apple

which Apple John Ternus inherits and where the cracks are

by

It arrived last night for Spanish time, but it is the news of the day. After 15 years at the head of Apple, Tim Cook will leave his position on September 1. It will be the moment in which a John Ternus50 years old and with 25 years in the company, will assume the position of CEO. From a sales shark, we will move to a hardware man. And, although Cook will leave Apple in an enviable position, Ternus will take the reins with significant debts and many open fronts. Let him Cook. Yeah jobs He was the idealist, cook was the executor. Most of the hardware of the Apple of the three decades with Tim at the helm was inherited from Jobs’ Apple, but Cook has made them shine. Products such as Apple Watch either AirPodsthe Mac line was turned upside down, giving sit down Intel to bet on its own chips and even the MacBook, which was not one of the economic legs of the business, has experienced tremendous success of the macbook neothe ‘cheap MacBook’. But under the mandate of a Cook who has developed his career doing numbers and not designing products, what has shone has been the services segment. Services > hardware. We already said it in 2024: services exceeded the combined revenue of practically all of Apple’s hardware except for the iPhone, which still represented approximately half of the company’s sales. Apple has gone from being a company recognizable for its devices to one in which the iPhone was still capital, yes, but in which wearables (AirPods and Apple Watch), accessories, home devices, iPad and MacBook They were capitulating to the software. Apple has been buying software (Pixelmator, for example) to strengthen its suite of applications, but it has also been moving to subscription models and opening new lines. AppleOnefor example, is the umbrella under which services such as Apple Music, Fitness+cloud storage, Apple Arcade either AppleTV. All with a monthly subscription, of course. Because the company has a complete line of services and subscriptions to compete against other alternatives. They are in all segments (music, video games, personal care, movies and series) and represent a lifestyle. But they also have their creative suite for compete directly with Adobe with apps like the aforementioned Pixelmator, Final Cut Pro or Logic Pro Proven effectiveness. If we talk about money, this translates into a beastly lever. Under Cook, Apple has grown, and it has grown in a big way: Income from 108 billion dollars in 2011 to more than 391,000 in 2024. Close to 4x. Capitalization which went from 320 billion to 3.7 trillion dollars in 2024, almost 11x. The margin left by services and subscriptions is greater and more direct than that of the iPhone, also allowing a constant cash flow that is not seen as much with hardware launches. An iPhone is purchased once a year, a subscription is paid every month. And all this leads us to talk about a different Apple than in 2011. If then it was the company of the iPhone and other devices, today it is a business that encompasses software and hardware. And there are not many who can boast of that. And Cook’s Apple has built this while creating an extremely streamlined and diversified supply chain with factories in China, India and Vietnam to adapt to regulations and market shocks. But although that is ‘Side A’ of Tim Cook’s film, there is a ‘Side B’. The limbo of the Vision Pro. At this stage we have seen many hardware successes, but also some devices that raise many questions. I don’t get into AirPower anymore (Do you remember AirPower?), but in the field of VisionPro. Apple’s VR/AR headset is a technological marvel, but also an expensive device, with compromises like that hanging battery and, above all, one that you don’t really know where it goes. Launched in 2024 at $3,500, they have not seen the light of day in many markets, they continue to cost the same and a renewal has not been presented in these almost two years. At that time, the other great promoter of VR, Meta, has moved to glasses that they have made a place for themselves among content creators and people who want to document their daily lives. It is evident that it is not the same product, but the underlying idea is: a wearable that you wear for much of the day to create content and consume it. In a different way (audio in the Meta, more video in the Vision Pro), but consuming content, in short. In these two years, we have not been able to give the Vision Pro a ‘killer app’, something that would really make them irresistible within the Apple ecosystem. Tariffs and Apple Intelligence. But if the Vision Pro can afford to be in limbo, these two issues cannot. First, the tariffs. That Apple is moving to other countries to stop depending so much on China is not a whim. Globalization has been breaking down with the most current protectionist policies. Not depending so much on China (which continues to happen) is necessary if Apple, like so many other companies, does not want tariffs to affect its products. It is a very complicated juggling act because the United States wants its companies to leave China, but China is consolidating like a huge marketwith Apple and NVIDIA being two of the companies more interested in continuing to penetrate it to get part of a huge pie. On the other hand, Apple Intelligence. Apple’s AI was introduced two years ago and not only is it not at the level of rivals like Google, they have to ask Google for help to revitalize Siri. Apple Intelligence It is the great pending task of an Apple that has not known how to compete in these early stages of the artificial intelligence race. 2026 is expected to be the year (even They have sent most of the engineers to an ‘AI … Read more

Who is Johny Srouji and why this great unknown has just become the second most powerful person at Apple

by

For those who have been following Apple for a long time, Johny Srouji is no stranger. For the rest of the world yes, but after the appointment of John Ternus as CEO of Applethis Israeli engineer has become the second most powerful person in the company. The question is obvious: who is Johny Srouji? Who is Srouji and why does he matter?. Born in Haifa, Israel, in 1964 to a middle-class Christian Arab family, Srouji studied computer science at the Israel Institute of Technology (Technion) and graduated Summa Cum Laude in both his engineering and master’s degrees. He worked at Intel and IBM before Apple hired him in 2008 with a very clear assignment: to design the company’s first chips. He did much more than that. The revolution made chip. That first chip designed by Srouji was the Apple A4, which debuted in 2010 in the original iPad and the iPhone 4. From there, Srouji forged one of the most prestigious hardware careers in the recent history of the technology industry. The A7 of 2013 was the first SoC in using 64-bit architecture, and then there would come the revolution of the Apple M1 with which the company definitively got rid of dependence on Intel in its Macs. But his work goes beyond. His official title until now was senior vice president of hardware technologies, but it did not reflect the real scope of his work. Srouji not only led the chip design. Also that of batteries, cameras, storage controllers, sensors, displays, cellular modems and other critical components of the entire family of Apple devices. Almost everything that makes these products work the way they do is largely due to the work of Srouji and his team. With the new position, his responsibility expands and he will now control the entire cycle: not only the hardware itself, but also the physical design. It’s a colossal challenge, but if anyone seems prepared to take it on, it’s Srouji. He was about to leave. In December 2025 Bloombeg reported that Srouji had informed Tim Cook that he was seriously considering leave Apple in the near future. Two days later, Srouji himself published a message to his team denying the newsbut the damage was done. For Apple to lose Srouji would have been a disaster, and it is very likely that this new position is in part Apple’s response to that alarm signal. Textbook talent retention, but raised to maximum power. New position, new structure. In it internal communication that Srouji has sent to team employees, the engineer detailed how he will organize the division into five areas: Hardware engineering: led by Tom Marieb, an Intel veteran who joined Apple in 2019. Siilicio: it will be directed by Sri Santhanam, a manager with a long career at Apple Advanced Technologies: Supervised by Zongjian Chen Platform architecture: led by Tim Millet Program management: will be managed by Donny Nordhues In that message, Srouji acknowledges that this “represents a significant change” but believes it will work thanks to the entire team. It seems that you are very clear about how you want to work with your team. A fusion with a lot of historical sense. The reunification of hardware engineering and the hardware technologies division under the same leader is not entirely new. It is the structure that Apple had for years under the direction of Bob Mansfield, former head of hardware. until 2013 and? then he took charge of the failed Project TitanApple’s car. That’s when those two areas were divided, something that allowed both Ternus and Srouji to progress in their domains, but also caused some structural tensions between teams that had to collaborate. Bringing them back together is a clear commitment to strengthening that collaboration. The great cover-up of Ternus’s appointment. It is normal that the vast majority of headlines go to Ternus, who will decide the future of the company from now on, but Apple is above all a hardware company. That Srouji now becomes his leader makes this engineer a person with enormous power within the company. The change is promising in terms of promoting that facet of the product that both he and Ternus dominate, and without a doubt interesting times await us at Apple. Image | Apple In Xataka | John Ternus, vice president of Apple: “The iPhone Air had been in development for years, but we had to say ‘no’ until now”

Samsung has shown a new device with AI. It is not what we imagined and is reminiscent of an Apple idea

by

When they tell us about a new device with artificial intelligencethe normal thing is that we think of a mobile phone, a laptop or, at most, the disappointing Rabbit R1 either Humane AI Pin. That’s why it’s interesting to stop when a company like Samsung teaches something that doesn’t quite fit into any of those boxes. What we have seen this time is not a common gadget, but a rather revealing clue as to how the South Korean giant could imagine a possible home interface of the future. What Samsung has shown in Milan is called Project Luna and, at least for now, it moves in the field of concepts. It is a desktop device with a mobile circular screen that acts as a head and can rotate to face the user. The company’s materials also show that this head not only rotates, but also changes orientation depending on the angle it needs. With that combination, Samsung draws a home device that wants to look less like a conventional speaker and more like an object designed to interact with the user. A concept that points further than a speaker One of the scenes that Samsung has used to show Luna places it on a kitchen table, connected to the user’s smartphone, playing music with an interface reminiscent of a record player and answering questions both by voice and on screen. In that same demonstration he also appears controlling the lighting in the room and suggesting food options for the day. Additionally, there are projectors scattered around the kitchen that display data such as the calories in the recipe or a calendar notice for a dinner party. And that’s where Luna begins to tell us something more interesting than her own design. In an interview with Fast CompanyMauro Porcini, Samsung’s chief design officer, explained that this concept represents more of “a vibe, a feeling of the type of design language we want to use.” The phrase matters because it lowers any immediate commercial reading and forces us to look at it differently. Rather than anticipating a launch, the firm seems to use this project to teach the type of language and relationship with the user that it wants to explore in future AI devices. And at that point it is difficult not to remember Apple. In August 2024, Mark Gurman told Bloomberg that the company was moving forward with the development of a home desktop device that would combine an iPad-like screen with a robotic arm. The proposal, according to that informationwas conceived as a home control center, a tool for video calls and a remote surveillance system, with a screen capable of tilting and rotating 360 degrees using actuators. It has not materialized as a product, but there is some underlying parallel with what Samsung is now teaching. The most interesting reading may not be in looking for an exact equivalence between what Samsung has taught and the rumors about Apple, but in stopping at the underlying trend. What we’ve seen suggests that home AI could end up taking a much more tangible form than the assistants or screens we already know. We are not yet talking about a consolidated category, far from it. But it does provide a fairly serious clue as to where the industry could move in the coming years. At this point, the temptation is to think: okay, that sounds good, but where exactly does something like this fit into our daily lives. Because we can imagine it on the kitchen counter, recommending a mealanswering a quick question or accompanying us while music plays, and the scene is even convincing. The problem is that that same house is already full of devices that already cover a good part of all that. Images | Samsung In Xataka | Meta spent 2 billion on a Chinese AI startup. China is clear that it was a conspiracy

The European Union is very determined that batteries are removable in 2027. And Apple is very determined that they are not

by

The European Union has detailed your plan so that phone batteries are removable starting in 2027. A regulation that will arrive in the midst of the greatest revolution in battery capacity in recent years: the democratization of silicon-carbon and the commitment to 100% models eSIM. For years, the industry has abandoned removable batteries for technical reasons. The unibody design has improved water and dust resistance, optimized internal space and increased structural efficiency. It is something that has not been enough for Europe. What’s going to happen. Europe has reconfirmed its plan to make phone batteries be removable by law in 2027. It is a hard blow for many manufacturers, who have been creating a “unibody” design industry for years to improve their resistance to water and dust, as well as a more efficient internal structure (every millimeter inside a smartphone is key). That the batteries have to be removable dynamites the plan that manufacturers have been following for almost a decade, since it is required by law that they be “easily removable.” What does this mean. The novelty in the text has to do, precisely, with what the European Union understands with this concept. The manufacturer will no longer be able to glue the batteries to the plate, something that required them to be removed using a heat gun to remove the adhesive. If a specialized tool is required to remove the battery, it will be provided by the manufacturer itself. The manufacturer will need to include clear instructions on battery removal and replacement. The software will not be able to hinder the process or block phone functions if the replacement has not been done at an official service. Hardware availability must be at least five years. The cost of batteries must be “reasonable and non-discriminatory” in price. Although there are many questions in the air and, as usual, few specificities, what is clear is that starting next year there will be a very powerful change at the industry level. And then, China arrived. China is leading a paradigm shift in smartphone battery technology: the arrival of silicon-carbon. This type of battery allows much higher energy densities in the sizes we already know. Or, in other words, that in the usual batteries we have more capacity than ever. Thanks to China, phone batteries are skyrocketing to cases 10,000mAh. The problem? European restrictions on the transport of high-density batteries are very strict, which is why many of the phones that are marketed in China with enormous batteries end up landing in Europe with noticeably smaller cells. And that, for the consumer, is a problem. Europe moves the industry, but China does not plan to stop. Europe is an important enough market that big technology companies have to completely rethink the hardware of their devices. The best example is Apple, which had to bend to USB-C in all territories of the world due to EU demands. An especially painful move, considering that it was the only manufacturer with its own charging port standard. The mandatory nature of removable batteries will, once again, lead to a change in the industry. But China faces the dilemma between slowing its progress with carbon silicon batteries (something it does not seem willing to do) to give in to European regulations, or assuming the extra cost of manufacturing a model for each region. The latter has been doing so for years and, in fact, pushes some consumers to opt for more complete versions from China. not so fast. Although the headline goes to the return of the removable battery, there is an exception for which manufacturers like Apple have been protecting themselves for years. Commission Communication C/2025/214, which develops and interprets article 11 of Regulation (EU) 2023/1542, contemplates an exit for durability: if the battery retains at least 80% of its capacity after 1,000 charge cycles (or 83% after 500 cycles) and the phone meets an IP protection standard, the manufacturer may be exempt from the requirement for user replacement. It is no coincidence that, two years ago, Apple will double the number of cycles of its batteriesgoing from 500 to 1,000. Thanks to this exception, both Apple and many manufacturers will be able to continue selling a good part of their phones as we know them today. Summing up. The smartphone industry is on the verge of chaotic change. The cheapest mobile phones seem to be forced to give up the unibody design, China has hit a wall to continue innovating in batteries that can leave the country (it is not easy to continue growing sizes and maintain cycle capacity), and the high-end seems that it will be able to escape if it maintains batteries that maintain 80% of their capacity after 1,000 cycles. A soap opera whose ending we still do not know, but for which the manufacturers will have to start writing the next chapter. In Xataka | How to charge your mobile phone to maximize battery life

Apple is two years behind its competitors. So he’s sending 200 engineers to an “AI camp.”

by

When we talk about AI Big Tech, there is one name missing: Apple. There are many “the wolf is coming” in this matter of artificial intelligencewith companies that are creating ‘hype’ with models that they consider very dangerous and, above all, with artificial general intelligence. However, the “the wolf is coming” par excellence in AI is the new Siri and Apple Intelligence. Apple is tired of being the last and has made the most radical decision two months before WWDC. Sending almost all Siri engineers to early summer camp. Issues. Apple has two approaches with AI. On the one hand, a more transparent one for the user that interconnects applications of your systems or that allows us to have advanced information about photos from our gallery. On the other hand, the avalanche of promises they made two years ago about Apple Intelligence. For a start, they were already late to the advertise your system a year and a half after the arrival of ChatGPT. To continue, there were functions that did not reach the devices, others that were delayed and even had to delete promotional videos that showed something totally false. This translated into an Apple that allied itself with OpenAI to integrate ChatGPT into Siri and, in January of this year, they teamed up with Google to put a huge band-aid: Apple’s next basic models will basically be Gemini. The user will not notice it – it would be a blow to the pride of those from Cupertino – but Google accounts will. The camp. If two years ago they were late, now they are running out of reaction time. This year we are seeing AI advancing day after day with both American and European and, above all, Chinese models. Apple must get in tune and, as they point out in The Informationhave made the decision to send 200 of Siri and Apple Intelligence engineers to a several-week “camp” focused on programming tools for AI. It is something that reflects the uncomfortable reality that Apple is experiencing right now. On devices they are doing well (even with a MacBook), are establishing themselves as one of the technological pillars of the United States and They have returned to work in Chinabut in the most important race in recent years, they are still behind. Therefore, it is urgent that the Siri team, which is earning such a bad reputation, gets its act together ahead of what could be one of Apple’s most momentous launches in years. And it’s not just sending developers to camp: it’s reformulating the company. The departure of John Giannandrea – one of the leaders of Apple’s AI strategy team – left a gap that has been filled by Craig Federighi, the company’s director of software engineering. Mike Rockwell, team leader of the VisionPronow leads the new Siri team. They are two Apple heavyweights who are very much on top of the AI ​​team, which makes clear the importance that Apple is giving to this issue. 60 stay at home. Obviously, the Apple Intelligence ‘laboratories’ are not going to be deserted these weeks. As The Information points out, about 60 members of the Siri development team will remain in their positions to continue shaping the new assistant and another 60 will be in charge of evaluating performance, ensuring that it meets the standards that Apple wants to implement. Because we are no longer talking only about the quality or functionality of the assistant and Apple Intelligence, but about the ambitious privacy goal. At the presentation of the softwareApple commented that it had built a cloud infrastructure specifically for AI with end-to-end encrypted data sending and that, when that was not possible, the data would be encrypted to obscure the user’s identity. According to the company, none of them would be visible even to its own workers. The new Siri, now it is. It is evident that Apple seeks to close the gap between its assistant and what the competition has – Google integrated Gemini into Assistant months ago – but they must also close that space between their reality and the ambition they showed when presenting Apple Intelligence. Either way, this year is expected to be the year of the new Siri. According to rumors, we will see during the first half of this yearbut we have been there for four and a half months and there is no trace. Now, everything indicates that Siri will be the star of Apple’s keynote at WWDC, the great software – and hardware, sometimes – event that will be held from June 8 to 12. Meanwhile, the world of AI continues to spin, and the most curious thing about all of this is what we mentioned at the beginning: Apple has no say. We’ll see if that new Siri manages to get them into the conversation. In Xataka | Customers demand that a human solve their problem. The surprising thing is that if humans serve them they think they are an AI

Apple is dying of success with the MacBook Neo. So much so that its manufacturing is in danger

by

Apple has a problem with MacBook Neo: You are selling it too much. The first Mac with an iPhone processor is being an overwhelming success, and it hits the keys that mobilize the average user: it is cheap, it can be used for practically all uses and… it is a Mac. The problem? That this laptop has the Apple A18 Pro It is no coincidence, and that it is selling so much is a problem for the supply chain. Why the A18 Pro. Apple is not manufacturing new A18 Pro chips for its MacBook Neo, it is recycling processors from the original production. If we look at its technical details, the MacBook Neo incorporates a five-core GPU and not six. When processors are manufactured in batches, not all of them work perfectly. Some may have specific failures in one of the CPU or GPU cores. Instead of throwing them away, Apple deactivates that defective core and can sell a trimmed version of it. This allowed Apple to create a laptop whose processor was practically at zero costa pillar for the profitability of the product. The problem. The demand for the MacBook Neo is exceeding Apple’s expectationsand the stock of the A18 Pro is starting to come to an end. According to Tim Culpan, production of this device is divided equally between Quanta and Foxconn, with an initial plan to produce about six million units. As of today, suppliers are not clear about being able to produce more MacBook Neo with the stock of A18 Pro processors. The dilemma. The Apple A18 Pro is manufactured in TSMC’s N3E process, three-nanometer technology, a chip whose production capacity is practically exhausted. Among Apple’s options would be to pay a premium to order urgent batches from TSMC, something that would allow production to resume but would end the key to the Neo: manufacturing an economical product with a profit margin. The second plan involves reallocating the wafers that Apple uses for other devices to the production of the Neo, another solution that does not seem ideal. If we add to this the current storage and RAM costs, the production of the Neo becomes complicated. No solution in sight. If demand for the MacBook Neo remains above expectations, Apple will have a decision to make. Raise Neo prices? Eliminate the budget 256 GB option? Offer new colors to revitalize the product? Be that as it may, the Neo makes one thing clear: the strategy of selling MacBooks at the lowest possible price works. And even more so when we are at that point where a mobile processor is, literally, a PC processor. In Xataka | The MacBook Neo is the biggest existential threat to the Windows laptop market. And the manufacturers have no answer

Congratulations, you already program without knowing how to program. Now prepare to wait six weeks for Apple to listen to you

by

James Steinberg is a New Yorker, 35 years old, and has two professions. The first, cat sitter. The second, develop applications through vibecodinga technique in which knowing what one wants and iterating with AI manages to replace (in part) deep knowledge of areas such as software architecture or programming. Steinberg is not the exception, but the new norm in a phenomenon in which amateur programmers are saturating the software distribution system. Let them tell Apple. Wanting is power. There was a time when publishing an app on the App Store was a rite of passage for an engineer or software developer. After months of fighting with Swift or Objective-C, the app was ready and all that was missing was the blessing of the App Store and its strict terms of use. Today that wall has fallen, because since the vibecoding has appeared, the creation of software is no longer about being able to do things, but about wanting to do them. However, this democratization of programming comes at a price: before the problem was writing code, but now the bottleneck is get the App Store to validate it. The growth rate of apps published in the App Store has grown extraordinary since the end of 2025. The impact of vibecoding is evident. Source: BI. The explosion of agentic software. Data from the consulting firm Sensor Tower confirm that we are facing an extraordinary situation. In January 2026, the volume of new apps launched in the App Store in the US grew 54.8% compared to the previous year. A very similar figure had already been recorded in December: a 56% increase compared to the same month in 2024. Here there is not suddenly a batch of experts fresh out of university programming as if there were no tomorrow, but rather a bunch of “amateur programmers” who have used vibecoding to program their apps in a matter of minutes or hours and who have uploaded them to the App Store. Apple has a problem. When Steinberg or any other developer tries to publish their app on the App Store, they run into a problem: Apple’s validation process is dragging out and the average wait time is around six weeks to achieve the desired “green light.” Apple, aware that this saturation can damage its reputation, has wanted to come forward with figures to calm the market’s spirits. Apple says one thing, developers another. According to the company, 90% of the proposals it receives from all these programmers are reviewed in less than 48 hours, and the average wait is, according to the company, 1.5 days. In the last twelve weeks, Apple employees have analyzed more than 200,000 weekly shipments, which seems to make it clear that, at least according to them, the bottleneck is not that big. The developers don’t seem to be of the same opinion, and in forums and social networks there is talk of how reviews of existing updates take up to a week and new releases enter a kind of administrative limbo that exasperates this new legion of programmers. Apps that are AI Slop? A potential reason for this slowdown in deadlines may not only be the quantity of apps, but their quality. Both among traditional programmers and probably within Apple itself, there is a fear that this new batch of apps “vibecodeadas” is largely another variant of the “AI slop” or “AI Slop” that has already been presented in the form of images or videos. For some experts, many of these apps are mediocre, have been generated with little supervision and simply seek to monetize search niches. The strict terms of the App Store may be criticizable, but they are a kind of retaining wall that could flood the App Store with absolutely irrelevant apps. The App Store facing the dilemma. Forrester analyst Dipanjan Chatterjee indicated in Business Insider that “this is not a problem that Apple can get out of by rejecting apps. As AI accelerates the creation of applications, the company will have to evolve from artisanal surveillance to curation at scale.” Or what is the same: either Apple automates part of the process, or waiting times will continue to increase. The other option: tighten the entry criteria for apps created with AI so much that it disproportionately penalizes the developers who use these tools… of which there are more and more. Wanted vibecoder. What seemed like a hobby for hobbyists is becoming an increasingly striking economic ecosystem. According to Business Insiderplatforms like Lovable already publish job offers in which they are looking for “vibecoders professionals”, which seems to validate this new type of programmer, no matter how much the traditional market criticizes him. But. This avalanche of applications created with AI may be striking, but comments from professional developers usually agree on the same thing: these apps are more difficult to maintain in the long term. Even Linus Torvalds, who had partially fallen into the networks of AI, I warned him: “AI will be a tool, and it will make people more productive. I think vibe coding is great for getting people to start programming. I think (the code it generates) is going to be horrible to maintain… so I don’t think programmers will go away. You’ll still want to have people who know how to maintain the output.” Image | James Yarema In Xataka | Vibe coding wants to help Open Source. But developers don’t want AI botches

is ceasing to be the ‘Chinese Samsung’ to be something more similar to ‘the Chinese Apple with a car’

by

Xiaomi’s 2025 has been a record in several aspects, but also the certification of something that we had been seeing coming for a long time: the end of the Xiaomi that we knew. And it gives way to a new, much more interesting Xiaomi. Why is it important. For years, Xiaomi was the company that made the margins of Apple and Samsung a war to fight. His promise was, above all, the price. Now, for the first time in its history, the smartphone segment has decreased by 2.8% in revenue while the electric car and AI segment has grown by 224%. The company that built its identity on bargain He has started talking about something else. The panoramic. Total revenue in 2025 exceeded 450 billion yuan (about 57.7 billion euros), 25% more than the previous year and the first time that the company has surpassed the 400 billion barrier. Adjusted net profit reached 39.2 billion yuan (about 4.95 billion euros), an all-time high. But the real headline is in the composition of that revenue: a year ago, the smartphone and IoT device business represented 91% of the total. It has now fallen to 76.8%. Fourteen percentage points in a single year is too abrupt a drop not to assume that we are facing a different scenario. Between the lines. The segment that Xiaomi calls “smart electric vehicle, AI and other new initiatives” has achieved its first year with positive operating profit: 900 million yuan (about 114 million euros). The figure seems modest, but in reality it hides an intentionally opaque financial architecture. That same segment has increased its operating expenses by 87.7% year-on-year, to 24.8 billion yuan. Included are the costs of the car, but also the billion-parameter MiMo language modela robotics program, the development of own chips and the AI ​​agent platform Xiaomi miclaw. That is to say: the profits from the car are financing the company’s AI bet. And in 2026 that balance could be broken: Xiaomi has committed 16,000 million yuan (about 2,020 million euros) only in AI and “embodied intelligence” this year, part of a three-year plan of 60,000 million. The contrast. While the car moves forward, the phone moves backwards. The gross margin of the smartphone segment has fallen from 12.6% in 2024 to 10.9% in 2025, and in the fourth quarter it plummeted to 8.3%. The reason is the memory crisis: the demand for AI data centers has generated a bullish supercycle in DRAM and NAND prices which is swallowing the profitability of any mobile manufacturer. In the end, the same AI boom that Xiaomi is trying to capitalize on is what is eroding its core business. The company that financed its expansion based on tight margins in mobile phones now discovers that those margins are unsustainable precisely because of the trend it wants to lead. For years, the label that best defined Xiaomi was “the Chinese Samsung”: a company with a very wide range of products, presence in all price segments and a business model built on volume. Now the accounts point in another direction. The growing weight of the ecosystem of services on a base of premium hardware, the car as an aspirational extension of the brand and the own AI models integrated into all devices draw something more similar to Apple: a closed ecosystem where the hardware is the gateway and the services are the margin. The CEO of Ford already drew this parallel. With the difference that Xiaomi also makes the car. Apple doesn’t do that. The context. This shift has not come overnight. We have been seeing for years how Xiaomi patiently built its premium jump, first with Leica cameras, then with a SU7 that aimed directly at Tesla and Porsche. What the 2025 results confirm is that this repositioning is no longer a declaration of future intentions: it is the present financial reality of the company. One detail: 60% of buyers of the SU7 They are iPhone users, a sign that Xiaomi is capturing the consumer who pays for ecosystems, not specifications. The big question. Can a single company simultaneously maintain an under-pressure smartphone business, scale an electric car operation with some fiscal uncertainty, and fund an AI program with indefinite to delayed returns? The 754 million monthly active users and the 1,080 million connected IoT devices that Xiaomi has are an argument for optimism, but maintaining three demanding fronts at the same time, with the business that finances them under siege, is the great challenge that Xiaomi has ahead of it for this new stage. In Xataka | Leica is teaching Xiaomi everything it knows: when the student no longer needs the teacher, the agreement will have fulfilled its function Featured image | Xiaomi

Apple has never liked anyone tampering with its machines

by

In 2019 Apple put on sale a spectacular Apple Pro that in its most ambitious version had an equally spectacular price: 62,648 euros. That equipment was still based on Intel chips, but it offered something unusual for Apple: the possibility of easily exchanging some of its internal components. Today Apple has removed it from its website without a press release, without farewells or tributes and without plans to replace it. 20 years later, the Mac Pro is dead. The slow agony of the Mac Pro. Abandoning this product has not been a surprise at all. It had been three years since the Mac Pro he had been “paralyzed” with the M2 Ultra chip while the rest of the Apple Silicon-based range moved forward. And the amazing thing was that Apple I kept offering it from 8,399 euros with that chip, 64 GB of unified memory, 1 TB of SSD and of course that characteristic tower format box with that “cheese grater” front that it was a meme in itself. Of expansions, nothing. This team was different precisely because of that capacity for expansion. It had PCIe slots for specialized cards, and the ability to add storage. That offered some flexibility for professionals who had specific needs that weren’t available on any other Mac. The surname Pro is blurred. In reality, Apple has been telegraphing this farewell for some time. The Mac Pro began to be out of stock months ago both in physical stores and online, but also a few days ago Apple abandoned Pro Display XDR monitors to simply call their new monitors Studio Display. The Pro surname disappears from the range of desktop computers, although it remains important for the rest of the catalog: we have MacBook Pro, we have iPhone 17 Pro/Max, and we have iPad Pro. And it does not seem that those surnames are going to disappear easily here. The M1 changed everything. The problem is that the departure of Apple Silicon and the revolutionary Apple M1 chip He went directly in the opposite direction. The unified memory on the chip makes it impossible to expand, and the M2 Ultra does not support external graphics. The PCIe slots remained the only argument, but they became something with very little meaning because the Thunderbolt 5 port already offered expansion capabilities to external peripherals. The evolution of the design, with that 2013 Mac Pro with the “trash” design, was fascinating, but not necessarily a success. The Mac Studio was the new Mac Pro for years. In reality, the death of the Mac Pro occurred rather in 2022, when Apple presented the first Mac Studio. From that moment on, the Mac Pro proposal was in questionand disturbing comparisons confirmed it: a Mac mini M2 for 719 euros doubled in performance single-core to a 2019 Mac Pro. The Mac Studio was comparatively cheaper and much more powerful, and that made one question inevitable: does the Mac Pro make sense? Apple just definitively answered that question. Three years later, yes, but he has done it. Little Computing Beasts. With its chips from the Pro, Max and especially Ultra ranges, Apple has shown that it is not necessary for a computer to be “big” to prove to be a computing beast. The Mac Studio have proven this for years, and the versions with the M3 Ultra and 512 GB of unified memory are extraordinary machines that today are not only perfect for “regular” professionals, but also for those who work, for example, with AI models locally. These chips have managed to demonstrate that the expansion capacity offered by the Mac Pro is a secondary argument, at least, for Apple. And therein lies the crux of the matter. Don’t touch our machines. Apple has never made it easy for users to upgrade or even repair their devices. Although in recent times it has taken positive steps in that direction, not even those “facilities” are worth it. The company’s obsession with total control of the software and hardware of its devices made the Mac Pro a “dangerous” product for them, because it opened unexplored paths that they surely never wanted to travel. By removing the Mac Pro and focus everything on Mac Studio They regain that control, because Mac Studio cannot be easily expanded. It is possible change SSD driveas in the Mac mini, but it is not a task “for all audiences.” In Xataka | The new Siri will not be Gemini with another face. Apple has helped Google to build what it could not do alone

The new Siri will not be Gemini with another face. Apple has helped Google to build what it could not do alone

by

The new Siri, according to rumors, was going to land with iOS 26.4. This version arrived on compatible iPhones yesterday and, to no one’s surprise, there is no trace of Gurman’s prediction. What we have woken up with is new details about the agreement between Google and Apple related to access to Gemini. And there are interesting details. The agreement. Quick context: Apple and Google have teamed up to give Apple access to Gemini. The company has been promising for years that Siri, integrated into Apple Intelligence, will be an assistant that lives up to expectations. But after delays and more delays, it became clear that Apple needed help. The multi-year collaboration allowed Apple Foundation Models to rely on Gemini models, running on the platform Private Cloud Computing from Apple. Beyond this, no details about the agreement were revealed. The new. According to The Informationthe collaboration between Apple and Google will be somewhat deeper than expected. So much so that Apple would have full access to the Gemini model within its facilities. One of the company’s main purposes would be to produce smaller models designed to run locally and oriented to specific tasks within Apple devices. Distillation. Apple would not have agreed with Google to access Gemini with an integrated Siri interface. The objective is to use the main model to “distill” more efficient models, with lower requirements and fast operation. In other words and based on this information, Apple has made clear Google’s superiority in its AI models. So much so that he has needed to access them directly to be able to create the solutions that he has been promising for two years. What’s coming. According to Gurman, Apple is finalizing the changes to Siri to present it on June 8 at its WWDC, the developer conference it holds annually. In it, we will supposedly see Siri as a chatbot integrated into iPhone and Mac, as a real alternative to ChatGPT, Claude and Gemini. Late, very late. Apple’s problem is not being late with AI. It is coming at a time when giants like Claude iterate practically daily and when it is more than difficult to surprise the world. All the promises of Apple Intelligence, that contextual Siri, and that deep integration with the phone are already achieved by some of its rivals, since Apple has been waiting for two years. The big question is whether or not it will be worth the wait. In Xataka | Apple confirms the date of WWDC26 and hints at something important: AI will not be the only focus

Log In

Forgot password?

Forgot password?

Enter your account data and we will send you a link to reset your password.

Your password reset link appears to be invalid or expired.

Log in

Privacy Policy

Add to Collection

No Collections

Here you'll find all collections you've created before.