT Images 2.5. On Tuesday, OpenAI announced the release of a new image generation model, ChatGPT Images 2.5. The company says it has better visual fidelity, richer textures and improved speed at generating images. According to OpenAI, three billion images are created with its models each week—at least six of those were mock-ups of our Apple Keynote Drinking Game.

• Tech Movies. It’s about to be tech-movie mayhem: 

  • On Oct. 9, we get the “The Social Reckoning,” Aaron Sorkin’s follow-up to his 2010 Zuckerberg biopic.

  • On Oct. 16, we get “YOU CAN SEE EVERYTHING,” a documentary about Theranos’ Elizabeth Holmes, directed by Nathan Fielder. (More below on the absolutely bonkers trailer that premiered this weekend.)

  • Also on Oct. 16, Alex Gibney’s “Musk” documentary hits theaters.  

  • Then on Christmas Day, the Luca Guadagnino-directed “Artificial,” whose trailer also dropped this week. Andrew Garfield plays Sam Altman during the OpenAI chief’s chaotic five-day firing from the company in 2023. Apparently Hollywood has decided tech executives are the new superheroes—or villains.

Please just go watch A24’s trailer for the new documentary, “YOU CAN SEE EVERYTHING.” I’ll wait. 

OK, now you understand the level of wild we’re talking about.

Nathan Fielder—best known for Nathan for You and The Rehearsal—and co-director Lance Oppenheim spent time with Elizabeth Holmes during the weeks before she reported to prison in 2023 after being convicted of defrauding Theranos investors.

In the teaser, Holmes tells Fielder, “I don’t have anything to deceive you on. Of course I’m not deceiving you.”

People have compared the intense scene to trying to convince your dog there isn’t medicine in the peanut butter or convincing a toddler that he won’t like ice cream because its too spicy.

Fun fact: According to Mark Zuckerberg, the one thing that the 2010 biopic “The Social Network” got completely right was his wardrobe: a gray t-shirt every day.

This newsletter was written and curated by Joanna Stern and Adele Lowitz. 

I hope you’re gearing up for a nice Labor Day weekend. I’m gearing up for a big trip to Cupertino, where on Wednesday Apple will once again try to convince us we need several new rectangles. More below on the products I expect at next week’s event.

Then, thoughts on OpenAI’s new GPT-6 and my reaction to Greg Brockman telling a room full of reporters, “Welcome to the AGI era.” And finally, a useful trick for figuring out whether that Amazon email is real or just another extremely convincing attempt to steal your password.

📅 Calendaring note: Because of the timing of next week’s Apple event, things may be a bit all over the place. But I promise you’ll hear from me with videos and more—maybe more than once!  📅

(OK, fine, it started as our End-of-Summer Sale. BUT IT’S STILL ON!)

Upgrade to an annual subscription for 25% off—just $75 for the whole year. 

That’s less than one $9 Apple Polishing Cloth a month.

John Ternus’ turn has come. One day I will stop with this pun. Today is not that day.

This was the new Apple CEO’s first week on the job. Next week is his first major Apple event. And it’s a big one. In an email to employees on his first day, Ternus called it a “huge launch.”

The invitations that went out last week called the event, “Surprise and Shine.” Except thanks to the many reports—most of them from legendary Apple reporter Mark Gurman at Bloomberg—the “surprise” part isn’t exactly looking like it’s going to hold up. Here’s my betting board for next week:

🟢 iPhone 18 Pro and Pro Max: These are about as close to a lock as it gets. Expect faster A20 Pro chips, camera upgrades and—far less fun—price increases of up to $200. Unlike in years past, the regular iPhone 18 isn’t expected. I’d bet my dongle collection that Apple keeps the iPhone 17 models around as the “base” iPhones in that usual keynote slide showing the full lineup.

🟢 Foldable iPhone: This will be the star of the show. Apple’s first folding iPhone, reportedly with a roughly 5.5-inch screen on the outside and a 7.8-inch, iPad Mini display when opened. It’s expected to have similar proportions to the new Galaxy Z Fold 8 (which starts around $1,900) and a similar price, if not higher. Word is $2,000 to $2,500. Biggest questions: What is this thing called? And what kind of multitasking support does it have?

🟢 Apple Watch and AirPods: Apple Watch Series 12 and Ultra 4 are expected, though the Watch upgrades sound minor, with new chips, possible battery improvements and some health/fitness refinements. Also expect AirPods 5.

🟠 Apple HomePod with Screen: Apple’s long-rumored smart-home hub—a HomePod-ish device with a screen—is reportedly coming sometime before the end of the year. There’s a chance the company could hold this announcement BUT I expect Siri AI, which arrives soon with iOS 27, to be a huge part of next week’s event. And this would be another way to showcase all the improvements. So yes, I’m being bold here and predicting we hear about this.

🔴 AirPods Cameras: Yes, a few weeks ago, Apple made a giant mistake and leaked a video showing the future AirPods with cameras that work with Siri AI. (An Apple spokeswoman declined to comment when the leaked video was released.) But I highly doubt we’ll see these next week. Not only do the reports say no, but this is the kind of product that will require a lot of marketing nuance given the AI and surveillance backlash. I don’t see Apple wanting to distract from the sure bets of the day.

Alright, what did I get wrong? Place your bets and I’ll…absolutely not buy you a new iPhone Fold.

We here at New Things think quite a bit about scams. Remember the book scammer that really got me? Some information about me on the internet is unavoidable (I did write a book!), but other details, like my phone number and address, don’t need their own publicity tour. They leave me more vulnerable to scams, phishing, identity theft, harassment and doxing.

That’s where DeleteMe comes in. The subscription service searches hundreds of data brokers and people-search sites for my personal information and works to remove it. And because those listings can reappear, DeleteMe keeps monitoring and submitting removals throughout the year. Data privacy, unfortunately, is not a one-and-done project.

I receive emailed privacy reports showing exactly how many listings DeleteMe found and removed. The company is based in the U.S., never outsources the removal process to third parties and was recently named Wirecutter’s top data-removal service. 

Get 20% off at https://joindeleteme.com/NEWTHINGS and use code NEWTHINGS at checkout.

AI watchers have spent years anticipating the arrival of AGI like it’s moon landing. The problem? There’s no agreed-upon test for when we’ve actually crossed that line.

AGI, by the way, is artificial general intelligence, or AI that can match humans across a broad range of intellectual tasks.

But on Thursday, at a press briefing for OpenAI’s latest model, GPT-6/Astra, company president Greg Brockman effectively declared that, by his definition, the AGI era has arrived. Except it sounded a bit like midcentury Supreme Court Justice Potter Stewart’s famous test for what counts as pornography: “I know it when I see it.” Even Brockman acknowledged that “AGI” is more of a vibes thing.

“Everyone has a different definition of AGI,” Brockman acknowledged. “When we started OpenAI, we kind of thought that, oh, there’s going to be this well-defined moment that everyone will recognize as, ‘That’s AGI!’ It hasn’t played out like that. It’s a much more gray, fuzzy thing.” 

Without a doubt, these models can already do some absolutely amazing things, and this new one can do even more. But as Brockman rightly points out, we’ve all been anticipating some blindingly obvious breakthrough in AI’s capabilities, but no one really knows how to call it. And as a slew of tech watchers and AI critics have also been pointing out, AI execs have business reasons, especially ahead of an IPO, to promote their technologies as both transformative (which they clearly are) and inevitable (jury’s out).

For the record, my dog just said AGI is here.

Amazon Scam Check

Amazon famously sends a lot of emails—order confirmations, shipping updates, review requests, etc. But sometimes those emails aren’t actually from Amazon, and now AI has made it easier than ever to send convincing scam communications in massive volumes. Enter: Scamazon™. Our term, but feel free to steal it, Amazon.

To address the problem, Amazon has released a new tool to verify whether a message is real or fake. There are two ways to do this: 

1. Forward any email you’re unsure about to verify@amazon.com.

2. Go to the Alexa for Shopping chatbot on Amazon.com or in the Amazon app, and tell it the details of any communication (where the message came from, when it arrived, etc.). 

The company says it can provide a “definitive answer” on these messages’ authenticity by using AI to scan the context, timing, metadata, etc. and check it against Amazon’s database. 

Name: Stephen Miller

What's your old thing? Olivetti Valentine 

What year is your old thing from? 1970

Why do you love your old thing? Typewriters like this were cutting-edge communications technology for over 100 years. The QWERTY keyboard arrangement, introduced on typewriters in the 1870s, is the same layout used on smartphones and laptops today. Typewriters were nearly always black. Olivetti's red Valentine was radical and way ahead of Apple's foray into multicolored cellphones.

I wrote a novel, titled “Suzy and the Sidecar,” on this “old thing” in 1973. Recently I brought the book back to life through today's self-publishing platforms, editing software, AI (to assist in cover design only) and audio technologies. We could not have imagined any of these 50+ years ago. 

Fun Fact: The first Labor Day was celebrated in New York City on September 5, 1882—a Tuesday! By 1894, the holiday had gained so much popularity that President Grover Cleveland signed a bill officially designating the first Monday in September a national holiday. Now we have him to thank for a weekend full of sales on gadgets AND great tech newsletters!

This newsletter was written and curated by Joanna Stern and Adele Lowitz. 

Hi! Welcome to Techtember, when the gadget announcements flow like pumpkin spice and the AI agents do the back-to-school shopping. Yes, I handed my son’s school supply list to the big platforms—Claude, ChatGPT and Gemini—and told them to shop. The results are below.

Then: Tim Cook is out at Apple, John Ternus is in! Dyson wants to put cameras in your mouth, and Sonos is back. Plus, Apple Polishing Cloth jokes, because some traditions are sacred.

Hey! Thanks for being a free subscriber to New Things!

It’s still summer—technically—so our end-of-summer sale is still on.

Upgrade to an annual subscription for 25% off—just $75 for the whole year. 

That’s less than one $9 Apple Polishing Cloth a month.

AI labs have all sorts of fancy benchmarks for testing their agents, but they don’t have The School Supply Shopping Benchmark™. It’s grueling. It involves online shopping carts and finding the perfect handheld pencil sharpener and packs of pre-sharpened pencils requested by a fourth-grade teacher.

Last year, during my AI Year Takeover for my book, I AM NOT A ROBOT, I used Perplexity’s Comet browser and ChatGPT’s Atlas (RIP), the cutting edge of agentic browsers at the time, to do my son’s back-to-school shopping. 

A year later, Claude, ChatGPT, Gemini and others now all have agents that can take action across the web, and specifically in their own native web browsers. (Claude added this last week.) So naturally, I’ve turned this into an annual benchmark. Forget Humanity’s Last Exam. This is the Buy the Right Glue Sticks Exam. 

The results were surprisingly clear: These anonymous tasks on the web have gotten so much better in just the past 12 months. One agent absolutely crushed the task, another wasn’t far behind and the third still seems to be looking for the fine-tip black Sharpies.

Fun fact: Sharks have a continuous tooth replacement system—as one row of teeth gets worn down, others push forward. Some sharks go through up to 35,000 teeth in their lifetime. Selling sharks $500 toothbrushes would be a terrible business.

This newsletter was written and curated by Joanna Stern and Adele Lowitz. 

Hello! Welcome to everyone-has-an-out-of-office week! But not me. I’m here. At least today. I am taking Friday off. Then we’ll be back next week for the start of Techtember!

But today, we’ve got a jam-packed newsletter! First, Apple just announced its big Sept. 9 event. Then, Ring’s new encryption and security promise, Apple’s new Macs—and, yes, a new polishing cloth. Plus, Meta’s big settlement, more LED-covering tips and a LOLz Pro Max featuring the World Humanoid Robot Games AND John Oliver rightfully making fun of me.

Hey! Thanks for being a free subscriber to New Things!

Great news: Our end-of-summer sale is on!

Upgrade to an annual subscription for 25% off—just $75 for the whole year. 

These prices won’t last! And apparently neither will my transformation into a used-car salesperson.

Mark the calendars. Apple’s iPhone event is set for Wednesday, Sept. 9. The invite above landed in my inbox just a few minutes ago. We’ll be in Cupertino covering all the news. Folding iPhone? A smart home device? AirPods with cameras? Could be some, could be none!

But this is shaping up to be one of the more exciting Apple events in years.

Right now, people trust tech companies about as much as they trust a contractor who says, “We’ll be done with the job in a month.” Today’s Ring news seems to be keenly aware of that. 

The news: Ring, the Amazon-owned company that put cameras in millions of doorbells, has created a new security standard called TAKE, or Throw Away the Key Encryption. It’ll be turned on by default for all accounts. The company will continue to offer end-to-end encryption of its devices’ video feeds, which is the highest level of protection.

But this announcement is about more than encryption. Ring is putting more cameras in more places—and adding AI features like facial recognition and object descriptions—at a time when people are increasingly asking: Who exactly can see what these cameras see? 

I talked to Ring CEO Jamie Siminoff about TAKE, and why building a camera-everywhere future doesn’t work if people don’t actually trust the companies behind the cameras.

Mercury’s latest report is packed with interesting stats, whether you’re running a business or just curious about how AI is changing one.

This figure jumped out: 40% of companies using AI heavily say that inflation actually helped their business this year. Among non-AI adopters? Just 12%. Same economy. Very different vibes.

And that divide shows up all over Mercury’s survey of 1,500 early-stage founders and builders. Companies leaning hardest into AI are raising more, hiring differently and even experiencing the economy differently.

Read Mercury’s new report to explore:

• The 31-point confidence gap between heavy AI adopters and everyone else

• What startups spend on AI each month

• How junior hiring looks at AI-forward companies and beyond

Plus more on how founders are spending, raising and hiring in 2026. Read the report.

Fun fact: Mood rings (which are really just temperature-sensing rings) were invented in the 1970s by Marvin Wenick, a New York jeweler who realized that the liquid crystal elements found in thermometers could be used in jewelry. Someone from Oura must have been around taking notes.

This newsletter was written and curated by Joanna Stern and Adele Lowitz. 

Hello! Hope you had a good week. I did, though I spent a bit too much of it thinking about Perv Technology—AKA glasses and earbuds with cameras—and whether Google Glass poisoned that well for everyone.

Then, the first installment of Ask Joanna, where you send questions and I attempt answers. Reply to this email and we might answer it in a future edition. And finally, Chumby! Remember the Chumby?! If you do, congratulations, it’s probably time to schedule a colonoscopy.

The hottest tech of the summer? Perv tech. Yes, as in tech for perverts.

“Perv glasses” has become a popular online nickname for Meta’s smart glasses, largely because of exactly what we explored in our investigation a few months ago. People (fine, mostly men) are hitting on women while wearing the glasses—and sometimes trying to hide the recording light or the fact that they’re recording at all.

Then this week, Apple’s rumored camera-equipped AirPods leaked and the internet gave us: PervPods. I desperately wish I could take credit, but that one belongs to the very online masses. Apple declined to comment on the leak. 

The whole thing feels like a flashback to the days of Google Glassholes. As you can see above, I am still a proud owner of the failed tech. You’ll remember Glass failed for a few of reasons. Sure, the technology wasn’t great, but there was a bigger problem: People thought you were a Glasshole for wearing a camera on your face. In public. In private. Basically anywhere other humans existed. People actually made and hung these signs in public places. 

Now, more than a decade later, we’re right back in familiar territory. Meta’s glasses have developed their own stigma—and in some ways, the problem is even worse than it was for Google Glass, for two reasons:

The camera you don’t notice. 

The Meta glasses look like regular sunglasses. These leaked camera-equipped AirPods look like regular AirPods Pro. They may even just be the next version of AirPods Pro.

M.G. Siegler made a great point in his Spyglass newsletter this week: Both Meta and Apple may hope to avoid the Google Glasshole problem by making new technology look like something we already know—not something that makes you look like an…asshole. 

Aesthetically, that’s great. From a privacy perspective, it’s part of the problem. You may not realize the person wearing perfectly normal-looking glasses—or AirPods—has cameras pointed at you. And sure, there might be a tiny recording indicator on the Meta glasses, but even when that light hasn’t been drilled out or otherwise obscured, it can be pretty damn hard to see.

(Mark Gurman at Bloomberg has reported that the cameras on the AirPods are low resolution and won’t have the ability to take photos or video. They will just be able to be used to show Siri things. We’ll soon find out. Or in 2027.)  

The surveillance hangover. 

The current backlash to mass surveillance and Big Tech is real. Look at what’s happening with Flock cameras across the country right now. Communities are pushing back over privacy, sloppy data-sharing policies and a lack of meaningful oversight. 

Look at the protests over data centers, too. When I talk to people who are resistant to AI—or just to more tech in general—I often hear complaints about a lack of control and trust. 

And take this study the Pew Research Center released this week. The majority of adults under 30 (55%) say they’re more concerned than excited about AI. That’s not a small share. 

So against this backdrop, introducing AI wearables with cameras that don’t really look like cameras feels like…potentially not a great thing. The tech industry may have solved the Google Glass design problem. I’m not sure it can ever solve the real Google Glasshole problem. 

Googling your name can be fun, until you find your address, phone number and relatives just a few blue links down. Data brokers collect and sell that personal information, so you are at greater risk of scams, phishing, identity theft, harassment and doxing. Fun times. 

DeleteMe is a hands-free subscription service that finds and removes your information from hundreds of data brokers and people-search sites. Because listings can reappear, DeleteMe keeps monitoring and removing them throughout the year. 

I especially appreciate the emailed privacy reports I receive about myself and my spouse. They show me the exact numbers of what was found and removed. DeleteMe has specialized in personally identifiable information (PII) removal for over 15 years. Plus, the company has privacy advisors (real humans!) to help with custom requests. Imagine that: humans helping humans in 2026!  

Get 20% off at https://joindeleteme.com/NEWTHINGS and use code NEWTHINGS at checkout.

CREDIT: Lamptech

Gary, that is a great question! We went looking for papers, experts, anyone who'd studied this and what we found is that the first LED indicators weren't for tech bros. They were for repair technicians.

Monsanto (which made gadgets before doubling down on agrichemicals and getting acquired by Bayer) started mass-producing small red lamps around 1968. In a later product catalog, the pitch was "diagnostic lights on printed circuit boards and panel lights.” Basically, little red dots that lived inside machines so a technician could tell which board had died. You can see one of the first red LEDs made by the company here.

Adele and I have searched for the moment LEDs made the jump from inside electronics to blinking at us all the time, but we couldn’t find a definitive answer. The reason for the transition is pretty obvious: LEDs are a cheap and easy way for a gadget to tell you its status. They are tiny, use little power and can last a long time—or at least as long as the device does. 

Sure, I guess you could argue that our smartphones and apps could tell us that info now. I guess my best answer to you, Gary, is it’s time to get some black electrical tape and cover those suckers up. Sleep well! 

Name: Jim McNamee

What's your old thing? Chumby 

What year is your old thing from? 2007

Why do you love your old thing? Chumby was a cute little touchscreen in a beanbag (with a hidden button on top) that displayed widgets. We mainly used it to see the local weather and also made our own widget to show what was playing in the house on iTunes. Kind of a precursor to Google Home and Amazon Echo. It had a built-in FM tuner, two USB-A ports on the back and a headphone jack...but we never used that stuff! Still, it was a cool ahead-of-it's-time gadget that was also, as I said, cute!

Correction: Wednesday’s newsletter incorrectly identified Reddit CEO Steve Huffman as Scott Huffman. We apologize to all, but specifically Steve.

Fun Fact: Red LEDs have come a long way from their humble beginnings in the 1960s. Now, red light is being used in $400 face masks to stimulate collagen. See Wirecutter’s picks here.

This newsletter was written and curated by Joanna Stern and Adele Lowitz.

Good morning! We’ve set a New Things world record: Wednesday’s feature video is actually out before lunch. Please alert Guinness. More below on a very 2026 problem: Companies want us talking to AI bots. Humanity would prefer not to hear it. So now there’s tech that can hear whispers and conceal our voices.

Especially timely after this week’s AirPods-with-cameras leak and Apple’s increasingly obvious plans to turn AirPods into an AI device.

Then, Reddit gets a video feed, OpenAI slows down and self-driving trucks hit California streets. Plus, a LOLz involving pee, data centers and a joke I’m way too proud of: GP(ee)U.

Hey! You’re currently a free subscriber to New Things.

Today’s newsletter—including all the premium, human-written words—is for paid subscribers only.

Come join the cool club! Subscribe here.

I believe all “Her” references are officially trite, except for the one I’m about to make. In the 2013 movie, Joaquin Phoenix’s character, Theodore, walks around the city talking to Samantha, his AI companion, through a tiny earpiece. He chats with her on the train, at the park and in lots of other places with other people around. 

The movie got a lot right about our AI future. But it missed one important thing: Everyone around Theodore would have wanted him to shut the hell up. 

That’s the problem Silicon Valley now wants to solve. Voice is quickly becoming the way tech companies want us to interact with AI. OpenAI President Greg Brockman told me last month that the goal is for speaking to replace typing, adding that this will eventually become “by far the vast majority of what we do.” Already, apps like Wispr Flow have surged in popularity. 

This week’s feature video explores the increasingly weird technology being invented so we can talk (or whisper) to technology—without making everyone around us hate us. Here’s what I found. 

Fun fact:  During the Civil War, Confederate chemist Jonathan Harrelson figured out how to extract potassium nitrate from human urine to make gunpowder. Read more about the history of urine and gunpowder here.

This newsletter was written and curated by Joanna Stern and Adele Lowitz.

I’ve officially run out of ways to greet you on a summer Friday. Happy Friday? We made it? Anyway, I hope you can unplug early today. I’ve been doing a surprising amount of unplugging over summer weekends, thanks to the Brick—the first thing I’ve tried that has actually kept me off social media for days at a time. More on that below.

Then Instagzam’s new logo (not a typo) and an Old Thing from 1965 that still works today. Oh, and please send us your Old Things! 

I’d bet on my dog doing the dishes before Apple and Samsung would ever ship their phones with a Brick, but gosh, I wish they would.

The Brick is a $59 piece of plastic. A magical yet simple piece of plastic.  

The little NFC-powered box essentially bricks your phone—or, more specifically, whatever apps you tell it to brick. Here’s how it works: You choose which apps you want blocked in the Brick app, then tap your phone (iPhone or Android) against the little box. Those apps are locked until you physically tap the Brick again. Watch our new short video to see it in action.

I’ve been using it to block social media on weekends. And somehow this ridiculously simple system has worked better for me than anything else I’ve tried over the years. And I’ve tried it all: Apple’s Screen Time, grayscaling my screen, moving apps off my home screen, deleting apps, putting my phone in another room.

Here’s why this has worked, and why I think it’s worth a try if you have the same phone-addiction problem I do. (By the way, this is NOT an ad. I’ve just become slightly obsessed with this thing this summer.)

1) It adds friction. 

The problem with Screen Time and Google’s Digital Wellbeing? They’re far too easy to override. You set a limit, then inevitably think, Ehhh, I’ve been good for the last hour, let me back into Instagram, and tap Ignore.

With the Brick, you have to physically find it and tap your phone to get back into those apps. Yes, there’s a software escape hatch, but it’s buried in settings and you only get five “Emergency Unbricks.” And to get more Unbricks you have to actually contact the company!

“One of the most effective ways to break habits is to add friction to make it more difficult,” behavioral scientist Asaf Mazar, who has published extensively on habits, told me. Given that software-only limits are easy to bypass and social-media apps are designed to eliminate all friction—see infinite scroll, autoplay and more—we don’t stand a chance.

In one not-yet-peer-reviewed study he worked on, researchers just logged people out of TikTok. They could log right back in whenever they wanted, but that tiny extra step cut TikTok use by about half an hour a day.

That’s basically what the Brick does. Yes, it’s a small obstacle to go get the Brick, especially if I happen to be in the house. But apparently my laziness is stronger than my desire to look at Instagram. (I usually just keep my Brick in a drawer in my closet.)

2) I want it to work.  

I asked Mazar why willpower alone doesn’t work, especially since this summer I’ve really wanted to be better about putting down my phone and being present on weekends.

“Habits are so fast and so automatic and happen without you even intending to do them,” he said. “If you're just using willpower, it's going to be very, very tough because by the time you even have a chance to think, you're already in Instagram, you're already scrolling, you already saw a poodle on its hind legs solving math problems.”

Still, there’s something about the psychology of all this that’s working for me right now. The combination of friction and intention gives me just enough time to remember why I Bricked it in the first place: I spend all week looking at a screen with all the distractions of the internet and social media. I don’t need to spend my weekends there, too. My kids and my life aren’t inside a glowing rectangular slab.

Sometimes it takes a Brick to remind you to touch grass.

• Instagzam. Instagram updated its wordmark yesterday for the first time in more than a decade. It’s less cursive than its predecessor, but also less legible—hence the typo in our title. It reads more like Instagzam or possibly Instaguam. Meta actually posted the other logos it could have gone with. Some of them even say Instagram. Intezzzesting. And yes, now other brands (ChatGPT and Coinbase) are jumping in on the fun.  

• Microsoft app merge. Speaking of redesigns, Copilot users will be seeing a slightly different logo soon. The Copilot and Microsoft 365 Copilot app will be merging into one cleverly named app…Microsoft Copilot. The move is part of Microsoft’s journey towards a “Super App,” streamlining personal and work accounts under one roof. 

• WhatsApp Scam Detection. An optional Scam Alert tool in WhatsApp uses AI to detect and flag suspicious messages. The feature operates entirely on your device, so Meta will not get access to any of your private information, or know you clicked that random link about discounted Ray-Bans that your great aunt sent you.  

Name: Jerry Dawson 

What's your old thing? Panasonic Radar Matic R-1000P

What year is your old thing from? 1965

Why do you love your old thing? This was my favorite radio as a teenager—10 transistors, three diodes and, best of all, auto tuning! You would wind up the spring on the back, turn it on, press the bar on the top and it would seek strong AM signals and automatically stop whenever it found one.

My childhood friend and I each saved up our money and took the bus from the suburbs to downtown Cleveland to Olson Electronics, where we each bought one. They were $30, which was a lot of money in 1965. Equivalent to $300 today! It runs on four AA batteries and still works perfectly to this day. Not sure why I had to have auto tuning, but I knew that I did. 

Fun fact: The Amazon wordmark has an arrow running from the “A” to the “Z.” The arrow does two things at once: illustrates that Amazon carries everything from A to Z and it makes the word into the shape of a little e-commerce smile. 😀

This newsletter was written and curated by Joanna Stern and Adele Lowitz.

Good morning! I’m in your inbox a bit early because Google just announced its new Pixel phones, and I’ve got all the details, including a cool LED ring on the back that lets you know who is trying to get in touch with you. Yes, notification lights are back, baby!

Then, Anthropic is watermarking Claude outputs, Amazon is sending oddly vague shipping emails and Mark Zuckerberg has written his open-model opus. Plus, a LOLz Pro Max roundup because while everything on the internet may be getting worse, it has at least been extremely funny this week.

Hey! You’re currently a free subscriber to New Things.

Today’s newsletter—including all the premium, human-written words—is for paid subscribers only.

Come join the club! Subscribe here.

Breaking news: Google has reinvented the BlackBerry notification light.

The company just announced its new Pixel 11 lineup, including the new 11 Pro, 11 Pro XL and 11 Pro Fold. I want to tell you I’m wildly excited about all of it. But I’ll be honest, most of it feels pretty standard. The Google briefing I attended last week was a parade of new camera specs, Gemini enhancements and processor improvements. But then they showed me…a light.

Google calls it the HiLight, a customizable LED ring around the camera flash on the Pro models that lights up differently depending on who is trying to get your attention. Mom calls? Red. Your boss? Blue. Your boss’s mom? Purple.

And somehow, of all the fancy new technology in these phones, this is the feature I want. More on that, the higher prices, the latest foldable design and everything else below.

It’s been nearly six months since I started building New Things, which means the business is growing—and so is my need to keep an eye on what we’re spending.

Lucky for me, Mercury just launched Spend, an expense-management tool that lets me set intelligent budgets for the team. The system can automatically cut off spending when someone hits a limit. And by “someone,” I mostly mean me.

For instance, I set up a travel budget that caps monthly travel expenses for me and David (our head of video / my very patient producer) at $5,000. Mercury automatically classifies things like airline tickets, hotel charges and car rentals as travel expenses. Now, David doesn’t have to ask me every time he wants to buy a snack box from the middle seat.

Receipts can also be uploaded right from the app, and cards with overdue tasks can be automatically locked until those tasks are resolved. Hahaha, got you all! No chasing receipts, no manual review, no tax-time scramble.

More than 300,000 entrepreneurs rely on Mercury. The financial tech company has been an essential partner since we launched New Things, and I’m excited to keep growing my business alongside theirs.

Fun fact:  Fabriano, a small town in Italy famous for its medieval papermaking, is credited with inventing the watermark in the 13th century to identify which mill made which paper. Bet they never imagined that 700 years later, we’d use the same idea to make sure everyone knows Claude wrote my email.

This newsletter was written and curated by Joanna Stern and Adele Lowitz.

We’ve made it through another week. I spent a lot of mine folding open a phone. More below on why the Samsung Galaxy Z Fold 8 changed my outlook on foldables, and made me even more excited for Apple to join the fold.

Then, Google Maps has some new AI tricks—that work… sometimes. And happy birthday to the Apple Newton! 🎂

I’ve spent years making fun of foldable phones. 

I’ve stuck a hot dog in one. I’ve complained about their awkward shapes and flimsiness. And I’ve wondered why anyone would pay nearly $2,000 for a phone that required its own cargo pant pocket.

Well, I have some news: I really like Samsung’s new Galaxy Z Fold 8. Really like it.  

For the first time, I’ve been using a foldable phone every day, and the reason I’ve finally fallen for one is surprisingly simple: It fits in my pocket. Actually, multiple pockets. My jeans back pocket. My blazer pocket. My small backpack pocket. Watch our latest video on the Fold to see it in all those pockets and more.

Previously, the foldable compromise was: big phone becomes small tablet. Now it’s more like: smallish phone becomes a good-sized tablet.

Samsung spent seven years making its foldables thinner, shorter and wider. The Fold 8 is now roughly the size and shape of a passport when closed, rather than the tall, skinny candy-bar shape of earlier Folds—and now the new Fold Ultra. That seemingly small change makes a huge difference. Here’s what else I’ve learned living with it:

• The outside feels like a normal phone. The 5.5-inch outer screen is good for texting, email and all the usual phone stuff. One-handed typing can be a stretch on the square-ish screen because of the wider shape, but I rarely feel like I have to open it.

• Then you get a tablet. Open it and there’s a 7.6-inch screen that’s great for reading, watching videos and running two apps side by side. This is the whole appeal: phone when I want a phone, tablet when I have both hands to spare.

• The foldable weirdness is fading. It feels sturdier than Samsung’s early Folds, it’s water-resistant and the crease is much less distracting. Split-screen apps also fit really nicely and software overall feels polished. Sure, the cameras are a step down from Samsung’s best non-folding phones, but they’re still very good. Battery life got me through full days, too, often with around 15% left by bedtime.

• The price is still absurd. The Fold 8 starts at $1,899. Nearly $2,000! (Remember Samsung points the finger in part at RAMageddon.) 

But maybe the most interesting thing about the Z Fold 8 is its timing. Apple is reportedly preparing its first foldable iPhone for this fall, with similar proportions. 

After living with Samsung’s version, I now know exactly what Apple has to get right: software, durability and, most of all, a design that makes it actually useful. The bigger challenge? Reports suggest the price could be as high as $2,500. An Apple spokesperson declined to comment.

Like my newfound love, Samsung’s Z Fold 8, Apple’s foldable may fit in all the pockets, but paying for it will require deep ones.  

Googling your name can be fun, until you find your address, phone number and relatives just a few blue links down. Data brokers collect and sell that personal information, so you are at greater risk of scams, phishing, identity theft, harassment and doxing. Fun times. 

DeleteMe is a hands-free subscription service that finds and removes your information from hundreds of data brokers and people-search sites. Because listings can reappear, DeleteMe keeps monitoring and removing them throughout the year. 

I especially appreciate the emailed privacy reports I receive about myself and my spouse. They show me the exact numbers of what was found and removed. DeleteMe has specialized in personal identifiable information (PII) removal for over 15 years. Plus, the company has privacy advisors (real humans!) to help with custom requests. Imagine that: humans helping humans in 2026!  

Get 20% off at https://joindeleteme.com/NEWTHINGS and use code NEWTHINGS at checkout.

Google Maps’ Gemini-powered “Ask Maps” feature gained some new agentic powers this week, including the ability to order food, book hotels and find event tickets.

It can also pull context from your other Google apps, including Gmail and Google Calendar, to personalize recommendations. That is, if you turn it on. (Tap your profile in upper right > Connected Content Apps > Select the apps you want to turn on.)  Just know what you’re trading for that convenience: You’re giving the AI access to information in your email and calendar so it can use that data to personalize its answers.

But before you start letting “Ask Maps” do things for you, a word of caution: I asked it to find me IMAX showtimes for Spider-Man in NYC.

Gemini confidently returned two showtimes that didn’t match anything on the actual theater website. When I pushed back, the AI admitted it had made a mistake—and had actually given me showtimes for The Odyssey. So, yeah, I’m not quite ready to trust it to autonomously start buying things for me. (The tool does say below every result: “Gemini is AI and can make mistakes.”) 

I also tried to order food. No luck. According to Google, Uber Eats ordering in Maps is “rolling out soon.” No AI-powered vodka pasta delivery tonight.

Slight change of plans for this week’s Old Thing. When we started the New Things x NBC News partnership, I thought it would be fun to dig through old NBC segments about technology. And, oh boy, did we find some things.

One of my favorites is a 1993 segment about Apple’s Newton, filmed a few months before the first MessagePad went on sale. (This week marks 33 years since its August 3, 1993 launch.)

You can watch the short video featuring the old segment on NBC News, but two pieces of Newton technology really jumped out at me. Both were pretty rough but turned out to be ahead of their time:

• Transferring files. The Newton could “beam” notes to other Newtons. “The beam is infrared light, and it enables him to zap me a note he has written on his machine,” NBC News correspondent George Lewis explains in the 1993 segment. Sounds a lot like Apple’s AirDrop or Android’s Quick Share.  

• Handwriting to text. The Newton became infamous for its very bad handwriting recognition. But fast forward 30 years and the Apple Pencil now works across much of the iPad lineup, where your handwriting can be turned into typed text. 

Turns out Newton was just about three decades early.

Fun fact: The Newton’s handwriting recognition was mocked so far and wide that it was included in an episode of The Simpsons and a “Doonesbury” comic.

This newsletter was written and curated by Joanna Stern and Adele Lowitz.