Apple’s latest research about running large language models on smartphones offers the clearest signal yet that the iPhone maker plans to catch up with its Silicon Valley rivals in generative artificial intelligence.

The paper, entitled “LLM in a Flash,” offers a “solution to a current computational bottleneck,” its researchers write.

Its approach “paves the way for effective inference of LLMs on devices with limited memory,” they said. Inference refers to how large language models, the large data repositories that power apps like ChatGPT, respond to users’ queries. Chatbots and LLMs normally run in vast data centers with much greater computing power than an iPhone.

The paper was published on December 12 but caught wider attention after Hugging Face, a popular site for AI researchers to showcase their work, highlighted it late on Wednesday. It is the second Apple paper on generative AI this month and follows earlier moves to enable image-generating models such as Stable Diffusion to run on its custom chips.

Device manufacturers and chipmakers are hoping that new AI features will help revive the smartphone market, which has had its worst year in a decade, with shipments falling an estimated 5 percent, according to Counterpoint Research.

Despite launching one of the first virtual assistants, Siri, back in 2011, Apple has been largely left out of the wave of excitement about generative AI that has swept through Silicon Valley in the year since OpenAI launched its breakthrough chatbot ChatGPT. Apple has been viewed by many in the AI community as lagging behind its Big Tech rivals, despite hiring Google’s top AI executive, John Giannandrea, in 2018.

While Microsoft and Google have largely focused on delivering chatbots and other generative AI services over the Internet from their vast cloud computing platforms, Apple’s research suggests that it will instead focus on AI that can run directly on an iPhone.

Apple’s rivals, such as Samsung, are gearing up to launch a new kind of “AI smartphone” next year. Counterpoint estimated more than 100 million AI-focused smartphones would be shipped in 2024, with 40 percent of new devices offering such capabilities by 2027.

The head of the world’s largest mobile chipmaker, Qualcomm chief executive Cristiano Amon, forecast that bringing AI to smartphones would create a whole new experience for consumers and reverse declining mobile sales.

“You’re going to see devices launch in early 2024 with a number of generative AI use cases,” he told the Financial Times in a recent interview. “As those things get scaled up, they start to make a meaningful change in the user experience and enable new innovation which has the potential to create a new upgrade cycle in smartphones.”

More sophisticated virtual assistants will be able to anticipate users’ actions such as texting or scheduling a meeting, he said, while devices will also be capable of new kinds of photo editing techniques.

Google this month unveiled a version of its new Gemini LLM that will run “natively” on its Pixel smartphones.

Running the kind of large AI model that powers ChatGPT or Google’s Bard on a personal device brings formidable technical challenges, because smartphones lack the huge computing resources and energy available in a data center. Solving this problem could mean that AI assistants respond more quickly than they do from the cloud and even work offline.

Ensuring that queries are answered on an individual’s own device without sending data to the cloud is also likely to bring privacy benefits, a key differentiator for Apple in recent years.

“Our experiment is designed to optimize inference efficiency on personal devices,” its researchers said. Apple tested its approach on models including Falcon 7B, a smaller version of an open source LLM originally developed by the Technology Innovation Institute in Abu Dhabi.

Optimizing LLMs to run on battery-powered devices has been a growing focus for AI researchers. Academic papers are not a direct indicator of how Apple intends to add new features to its products, but they offer a rare glimpse into its secretive research labs and the company’s latest technical breakthroughs.

“Our work not only provides a solution to a current computational bottleneck but also sets a precedent for future research,” wrote Apple’s researchers in the conclusion to their paper. “We believe as LLMs continue to grow in size and complexity, approaches like this work will be essential for harnessing their full potential in a wide range of devices and applications.”

Apple did not immediately respond to a request for comment.

https://arstechnica.com/?p=1992550

According to a report in Bloomberg, Tang Tan, vice president of Product Design, is leaving Apple, and his departure heralds a shuffle of executives heading up some of the company’s most important products.

Sometimes, you might wonder just how much a specific executive influences the grand scheme of things, but the report claims that people within Apple see Tan’s departure as “a blow,” clarifying that he “made critical decisions about Apple’s most important products.” His team reportedly had “tight control” over the look and functionality of those products.

Tan oversaw major aspects of iPhone and Apple Watch design, and he was the executive overseeing accessories and AirPods, as well. He reported to John Ternus, Apple’s senior vice president of Hardware Engineering, who is likely a more widely known name.

Richard Dinh, “Tan’s top lieutenant and head of iPhone product design,” will report directly to Ternus and take on some of Tan’s duties, while Kate Bergeron, previously involved in Mac hardware engineering, will take on the Apple Watch.

Apple has seen several executive departures from its product design and engineering groups recently, so many aspects of upcoming iPhones and other products will be designed with new eyes and perhaps new sensibilities, though what that might lead to remains to be seen.

Apple recently shifted the iPhone from the company’s proprietary Lightning port to a more standard USB-C, and it changed the materials for its Pro line of phones. Despite tweaks like that, the iPhone’s design and functionality has not changed significantly in the past five or so years.

The iPhone 16 line in 2024 is expected to shake things up a little more, at least regarding the phone’s look and feel. Rumors have suggested that the new phones may have larger screens (and bigger chassis overall) and perhaps haptic buttons instead of the current physical buttons. Other changes could be in store, and Apple’s plans are likely not yet finalized.

https://arstechnica.com/?p=1989815

You won’t want to miss out on these accessories for your iPhone 15 Pro. https://www.entrepreneur.com/science-technology/this-5-piece-apple-compatible-accessory-bundle-will-save/464814

Three years ago, Apple introduced a privacy-enhancing feature that hid the Wi-Fi address of iPhones and iPads when they joined a network. On Wednesday, the world learned that the feature has never worked as advertised. Despite promises that this never-changing address would be hidden and replaced with a private one that was unique to each SSID, Apple devices have continued to display the real one, which in turn got broadcast to every other connected device on the network.

The problem is that a Wi-Fi media access control address—typically called a media access control address or simply a MAC—can be used to track individuals from network to network, in much the way a license plate number can be used to track a vehicle as it moves around a city. Case in point: In 2013, a researcher unveiled a proof-of-concept device that logged the MAC of all devices it came into contact with. The idea was to distribute lots of them throughout a neighborhood or city and build a profile of iPhone users, including the social media sites they visited and the many locations they visited each day.

In the decade since, HTTPS-encrypted communications have become standard, so the ability of people on the same network to monitor other people’s traffic is generally not feasible. Still, a permanent MAC provides plenty of trackability, even now.

As I wrote at the time:

Enter CreepyDOL, a low-cost, distributed network of Wi-Fi sensors that stalks people as they move about neighborhoods or even entire cities. At 4.5 inches by 3.5 inches by 1.25 inches, each node is small enough to be slipped into a wall socket at the nearby gym, cafe, or break room. And with the ability for each one to share the Internet traffic it collects with every other node, the system can assemble a detailed dossier of personal data, including the schedules, e-mail addresses, personal photos, and current or past whereabouts of the person or people it monitors.

In 2020, Apple released iOS 14 with a feature that, by default, hid Wi-Fi MACs when devices connected to a network. Instead, the device displayed what Apple called a “private Wi-Fi address” that was different for each SSID. Over time, Apple has enhanced the feature, for instance, by allowing users to assign a new private Wi-Fi address for a given SSID.

On Wednesday, Apple released iOS 17.1. Among the various fixes was a patch for a vulnerability, tracked as CVE-2023-42846, which prevented the privacy feature from working. Tommy Mysk, one of the two security researchers Apple credited with discovering and reporting the vulnerability (Talal Haj Bakry was the other), told Ars that he tested all recent iOS releases and found the flaw dates back to version 14, released in September 2020.

“From the get-go, this feature was useless because of this bug,” he said. “We couldn’t stop the devices from sending these discovery requests, even with a VPN. Even in the Lockdown Mode.”

When an iPhone or any other device joins a network, it triggers a multicast message that is sent to all other devices on the network. By necessity, this message must include a MAC. Beginning with iOS 14, this value was, by default, different for each SSID.

To the casual observer, the feature appeared to work as advertised. The “source” listed in the request was the private Wi-Fi address. Digging in a little further, however, it became clear that the real, permanent MAC was still broadcast to all other connected devices, just in a different field of the request.

Mysk published a short video showing a Mac using the Wireshark packet sniffer to monitor traffic on the local network the Mac is connected to. When an iPhone running iOS prior to version 17.1 joins, it shares its real Wi-Fi MAC on port 5353/UDP.

In fairness to Apple, the feature wasn’t useless, because it did prevent passive sniffing by devices such as the above-referended CreepyDOL. But the failure to remove the real MAC from the port 5353/UDP still meant that anyone connected to a network could pull the unique identifier with no trouble.

The fallout for most iPhone and iPad users is likely to be minimal, if at all. But for people with strict privacy threat models, the failure of these devices to hide real MACs for three years could be a real problem, particularly given Apple’s express promise that using the feature “helps reduce tracking of your iPhone across different Wi-Fi networks.”

Apple hasn’t explained how a failure as basic as this one escaped notice for so long. The advisory the company issued Wednesday said only that the fix worked by “removing the vulnerable code.”

This post has been updated to add paragraphs 3 and 11 to provide additional context.

https://arstechnica.com/?p=1979099

A group of academic researchers has devised a technique to extract sounds from still images captured using smartphone cameras with rolling shutter and movable lens structures.

The movement of camera hardware, such as the Complementary Metal-oxide–Semiconductor (CMOS) rolling shutters and the moving lenses used for Optical Image Stabilization (OIS) and Auto Focus (AF), create sounds that are modulated into images as imperceptible distortions.

These types of smartphone cameras, the researchers explain in a research paper (PDF), create a “point-of-view (POV) optical-acoustic side channel for acoustic eavesdropping” that requires no line of sight, nor the presence of an object within the camera’s field of view.

Focusing on the limitations of this side channel – which relies on a “suitable mechanical path from the sound source to the smartphone” to support sound propagation, the researchers extract and analyze the leaked acoustic information identifying with high accuracy different speakers, genders, and spoken digits.

The academics relied on machine learning to recover information from human speech broadcast by speakers, in the context of an attacker that has a malicious application running on the smartphone but does not have access to the device’s microphone.

However, the threat model assumes that the attacker can captures a video with the victim’s camera and that they can acquire speech samples of the target individuals beforehand, to use them as part of the learning process.

Using a dataset of 10,000 samples of signal-digit utterances, the researchers performed three classification tasks (gender, identity, and digit recognition) and trained their model for each task. They used Google Pixel, Samsung Galaxy, and Apple iPhone devices for the experiments.

“Our evaluation with 10 smartphones on a spoken digit dataset reports 80.66%, 91.28%, and 99.67% accuracies on recognizing 10 spoken digits, 20 speakers, and 2 genders respectively,” the academics say.

Lower quality cameras, the researchers say, would limit the potential information leakage associated with this type of attack. Keeping smartphones away from speakers and adding vibration-isolation dampening materials between the phone and the transmitting surface should also help.

Smartphone makers can mitigate the attack through higher rolling shutter frequencies, random-code rolling shutters, tougher lens suspension springs, and lens locking mechanisms.

“We believe the high classification accuracies obtained in our evaluation and the related work using motion sensors suggest this optical-acoustic side channel can support more diverse malicious applications by incorporating speech reconstruction functionality in the signal processing pipeline,” the researchers added.

Related: Researchers Demo Electromagnetic Fault Injection Attacks on Drones

Related: Open Source Tool For Hunting Node.js Security Flaws

Related: New Speculative Execution Attack Against Apple M1 Chips

https://www.securityweek.com/researchers-extract-sounds-from-still-images-on-smartphone-cameras/

Apple has patched a potent chain of iOS zero-days that were used to infect the iPhone of an Egyptian presidential candidate with sophisticated spyware developed by a commercial exploit seller, Google and researchers from Citizen Lab said Friday.

The previously unknown vulnerabilities, which Apple patched on Thursday, were exploited in clickless attacks, meaning they didn’t require a target to take any steps other than to visit a website that used the HTTP protocol rather than the safer HTTPS alternative. A packet inspection device sitting on a cellular network in Egypt kept an eye out for connections from the phone of the targeted candidate and, when spotted, redirected it to a site that delivered the exploit chain, according to Citizen Lab, a research group at the University of Toronto’s Munk School.

A cast of villains, 3 0-days, and a compromised cell network

Citizen Lab said the attack was made possible by participation from the Egyptian government, spyware known as Predator sold by a company known as Cytrox, and hardware sold by Egypt-based Sandvine. The campaign targeted Ahmed Eltantawy, a former member of the Egyptian Parliament who announced he was running for president in March. Citizen Lab said the recent attacks were at least the third time Eltantawy’s iPhone has been attacked. One of them, in 2021, was successful and also installed Predator.

“The use of mercenary spyware to target a senior member of a country’s democratic opposition after they had announced their intention to run for president is a clear interference in free and fair elections and violates the rights to freedom of expression, assembly, and privacy,” Citizen Lab researchers Bill Marczak, John Scott-Railton, Daniel Roethlisberger, Bahr Abdul Razzak, Siena Anstis, and Ron Deibert wrote in a 4,200-word report. “It also directly contradicts how mercenary spyware firms publicly justify their sales.”

The vulnerabilities, which are patched in iOS versions 16.7 and iOS 17.0.1, are tracked as:

• CVE-2023-41993: Initial remote code execution in Safari
• CVE-2023-41991: PAC bypass
• CVE-2023-41992: Local privilege escalation in the XNU Kernel

According to research published Friday by members of Google’s Threat Analysis Group, the attackers who exploited the iOS vulnerabilities also had a separate exploit for installing the same Predator spyware on Android devices. Google patched the flaws on September 5 after receiving a report by a research group calling itself DarkNavy.

“TAG observed these exploits delivered in two different ways: the MITM injection and via one-time links sent directly to the target,” Maddie Stone, a researcher with the Google Threat Analysis Group wrote. “We were only able to obtain the initial renderer remote code execution vulnerability for Chrome, which was exploiting CVE-2023-4762.”

The attack was complex. Besides leveraging three separate iOS vulnerabilities, it also relied on hardware made by a manufacturer known as Sandvine. Sold under the brand umbrella PacketLogic, the hardware sat on the cellular network the targeted iPhone accessed and monitored traffic passing over it for his phone. Despite the precision, Citizen Lab said that the attack is blocked when users turn on a feature known as Lockdown, which Apple added to iOS last year. More about that later.

There’s little information about the iOS exploit chain other than it automatically triggered when a target visited a site hosting the malicious code. Once there, the exploits installed Predator with no further user action required.

To surreptitiously direct the iPhone to the attack site, it only needed to visit any HTTP site. Over the past five years or so, HTTPS has become the dominant means of connecting to websites because the encryption it uses prevents adversary-in-the-middle attackers from monitoring or manipulating data sent between the site and the visitor. HTTP sites still exist, and sometimes HTTPS connections can be downgraded to unencrypted HTTP ones.

Once Eltantawy visited an HTTP site, the PacketLogic device injected data into the traffic that surreptitiously connected the Apple device to a site that triggered the exploit chain.

Predator, the payload installed in the attack, is sold to a wide array of governments, including those of Armenia, Egypt, Greece, Indonesia, Madagascar, Oman, Saudi Arabia, and Serbia. Citizen Lab has said that Predator was used to target Ayman Nour, a member of the Egyptian political opposition living in exile in Turkey, and an Egyptian exiled journalist who hosts a popular news program and wishes to remain anonymous. Last year researchers from Cisco’s Talo security team exposed the inner workings of the malware after obtaining a binary of it. https://arstechnica.com/?p=1970625

Today marks the in-store launch of the iPhone 15 and iPhone 15 Pro, plus the likely delivery date for at least the earliest preorders. Preorders went live a week ago, on September 15.

You’ll be waiting for a while if you want the Pro model and didn’t preorder, though.

In Chicago, delivery dates for new orders of the iPhone 15 Pro and iPhone 15 Pro Max from the online Apple store are currently estimated to be between October 23 and 30—more than a month from now. Next-day in-store pickup is still a possibility for most configurations, except for the 1TB iPhone 15 Pro Max.

The regular iPhone 15 and iPhone 15 Plus don’t seem to have the same problem, though. I was offered immediate shipping or pick-up for every configuration I tried. All these estimates could be different not long after this is published, of course.

It’s tempting to look at that information and conclude that the Pro models will be more popular during this year’s cycle, but that’s not necessarily the case. It depends on how many units of each model Apple has produced, of course, and it stands to reason that early adopters who jumped right on preorders last week are enthusiasts who might be more interested in the Pro models.

A handful of companion products to the iPhone 15 lineup are also available today, including USB-C AirPods Pro and MagSafe chargers.

We currently have the iPhone 15, iPhone 15 Plus, iPhone 15 Pro, and iPhone 15 Pro Max in hand and are working on a review that will go live next week.

In case you missed the announcement a couple of weeks ago, the iPhone 15 brings several of the “Pro” features from the iPhone 14 Pro to the current base iPhone, including the Dynamic Island to replace the notch, Apple’s A16 chip, and a 48-megapixel camera sensor that is used to facilitate 2x zoom, among other things. It also ditches the long-standing proprietary Lightning connection in favor of the industry-standard USB-C.

The iPhone 15 Pro distinguishes itself from the base model with a new configurable “Action Button” to replace the mute switch, a faster USB-C port, a more robust camera system, a faster A17 chip, which claims notably improved graphics performance, and a new titanium enclosure. The phones’ general sizes, designs, and shapes are very similar to what we saw last year.

https://arstechnica.com/?p=1970510

Improving a device’s modularity and repairability isn’t just a hobby horse for right-to-repair activists—it can also significantly lower costs when something breaks. Case in point: the iPhone 15 Pro, which is said to pick up some of the same internal changes that Apple made to last year’s non-Pro iPhone 14 to make repairs easier.

Replacing the back glass in older iPhone X-style designs previously involved going in through the front of the phone, a tricky and involved process that made it expensive to pay for and extremely difficult to do by yourself. The iPhone 14 and iPhone 15 Pro change this, making it as easy to remove the back of the phone as it is to remove the screen and giving easier access to the battery and other components to boot.

To find the upshot, compare repair estimates on Apple’s iPhone Repair & Service page (via MacRumors). Fixing damaged rear glass on an iPhone 14 Pro or iPhone 14 Pro Max costs you $499 or $549, respectively, if you didn’t buy AppleCare+ protection for your phone. That’s half of what those models cost to buy brand new. For an iPhone 15 Pro and 15 Pro Max, that charge falls dramatically, down to $169 or $199.

That’s an even larger drop than we saw for the non-Pro iPhone models, where repair fees fall from $349 for an iPhone 13 to $169 for an iPhone 14. The costs for other repairs, including for the screen, the rear camera assembly, or the battery, are the same for the iPhone 15 Pro as they were for the iPhone 14 Pro.

Because of all the internal changes, the teardown artists at iFixit declared the iPhone 14 “the most significant design change to the iPhone in a long time,” even though the phones look pretty similar on the outside to the iPhone 13 models that they replaced.

In addition to lowering Apple’s official repair prices, the repairability improvements should also lower cost and complexity for people who buy their own parts through Apple’s Self Service Repair Store or third parties like iFixit. Just be aware that for some repairs, including anything that replaces a FaceID sensor, TouchID sensor, or a touchscreen, you may need Apple’s first-party tools to re-pair and recalibrate these components before they’ll work properly, and third-party battery replacements may also generate warning messages.

https://arstechnica.com/?p=1969009

For many people, the iPhone 12 effectively disappeared from the market on Tuesday, when Apple introduced iPhone 15 models and stopped selling the 12, first released in October 2020. In Europe, however, the iPhone 12 remains a notable device, as a number of countries are following France’s lead in looking into the device’s electromagnetic profile. With a software update coming, it may go back on sale soon.

What kicked off the unexpected concern about a nearly 3-year-old phone was France’s National Frequency Agency (ANFR). On the same day as Apple’s fall product announcements, the ANFR informed Apple that the iPhone 12 exceeds European Union regulations for Specific Absorption Rate (SAR), the rate at which a human body would absorb radiation from a device. A translated version of the ANFR report has the agency calling on Apple to withdraw the iPhone 12, “quickly remedy this malfunction,” and if not, “recall copies already sold.”

There are two measures of SAR for a device operating in the same frequency range as an iPhone, per EU standards. The “head and trunk” value, taken to protect against “acute exposure effects on central nervous tissues” when a phone is against the head or in a pants pocket, must not exceed 2 Watts of power per kilogram of body tissue, averaged over six minutes. When the phone is held in the hand or in clothing or accessories, for a “limbs” value, it’s 4 W/kg.

France’s ANFR measured the iPhone 12 exceeding the “limbs” limit at 5.74 W/kg. The ANFR stated that it would ensure the iPhone 12 was no longer available for sale in France and would oversee “corrective updates” it expects from Apple. Jean-Noel Barot, a digital and telecommunications minister in France, told newspaper Le Parisien that software updates could fix the issue, according to Reuters.

Apple responded swiftly to ANFR’s claims, telling multiple press outlets earlier this week that the iPhone 12 was certified by multiple international bodies and that it had provided the ANFR with documentation showing the device within regulatory limits, both from within Apple and independent lab results. Apple eventually told Reuters that it would issue a software update “to accommodate the protocol used by French regulators.” The company noted that it believed the issue was “related to a specific testing protocol used by French regulators and not a safety concern.” French officials stated that they were eager to test the software update and lift their ban on sales of the iPhone 12.

The EU’s standards note that within a phone’s typical frequency range, the main danger of excess radiation is not changes to cells or chemicals in the body, leading to cancer, but “whole-body heat stress and excessive localized heating of tissues.” The vast majority of mobile phone research indicates no adverse effects from regular exposure to the non-ionizing frequencies phones use to communicate. But a series of studies, however inconclusive or problematic, have raised unnecessary concern and garnered media attention. The World Health Organization states that “no adverse health effects have been established as being caused by mobile phone use.”

France’s notice has spurred action by other countries. Germany’s Federal Office for Radiation Protection said Wednesday that “the question of the need for change is currently the subject of discussions,” Reuters reported. Belgium’s state secretary for digitalization, Mathieu Michel, told Reuters that he reached out to regulators to review not just the iPhone 12 but all Apple smartphones and other devices. Denmark and Italy have said they are investigating but have taken no formal actions.

In the US, SAR limits set by the Federal Communications Commission are 1.6 W/kg. The iPhone 12’s submitted SAR levels were measured at 1.554 W/kg at their peak, generally when using a hotspot or engaging in “Simultaneous Transmission.” The iPhone 12 did, of course, clear the FCC for release in 2020.

French regulators have recently shown enthusiasm for demanding more from US-based tech companies. They’ve asked Google and Facebook to offer one-click cookie rejection and put repairability scores on smartphones and appliances and told the US and other nations that they want to see global AI regulations by year’s end.

This story was updated on September 15 to include Apple’s response and decision to issue a software update.

https://arstechnica.com/?p=1968361

Alongside the announcement of the iPhone 15 and iPhone 15 Pro, Apple quietly ended the iPhone 13 mini‘s run today. That marks the end of life for arguably the best premium small phone designed for one-handed use.

It’s not a surprise, of course. It became clear shortly after the launch of the first iPhone with the “mini” label—the iPhone 12 mini—that it wasn’t selling that well. Market research has consistently shown that most users want bigger screens and batteries, which are incompatible with a smaller phone.

Further, the mini split the small-phone market with the much cheaper iPhone SE—even though the mini offered drastically improved features, like an OLED screen and better cameras.

Apple went ahead and shipped an iPhone 13 mini, and it even kept that phone in the lineup when it introduced the iPhone 14 last year. But now the iPhone 13 mini is nowhere to be seen in Apple’s online store, even though the larger iPhone 13 still exists as a down-market alternative to the new flagships.

The iPhone SE is also still there, but its features are not too far removed from the iPhone 8 that came out six years ago. Supply chain analysts and journalists with inside knowledge have generally agreed that Apple could update the iPhone SE as soon as next year—but most rumors have suggested that it would go bigger, not smaller.

A dirge for small phones

The fact that the iPhone 13 was kept in the lineup may give lovers of small phones some hope that the iPhone SE refresh might be essentially the iPhone 13 mini or something like it. We’ve seen no evidence to support that beyond that it might seem odd Apple cut one but not both of the 13 models from the lineup.

The iPhone 12 and 13 minis were always niche products. Many op-eds have been written at tech sites (including this one, by me) saying it’s important for Apple to offer a one-handed phone size, even if it’s less popular than the bigger models.

Apple’s early strategy of offering a one-size-fits-all iPhone made more sense when there was less competition from Android handsets and when owning a smartphone wasn’t yet the default. Now, these devices are essential for modern life, and it makes sense to offer variations to fit different lifestyles, priorities, and, yes, even hand sizes.

That said, supporting an additional size adds extra supply chain and production overhead—and it may not be the most financially prudent decision for Apple if it’s confident that it can covert mini owners to join the majority and opt for larger phones instead.

Chances are that axing the iPhone 13 mini won’t do much to hurt Apple’s earnings, but it leaves a minority of consumers without an important option, and I still think that’s bad news.

https://arstechnica.com/?p=1967913