Whether we’ve secured our Brother printers, our optimism or skepticism about Apple fixing accessibility issues in betas, thoughts on a MacBook powered by an iPhone chip, and the books we’re planning to read—or reread—this summer.
The modern Apple strategy is to re-use older technologies to create more affordable products… Why does the M1 MacBook Air [still] exist? Because Apple wants to have a product available at a (relatively) low price point…
Now let’s imagine a world with a M3 MacBook Air in it. Does Apple discontinue the M2 model, or push it down into the $999 range? Does Apple discontinue the M1 Air at that point? In the Intel era I’d have answered yes, but the Apple silicon era is something different. The truth is, even now, the M1 is more than enough for most potential Mac users.
Just as a thought experiment, consider what Apple might do if it was planning to import the iPhone SE strategy to the Mac. It would take some older, but still quite capable technologies—say, everything that makes up an M1 Mac. The device’s parts are carefully scrutinized with an eye toward eliminating cost wherever possible, without sacrificing a basic Apple level of quality.
More-Affordable MacBook… Expected to enter mass production in late 4Q25 or early 1Q26, with an approximately 13-inch display and powered by the A18 Pro processor. Potential casing colors include silver, blue, pink, and yellow… The more-affordable MacBook is projected to account for 5–7 million units for 2026.
The more things change, the more they remain the same. My thoughts about this rumor are very similar to my thoughts 22 months ago. First off, the M1 MacBook Air can’t be sold forever. I’m sure the margins on a five-year-old product are great, but Apple and TSMC surely want to stop making M1 chips at some point! So how do you make a new product that’s still well below the $999 of the (incredible value) M4 MacBook Air?
Using the same A18 Pro processor found in the iPhone 16 Pro might be a good start. Let’s look at the relative speed of the A18 Pro versus the M series found in Apple silicon Macs:
Well, would you look at that? The A18 Pro is 46% faster than the M1 in single-core tasks, and almost identical to the M1 on multi-core and graphics tasks. If you wanted to get rid of the M1 MacBook Air but have decided that even today, its performance characteristics make it perfectly suitable as a low-cost Mac laptop, building a new model on the A18 Pro would not be a bad move. It wouldn’t have Thunderbolt, only USB-C, but that’s not a dealbreaker on a cheap laptop. It might re-use parts from the M1 Air, including the display.
I like that Apple sells a laptop at $649, and I think Apple likes it, too. A new low-end model might steal some buyers from the $999 MacBook Air, but I’d wager it would reach a lot of customers who might otherwise not buy a full-priced Mac—the same ones buying M1 MacBook Airs at Walmart.
Why the macOS Tahoe Menu Bar is the start of something big, Apple may lower the bar when it comes to an entry-level Mac laptop, and we try to parse the reasoning behind Apple’s latest set of EU App Store rules and regulations.
Some have lambasted these rules as “needlessly complex” or “clear as mud” or “I don’t know what lambaste means, it sounds like what you do before broiling those chops, but I’m still angry!” But I’ve completed a careful analytical reading of the new terms, and feel confident that the revised agreement is nothing more than sheer elegance in its simplicity. Allow me to elucidate.
You, the developer, get paid. Apple takes some percentage of that—ah ah ah, yes, I know the question “how much?” is already burning on your lips, but stay with me: I am simplifying.
The amount of that percentage depends on a few basic factors: are you in the Small Business Program? Did you agree to the alternative business terms?…
Apple has intended the Apple Watch to be a nifty complement to your iPhone from its release; interaction with the Mac came later. (With an iPad? Well, it’s not entirely neglected.)
One of the key ways you can use your Apple Watch with a Mac is to unlock it. Two different readers have expressed frustration with how that’s working.
Six Colors subscriber HiddenJester notes:
It’s always had some limitations, where sometimes the screen says the wireless signal is too weak or it occasionally demands a password, but it generally works pretty well. Except … it quit working for me last week, and no amount of toggling the switch off and back on again works.
I noticed this morning while I was looking at the password prompt that it occasionally flashed “Unlocking with Apple Watch …” underneath the password field, but it immediately goes away.
They have tried rebooting their Apple Watch and Mac and are still locked out.
A similar report came in from Six Colors reader Coach Mike via Mastodon:
Any thoughts/ideas on how to fix an Apple Watch that suddenly stops unlocking a Mac?
Rebooted Mac couple of days ago; rebooted Watch this morning. (And yes, Watch is “unlocked, on my wrist and powered on”.)
Why would a convenient feature stall out? My suspicion is that has to do with how Apple manages to bypass using macOS account passwords (or encrypted versions of them) while also ensuring that your Apple Watch remains under your control.
Watch it unlock your Mac
To enable this feature, called Auto Unlock in Apple’s documentation, you first need to meet these requirements:
The Mac must be a model released from mid-2013 or later and have macOS 10.12 Sierra or later installed.
All Apple Watch models offer this feature, but must be running at least watchOS 3 or later.
Your Apple Account has to be set up with two-factor authentication and you’re using the same Apple Account on both devices.
You must have set a passcode for your Apple Watch.1
Both your Mac and Apple Watch must have Bluetooth and Wi-Fi enabled.
(Update: Continuity features like Auto Unlock sometimes require devices are on the same Wi-Fi network; Auto Unlock doesn’t explicitly list this as a requirement. However, Coach Mike discovered after this article first appeared that his watch had shifted to the Guest network at his house. He used Settings on the Apple Watch to change back to his primary network, at which point Auto Unlock worked again!)
With all of that in place, you go to System Settings > Touch ID & Password on a Mac with an integrated Touch ID sensor or with a Magic Keyboard with Touch ID2, or System Settings > Login Password if there’s no Touch ID sensor attached. You can then enable the switch under the Apple Watch label at the bottom to let your wearable unlock your Mac, as well as unlock applications like Password or allow autofilling passwords and verification codes. (The full text reads “Use Apple Watch to unlock your applications and your Mac.”) If you have more than one Apple Watch, such as a day and night watch, you can enable them separately.
Use the switch for your Apple Watch under the Apple Watch label to unable automatic unlocking. (Image: Apple)
As with general Mac password security, every time you restart your Mac, shut down and power it back up, or log out of an account and back in, you must re-enter the account’s password—you cannot use Auto Unlock until after that point.
Now, you just tap a key on the keyboard, move a mouse, or tap a trackpad, and your Mac automatically unlocks.3 This won’t work if a remote device is controlling your Mac’s screen or, peculiarly, if you are using Internet sharing, where you use your Mac to pass an Internet connection or one or more Mac hard-wired interfaces or via Wi-Fi.
If your Apple Watch is close to your Mac, your Mac should simply unlock.
When you’re in a regular Mac session and Touch ID or an administrator password is requested, you also receive a notification on your Apple Watch and can double-press the side button to approve it.4
That should be the beginning and end of it. However, as our two readers note and I have seen myself, sometimes it just breaks.
This may be due to the layered security that enables this to happen at all.
A fragile tunnel?
Auto Unlock isn’t just about letting your Apple Watch unlock your Mac. The same technology allows your iPhone to unlock your Apple Watch after it restarts or starts up, or when it loses a connection with your body, such as after you’ve charged it and put it back on your wrist. (Go to the iPhone Watch app > Passcode > Unlock with iPhone.) It’s also used in some cases when you’re wearing a mask to let you unlock your iPhone with your Apple Watch.
Auto Unlock doesn’t involve storing or revealing passwords. Instead, setting up a connection between your devices allows them to pass secrets that prove their identity to each other coupled with proximity detection to ensure your devices are near one another. When this system fails, intentionally or not, it provides unuseful feedback—or none at all.
When you turn on Auto Unlock features, Apple creates an end-to-end encrypted tunnel using what it calls a Station-to-Station (STS) protocol. One set of long-lasting keys gets created when the feature is turned on to allow STS to work over time. However, the protocol doesn’t simply rely on that. Instead, it generates a unique session key to use whenever Auto Unlock is invoked. Because this feature works with Macs without the Secure Enclave5, the end-to-end tunnel uses Secure Enclave at both ends when available, but can also terminate on earlier Macs at the Mac’s kernel (its core operating system component).
This operation relies on Bluetooth—in particular, Bluetooth Low Energy (BLE). It also relies on peer-to-peer Wi-Fi, which is used to let two devices roughly calculate the distance between them, helping deter longer-range cracking attempts. Those attempts are themselves highly unlikely to succeed due to the layered encryption employed: all transactions over both networking types are separately encrypted.
The initial secret is sent by a “target”: the device that will allow itself to be automatically unlocked, such as a Mac to an Apple Watch. This secret passes over the STS tunnel using a BLE connection. When an unlock request is triggered, a new peer-to-peer Wi-Fi connection is established, and the distance is calculated. If the two devices are close enough to each other—seemingly a few feet, but it’s not documented. (Apple describes this as your Apple Watch being “very close to your Mac.”)
If the conditions are met, the unlocking device sends the target back the unlock secret it previously received. The target sends a new challenge that the unlocked device solves. Voila! The target unlocks, and transmits a new secret for the next go-round.6
There could be a few things going wrong that result in a silent failure or a failure with the wrong error message:
Despite being close enough, the two devices decide they are more than the requisite distance apart. This should result in a message stating that the Apple Watch connection is too weak. However, I’ve seen that error even when my wrist is inches from the Mac, so it must be displayed when it is provably false or instead of the correct error.
A glitch causes the handshake to fail between the two devices. Instead of announcing it, perhaps Apple chooses to let the process break without a message to avoid providing feedback that attackers could use to hone an exploit.
Gremlins?
There’s one way around it.
The time is out of joint
The key problem—pun definitely intended—is that enabling Auto Lock sets a trigger that’s later released. Disabling the Auto Lock feature deletes associated keys, including breaking the STS tunnel. The correct order of reset is:
Disable System Settings > Touch ID & Password (or Login Password) > Apple Watch > your device.
Restart your Mac and your Apple Watch.
Log into your Mac and enter the passcode for your Apple Watch.
Return to the Auto Unlock area and re-enable.
If you haven’t tried this sequence in precisely this order, now is the time to do so. If you have and it hasn’t worked, try it again. We all know that repeating the same actions with computers shouldn’t produce a different outcome. Yet, like rotating a USB Type-A plug three times to get it to fit, some parts of the technology world defy explanation.
[Got a question for the column? You can email glenn@sixcolors.com or use/glennin our subscriber-only Discord community.]
Apple doesn’t require a passcode to use an Apple Watch, but you miss out on many features and basic security protections without one. ↩
Sometimes, I can be sitting on our house’s main (and sole) floor above my day-lit basement office, and my Apple Watch says it’s mysteriously unlocked my computer. ↩
Approving these logins requires a slightly later system version than for Auto Lock: macOS 10.15 Catalina or later and watchOS 6 or later. ↩
Intel Macs received a full Secure Enclave via the T2 Security Chip, starting with models introduced in 2018. All Apple silicon Macs have a Secure Enclave subcomponent in their M-series chip. ↩
You’re not just watching the show, you’re attending it. Sit in the front row with other fans and direct your own experience by switching between cameras in real-time, getting up close to the conversation on the stage… VR filmmakers have always struggled with where to put the cut. We give that agency to the viewer, and the effect is unexpectedly magical.
I was sitting in the front row of the theater off to the far left, right behind Sandwich’s capture camera, and Theater remarkably replicates that experience. When you swipe to move between cameras, you also move locations in Theater’s immersive venue, maintaining geographical sense and the illusion of watching John Gruber and his guests on stage. It’s not immersive video in the way Apple has defined it, but it’s 3-D video playing on a stage inside an immersive environment, so it’s pretty close!
According to Sandwich, “the production uses a combination of stereoscopic 6K cameras and iPhones with AI 3D conversion.” There are more details on how it was all put together in a video by Adam Savage and Tested.
Apple has a movie out? Who knew? The company makes a better beta and begrudgingly complies with EU rules.
Unhappy meal
Welcome back to the summer of “F1”, already in progress. If you’re just joining us, F1 F1 F1 EFFFFF ONNNNNNEEEE!!!
VROOM-VROOM! WHHHHRRRRRRRROOOOOOM!
OK, lemme give you the 411 on “F1”.
[scoots closer, puts arm around reader]
Look, I don’t know if you’ve heard… but Apple… yes, the company that makes those things. And the other things. That’s right. That’s the one.
Anyway, you probably haven’t heard this, but Apple… has a new movie out.
I know. No one could be more surprised than I. I wish they would have told someone.
It is called… you might want to get a pencil and write it down so you don’t forget because it’s a bit of a sleeper… it is called…
“F1”.
“F1 – The Movie” if you’re nasty.
I’m not going to link to a story about it because online outrage is not exactly a new or even particularly noteworthy phenomenon, but just know that people are mad.…
With over 5,000 five star reviews; Magic Lasso Adblock is simply the best Safari ad blocker for your iPhone, iPad and Mac.
Designed from the ground up to protect your privacy, Magic Lasso blocks all intrusive ads, trackers and annoyances – stopping you from being followed by ads around the web.
Why American sports is so dominated by gambling content, DAZN and the Club World Cup, TNT gets spun off but may still bid for sports rights, Netflix gets into French broadcast TV, the NFL faces an existential crisis in court, and our TV picks.
Tiered App Store fees. For today’s full-service App Store, developers will now pay 13% on sales, reduced to 10% for Small Business Program members. Or developers can opt into “Tier One”, which comes with a 5% fee but does not support a raft of App Store features we’ve come to expect, like automatic app updates, App Store promotions, placement in search suggestions, ratings and reviews on product listings(!), and more.
Core Technology Commission. Apple is going to move all developers over to a new tax called the Core Technology Commission, in which developers who opt to sell apps outside the App Store will pay 5% of sales made through in-app promotions. The €0.50-per-install Core Technology Fee will be dropped as of January 1.
Free linking. Developers can promote offers broadly, are no longer limited to a single static URL without tracking parameters, and can freely design the interfaces for those links and promotions.
New business terms. Developers have to pay a 2% fee for digital goods and services purchased by new users for the first six months after a user first downloads an app; members of the Small Business Program don’t have to pay this fee.
“The European Commission is requiring Apple to make a series of additional changes to the App Store. We disagree with this outcome and plan to appeal.”
Apple always disagrees and always appeals, but these are pretty big changes. The introduction of a lower App Store tier with lower fees (but more spite?), combined with the reduced rates to the regular App Store fee structure, is especially fascinating. One has to wonder if Apple would’ve had as much trouble in the EU if it had made changes like this much sooner, but here we are.
The home screen layout has been revamped, with upcoming games now separated by league. You can further personalize the list by re-ordering the sections to their preference, so you can make the sports you are most interested in appear first, and starred teams always show at the very top. And, just in time for Wimbledon, the app has now added tennis scores.
The tennis coverage adds support for both the men’s and women’s tours, but you can’t follow individual players and it’s just for coverage of Grand Slams and the next tier down, 1000 level events. Backgrounds will change based on the court surface, so the green of Wimbledon will be up first.
The baseball screen has been updated to display the pitcher and hitter matchup directly beneath the score.
Though Apple Sports probably launched a bit prematurely, it has grown a lot over the past year and I find myself turning to it when I want a quick glimpse at what’s going on in a sport or league I care about.
Unfortunately, the app is still only available on iPhone and in the U.S., UK, and Canada.
Like the machine says, “Screenshot of a workflow automation tool with steps for receiving images, resizing, and describing them.”
Of all the features I’m excited about using in macOS 26, the one that most intrigues me is the Use Model action in Shortcuts. Use Model does exactly what you think it does: you toss data into it, and an AI model somewhere (on your Mac, on Apple’s Private Cloud Compute servers, or even at an OpenAI server farm) will take that data and turn it into… something.
The other day, I realized that this new feature would allow me to expand my existing automation that uploads images to the Six Colors web server by adding a description of the image. If there’s something AI stuff is pretty good at, it’s describing images. So I decided to give it a try.
Since Apple’s On-Device model doesn’t support image uploads, I used Private Cloud Compute. Uploading full-sized images to Private Cloud Compute led to very long processing times, so I inserted a step that resizes the image before sending it.
With all that done, the shortcut was able to generate a remarkably accurate description quickly—but it wasn’t quite ready for my use. The descriptions were frequently too long, and they included double quotes that would break if pasted inside the quotes of the alt attribute in an HTML image tag.
A baseball stadium filled with spectators under a clear blue sky. The field is green with a brown dirt infield, and players are positioned on the field. The stands are packed with fans wearing team colors.
No problem. I refined my prompt: “Describe this image for use in the alt text tag on a webpage. Limit yourself to two sentences at most. If it’s a screenshot, please include all the words. Don’t use double quotes, but only single quotes.”
A white dog with black spots standing on a tiled floor in front of wooden cabinets. The dog is wearing a blue collar with a tag. There is a chair with a blue cushion to the right. The floor has a brown and beige pattern.
This changed the output, but did it help? Some of the results were still way too long, and most of them contained double quotes regardless of my commands. I tried a few times to control the length of the result and the use of double quotes, but it was never consistent. Computer programs behave identically every time, but AI does not!
A room with a wooden shelf displaying vintage computers and monitors. A black office chair is in front of the shelf. The wall has framed pictures.
Eventually, I built a shortcut that had to be a hybrid of both approaches. My new prompt is: “Describe this image for use in the alt text tag on a webpage. Limit description to 200 characters maximum. If it’s a screenshot, please include all the words.” The result is frequently still too long, so now my shortcut checks the character count and if it’s too long, it passes the text to Apple’s on-device model with the prompt “Shorten this text to be less than 250 characters.” That works fairly well.
The final step is to use Shortcuts to search for the double-quote symbol and replace it with single quotes, because I really can’t rely on the AI to pull that one off. As with so much AI stuff, it’s simultaneously an amazing piece of technology and incredibly stupid.
And then there’s the final step: human intervention! My automation is meant to generate the HTML I want to insert in my stories, but that means I always see it and can adjust it as needed. There’s always a human being looking at that text and making sure it makes sense, which is good, because there will be errors. I got one description that inexplicably added on hashtags(!) and another that helpfully included the current time in a description of an Apple Watch screenshot—only to get the time entirely wrong.
Adding these kinds of tools to Shortcuts is going to be a real learning curve for all of us. And I’m sure app developers are feeling the same way about their new access to Apple’s on-device models. AI is a tool that can do some amazing things—but programmers (and Shortcuts creators) will still need to apply adult supervision.—Jason Snell
Handling expenses with AI: But at what cost?
Like Jason, I’ve been playing around with the Use Model action in Shortcuts. For me, the biggest attraction was the idea of conquering a workflow that I’d never been able to nail down before: filing my expense receipts.
You know me, I like a good expense tracking spreadsheet, but one of the pain points in my workflow is getting the receipt (often from my email), turning into a PDF (if it isn’t already), then laboriously entering the details into my spreadsheet. In the past, I’d attempted to create a shortcut to simplify entering this information, but the only real way to isolate the data I wanted was to enter it manually via a succession of prompts for the vendor, amount, and so on. Not exactly a timesaver from tabbing through spreadsheet columns.
But pulling information out of a document—especially information that might appear anywhere in a variety of forms—seems like something an AI model would be good at, so I decided to take another crack at it with Shortcuts’s new AI capabilities.
I started out my workflow by grabbing all the text from a PDF or web page, then passing it to the Private Cloud Compute model. (I attempted to use the On-Device model at first, but it was both very slow and not quite as good at formatting the response in the manner I wanted.)
At first, my goal was to get comma-separated values for the date, vendor, and dollar amount that could then be passed into my spreadsheet, but when that didn’t quite work I tried both tab- and newline-delimited and sent them to the clipboard, then tried to paste the values into my Numbers sheet. Turns out that doesn’t work great: Numbers really dislikes pasting things into multiple columns.
But after doing a bit of digging, I discovered the ideal format to pass to the “Add Row to Numbers Spreadsheet” action is actually a list. So I told the model to pass back my values as comma-separated values, then used the Split Text action to separate them on the commas, which automatically generates a list. At that point, I could simply pass that list to the Numbers action, and the values would get put in the first three rows of my sheet.1
The prompt I settled upon—after a lot of tweaking—was this:
This information is a receipt for payment. Please return the amount of the expense in US dollars, the date of the expense (look for dates in email headers if it’s formatted as such, otherwise use the curent date), and who the vendor was (use the subject line and from line in email headers if formatted as such). Also, create a new filename in the format: YYYY-MM-DD-[first five alphanumeric characters of vendor in uppercase]-[full dollar amount with no decimal point or dollar sign]
Format the values as follows, separated by commas:
date in MM/DD/YYYY,vendor,full dollar amount (make sure to include any decimal point but not the string USD),filename
Okay, so you probably noticed I put a whole separate section in there about a filename. That’s because I had the bright idea to see if the shortcut could rename the original file I gave it my receipt format.2
However, in typical AI fashion, this has introduced some problems. For one thing, the AI model gets confused that I’m asking for the dollar amount in two different formats. While I want the spreadsheet value for to be normally formatted—say $209.49—while I want the filename version to simply be the digits 20949. Unfortunately, sometimes it dumps decimal-less version into the spreadsheet, yielding a heart-stopping surprise expense of $20,949.
Another time-honored AI problem: counting. Oh, if these machines could count to five, perhaps they could rule the world. When I ask it to use the first five alphanumeric characters of the vendor, it first gave me six characters. Sometimes it will give me five…but not all the time.
And therein lies the rub with all of this. The results are neither reliable nor necessarily repeatable. The same data run through this shortcut multiple times provides different answers: I’d think that anathema (not to mention madness inducing) to the sensibilities of any programmer. Given the same data, the algorithm should yield the same thing every time, but the non-deterministic nature of AI models throws that out the window. Perhaps I could engineer an even better prompt that would get closer, but unlike other programming tasks I’m not even sure that I could test this enough to feel confident. I could run it 99 times and the 100th it could spit back something different and totally incorrect.
Given this, I find it no wonder that Apple punted on some of its promised Apple Intelligence features from last year. What I’m attempting is a relatively simple task for an AI model and it still gives me the wrong information sometimes. How confident would you be if it tells you when your mom’s flight lands?
I seem to have had some success in tweaking the prompt by explicitly asking for separate dollar amounts, but I’m still hardly filled with confidence. What I’ve ended up with is essentially a not very bright assistant whose work I need to double check. And if I can’t trust that work, does it actually save me time?—Dan Moren
Perhaps my favorite discovery of writing this workflow: for reasons that will shortly become apparent, I ended up with a list of four items, but wanted to pass only the first three to Numbers. However, there’s no facility in Shortcuts—that I could find—to remove a list item. Until, that is, I stumbled across a fascinating workaround. The “Filter Files” action, as its name suggests, is intended for files, but it actually works on any list you pass it. So I gave it my list of four values and told it to limit it to three items. Worked perfectly. ↩
I originally tried doing the filename separately with the On-Device model, but again, it proved unreliable. ↩
With the announcement that macOS Tahoe will be the last Mac OS version to support Intel Macs, Apple’s preparing to close the books on the third chip transition in Mac history.
It doesn’t get a lot of attention, but Apple is absolutely the best company in the world at picking up stakes and moving its platforms somewhere else. Over its 41 years of existence, the Mac has run on four entirely different processor architectures (not to mention two different operating system foundations), all the while remaining more or less the same familiar Mac we know and love.
This is not an easy feat to accomplish once, let alone three times. Apple’s gotten very good at this. Twenty years ago, it was the switch to Intel. Five years ago, the switch to Apple silicon started. And of course, way back in the mists of time when I was a brand-new hire at one of Macworld’s predecessor publications, Apple made the leap for the very first time.
