cameras

Razer has had a crazy idea and that is to put AI cameras in headphones. I have tried them and they have given me something to think about.

by

Project Motoko. I like to think it’s a reference to Motoko Kusanagi, the protagonist of ‘Ghost in the Shell‘, but in any case, that is the name given to Razer’s new concept. Indeed, they are headphones with two cameras and artificial intelligence whose proposal is quite interesting: what if, instead of smart glassesShould we wear smart headphones? The company has taken advantage of the MWC 2026 that took place these days in Barcelona to show them and I have had the opportunity to get my hands on them at the Qualcomm stand (we will see why later). At the moment, the prototype, because that’s what it is, a prototype, has certain rough edges to iron out, but I really liked the underlying idea. Let’s go in parts. Project Motoko by Razer | Image: Xataka The background idea. As is obvious in the photos, I wear glasses. Normal glasses, although prescription ones. If I wanted to use connected glasses I would have to change my glasses and buy a frame, which is not cheap, in addition to prescription lenses. Well, like me, half of the world’s population. That is to say, smart glasses have a small penetration problem: They have to convince glasses wearers to change their glasses. They have to convince those who don’t wear glasses to wear glasses. Razer’s idea. It may be easier to convince the user to use smart headphones instead of glasses. These devices are agnostic about whether people see better or worse and, in reality, they can offer a similar and even better experience in certain aspects, because being larger they can offer more autonomy and power. Currently, the Meta Ray-Ban 2 They move in the eight-hour range, for example. This is what the Project Motoko prototype looks like | Image: Xataka The trade-off, of course, is wearing big headphones all day. They are less concealed and you are not going to wear them at important moments in your life (or yes, we listen but we do not judge). Be that as it may, the glasses have an advantage there, but that does not make Razer’s proposal make any less sense and may even have a fit not in gaming or in everyday life, but in terms of accessibility. What is this about?. Project Motoko are over-ear headphones (quite comfortable, I must add) with two 12-megapixel wide-angle cameras at eye level, one on each side, and several far- and near-field microphones. It’s like having a pair of eyes connected to AI that see what we see. The experience will obviously vary depending on whether we are paid or free users of chatbots. Instead of using proprietary AI, the device can connect to all platforms, namely Grok, ChatGPT, Google Gemini, and even Perplexity. Part of the process is done in the cloud, but thanks to an undetermined (for now) Qualcomm chip, there will also be local processing capabilities for certain commands. The cameras are at eye level | Image: Xataka The operation is simple. You look at something, say a restaurant menu; You ask the AI ​​something out loud and it answers you. During the demo we asked the headset if an ingredient on a table was suitable for lactose intolerant people, and even what we could make with the objects in our inventory in ‘Minecraft’, and it responded without problems. It also recognized buildings, places and text, translating a Japanese menu and giving us recommendations based on our preferences. The prototype is still missing, but it works, it works. Razer is still ironing out some connectivity and interaction issues, but the company is positive that they will release it at some point. They are not clear when, but the product is moving in the right direction, as explained by Razer. Detail of the position of the camera and microphones | Image: Xataka The rough edges. The demo had some flaws, such as the headphones were not capable of recording live video and did not capture the image if we did not ask them to, let me explain. To generate a recipe with the ingredients on a table, you had to expressly tell it to take a photo and then the command. That is not natural language. It is not natural to say “take a photo and tell me yes”, but a normal interaction would be “hey, what can I do with this?” The idea is that we invoke the AI ​​using a button located on the headphones, so it would make sense that, in a final product, when you press that button the headphones begin to record the live image. Not a static one, but a video feed like Gemini Live does. And in that sense, the warning for third parties that they are being recorded with the headphones is not defined at the moment either. A white light turns on in Meta’s glasses when you record, for example. In any case, it doesn’t seem like something that can’t be fixed via software for a final product. The release date is not confirmed, nor is the price. Project Motoko | Image: Xataka Maybe the chicha is not in everyday life. Although it is tempting to think of a companion product for everyday use, especially if you work with headphones or usually wear them on the street (not my case), where I think Project Motoko could have a huge impact is in two areas: video generation to train humanoid robots and accessibility. On the one hand, headphones capture what we see (more, in fact, as they have a greater field of vision), so by recording how a manual industrial process is carried out, the necessary resources could be generated to train machine learning algorithms focused on robots. After all, an AI learns by watching the same action thousands, millions of times, but for that to be possible it has to have videos, many very specific videos which, of course, are not abundant. On the other hand, people with vision problems have a powerful ally in … Read more

Xiaomi’s high-end is back on sale with a bang. A powerful mobile with 1 TB and Leica cameras

by

Until next January 23, MediaMarkt will have its campaign active Downhill. Among all the offers there are some that have caught our attention, such as the Google Pixel 10 Pro or, in this case, that of Xiaomi 15T Pro of 1 TB, which has dropped in price to 799 euros in what is one of the best offers that the store has launched on this mobile to date. Other stores like amazon either PcComponents They have also lowered the mobile phone to the same price. The price could vary. We earn commission from these links A mobile with 1 TB He Xiaomi 15T Pro It is a high-end mobile from the brand that stands out both for what it offers at the hardware level and for the price, especially if we take this offer into account. At the design level it does not differ much from the previous generation, but it does. It comes with a different technical sheet, and quite good. It features a large 6.83-inch screen that offers both 1.5K resolution and a 144 Hz refresh rate. It is compatible with Dolby Vision and HDR10+ and inside we find the MediaTek Dimensity 9400+ processor along with 12 GB of RAM and 1 TB of internal storage. Its 5,500 mAh battery supports 90W fast charging and 50W wireless charging, its speakers are compatible with Dolby Atmos and its camera module, in addition to having the Leica collaborationis made up of a 50 MP main sensor, a 50 MP telephoto lens and a 12 MP ultra wide angle lens. You may also be interested XIAOMI Watch S4, Bluetooth Version, Advanced Professional Sports Mode, Quick Change 2.0 Bezel, 1.43 Inch Circular AMOLED Screen, Heart Rate and Sleep Monitor, Black The price could vary. We earn commission from these links XIAOMI OpenWear Stereo Pro, Open Ear Headphones, Bluetooth with Hook, Comfortable and Stable Use, 45h Battery, IP54 Water Resistant, Multiple Drivers, Hi-Res HiFi Audio, Titanium Gray The price could vary. We earn commission from these links Some of the links in this article are affiliated and may provide a benefit to Xataka. In case of non-availability, offers may vary. Images | Xiaomi In Xataka | The best mobile phones, we have tested them and here are their analyzes In Xataka | The best Xiaomi mobile in quality price: purchasing and comparison guide

to what extent will we accept that AI cameras fine us

by

Spain has begun to automate the surveillance of minor infractions with AI: Technology stops optimizing traffic and starts monitoring it. Why is it important. This marks a paradigm shift. Until now, urban AI was used to improve mobility (adjust traffic lights, predict traffic jams, reduce emissions). Now he goes from assistant to inspector. And it does so with a key nuance: it does not pursue major crimes or flagrant dangers, but rather small daily infractions that previously escaped control due to cost and surveillance capacity. AI reduces the marginal cost of sanctioning to practically zero. Once deployed, everyone can be observed all the time. The facts. The Barcelona pilot test involved four buses of lines H12 and D20 equipped with cameras that identify, through AI, vehicles blocking reserved lanes. In Madrid, the City Council has installed smart traffic lights that count pedestrians in real time and has announced systems that will detect seat belt use. The DGT has taken another step. It has deployed four cameras on the A-1, A-2, A-6 and A-42 highways that monitor the crossing of continuous lines. The system works with two cameras per section: one records the license plates at the beginning, another at the end. If a car changes lanes between both points, the fine is automatic. It is 200 euros per violation. In figures. Spain already has 3,395 devices to control violationsaccording to Faconauto. Of them, more than 1,300 are DGT surveillance points between fixed and mobile radars. Added to this are more than 200 cameras that monitor belts and mobile phones, Pegasus helicopters and now these new continuous line detection systems. Barcelona has not yet activated the sanctions on its buses, but the volume of violations detected (80 daily in just four vehicles) anticipates what is coming. Between the lines. There is a delicate balance that is being renegotiated without us having barely opened the debate. On the one hand, more compliance with fewer agents: administrative efficiency is indisputable. On the other, the sensation of an omnipresent eye. The difference with the classic radar is not so much technical as range. The radar monitors specific points where there is proven risk. These new systems turn the entire city into a guarded zone: each bus is an inspector, each intersection a control point. AI does not change what is sanctioned, it changes where and how much. Move from selective surveillance to ubiquitous surveillance. Yes, but. To what extent will citizens accept being recorded and punished by a machine? It’s not just a legal issue, but a cultural one: trust in the algorithm versus human interpretation. Who audits the system’s decisions? What room is there for error or appeal? Technology is not neutral: each deployment reflects political priorities about what deserves to be monitored and sanctioned. The big question. What is relevant is not whether this is good or bad in the abstract, but what it tells us about the new contract between citizen, city and AI. AI stops being an abstraction and enters the daily urban experience. The citizen goes from user to observed subject. And the unresolved question is who sees, who decides, who corrects and, above all, how far we are willing to go when automating the chase is so easy and cheap. In Xataka | The “made in China” business of the DGT’s V-16 beacons: homologating the same product 24 times and selling it under different brands Featured image | Barcelona City Council

Barcelona is tired of cars not respecting the bus lane. So it’s installing AI cameras on its buses

by

2,500 violations in one month. That is, Barcelona City Council could have fined more than 80 drivers every day. It has not done so because, at the moment, the project is in the testing phase, but it has served the City Council to confirm its suspicions: the invasion of the city’s bus lanes is a constant and a real problem. To confirm this, Barcelona City Council has launched a pilot test equipping four buses with cameras enhanced with artificial intelligence. They are units of lines H12, which crosses the city on Gran Via de les Corts Catalanes, and D20, which in this case covers the entire route of Avinguda del Paral-lel and a nearby section around the port. The project was already presented last February but it wasn’t until summer when we knew all the results. Four buses and 2,500 violations The data published suggests that something is not working in the Catalan city. They collect in Hybrids and Electrics that just four units collected more than 2,500 violations in just one month. These infractions have not been sanctioned because, as we said, it was a pilot project that will serve to define new strategies and look for solutions. The truth is that it is obvious that the Barcelona transport company has to deal daily with all types of vehicles that illegally occupy a space that should serve to prioritize the passage of public transport. To record these infractions, the cameras recorded the movement and position of the vehicles as the bus moved forward. They specified in the presentation of the project that cameras have not been used in the study to identify people or license plates. The software, they say, only counted if there was an object blocking the road and is capable of discerning illegal occupations, such as a parking lot, from those permitted (taxis, right turns…). With the data they want to propose solutions among which, of course, the possibility of turning buses into real moving radars has not been mentioned. Although buses have highlighted the problem, the truth is that Barcelona already has the so-called “multacar”cars with cameras that record violations of the bus lane and that, these do, issue penalties with the violations found in their path. And the city council also has similar vehicles that They control regulated parking areasinstantly checking if a car has exceeded the allowed time and, in that case, fining it. The big difference is that installing this system on the buses themselves would allow the fleet to become an exercise in constant surveillance throughout the service. And taking into account the data recorded, the volume of sanctions would be expected to be very high in that case. Photo | TMB In Xataka | FlixBus wants to operate international bus lines in Spain. He has encountered two enemies: Alsa and Avanza

We believed that the iPhone 17 and the Air shone by cameras and design. We have just discovered that they hide an exclusive security function

by

For a long time we have lived with the illusion that there are impenetrable computer systems. The reality is less resounding: in security, everything is reduced to how much effort, time and resources it requires to force a lock. Just as it is not the same to open the door of the house as the vault of a bank, in the digital world There are more or less resistant barriers and unexpected shortcuts that avoid brute force. The objective of the defense is not perfection, but to raise the toll to break it is impracticable. From there, the risk never disappears, it is managed. With that practical look, Apple has been adding layers to make every step of the attacker more and reduce its maneuvering margin. According to the Cupertino companythe most sophisticated exploitation chains that have observed against iOS come from the mercenary spyware and rely on memory vulnerabilities. Although they do not explicitly mention it, they surely refer to threats such as Pegasus of the company NSO. And the answer they have raised is a new piece in that wall: a reinforcement that integrates hardware and system to monitor the integrity of memory and cut overwhelms or undue accesses before they thrive. Memory Integrity Enforcement on iPhone 17 and iPhone Air Apple has presented Memory Integrity Enforcement (Mie) as part of the new iPhone 17, iPhone 17 Pro and Pro Max and iPhone Airan integrated memory defense directly in its hardware and operating system. This development is the result of five years of joint work among their teams of Chips and Software Engineeringwith the aim of drastically raising the cost and complexity of attacks based on memory corruption. Mie promises to act continuously and transparently, covering critical areas such as kernel and more than 70 processes in user space, all this without compromising energy consumption and device performance. The Miene nucleus combines several layers that work in a coordinated manner to reinforce security. The typated memory assigners are systems that organize the data according to their type, as if each object class had a specific drawer. This organization makes it more difficult than an error in a program allows one data to overwrite another. If a failure occurs, the system can detect it before it becomes an attack. On this basis acts the Enhanced Memory Tagging Extension (EMTE), a hardware technology that adds an extra layer of memory control. Emte works by assigning a “secret label” to each memory block. Every time an app or the system wants to access it, you must present the correct label; If it does not coincide, Hardware blocks attempt And the system can close the process. This permanent and synchronized check allows to detect and stop classic attacks such as buffer overflows or use after release (USE-AFTER-FREE), which are usual techniques to take control of a device. The allocators protect the use of large -scale memory, while EMTE provides precision to the smallest blocks, where the software itself does not respond with the same effectiveness. This permanent and synchronized check allows to detect and stop classic attacks such as buffer overflows The bet responds to a landscape of threats where the highest levels against iOS are faces, complex and directed, historically associated with state actors. These chains usually share a common denominator: they exploit interchangeable memory vulnerabilities that have been present throughout the industry. The intention of Mie is cutting the progression in early stages, when the attacker still has little room and depends on chaining multiple fragile steps to gain control. Apple graph showing real exploitation chains and the points where it blocks them The scope of protection includes kernel and extends to key system processes that are usually entry objectives. In addition, Apple makes available to developers the possibility of testing and integrating these defenses through the Enhanced Security option in Xcode, including EMTE capabilities in compatible hardware. That is especially relevant to applications where a user can be direct objective, as messaging or social networkswhich often appear at the beginning of the exploitation chains. To sustain the labeling and synchronous check -up without perceptible impact, Apple redesigned the A19 and A19 Pro allocating CPU area, CPU speed and memory for label storage. The company precisely modeled where and how to deploy emte, so that the hardware meets the demand for checks. The software, on the other hand, takes advantage of the assignments typated to raise the bar of protection against memory corruption, while the hardware assumes fine verification. As we point out above, this should maintain the expected experience in performance and autonomy. The project was evaluated with its offensive research team from 2020 to 2025. First with conceptual exercises, then with practical attacks in simulated environments and, finally, on hardware prototypes. This prolonged collaboration allowed to identify and close complete exploitation strategies Before launch. According to Apple, even trying to rebuild known real chains, they failed to restore them reliably against Mie, because too many steps were neutralized at the base. Even so, Apple remembers that perfect security does not exist. Very rare cases could survive, such as certain overflows within the same allocation. For previous generations without EMTE support, the company promises to continue expanding software -based improvements and safe memory allocatives, with the aim of bringing part of these benefits to previous devices without affecting its stability. Ultimately, Mie does not eliminate riskbut it does redraw the rules of the game by raising the cost and difficulty of memory corruption techniques. For those who buy an iPhone 17 or an iPhone Air, this translates into always active protection and, according to Apple, invisible for the user. Images | Xataka with Gemini 2.5 In Xataka | Or pay or we will use your works to train AI: the threat of hackers to an artist website In Xataka | How to change all our passwords according to three cybersecurity experts

I’ve been using Yi security cameras for years. It was delighted until the app became an advertising hell

by

More than five years ago I bought my first Security Chamber To monitor my cat when he got sick. It was YI brand and it worked great; He looked good, had movement alerts and could speak through the mic. With the passage of time I had more cats and I bought two more cameras of the same brand to cover the rest of the house. The problem is that The APP Yi Home has added advertising in its app. How much? All. It is not an exaggeration I was traveling recently and I entered Yi Home’s app to look through the cameras. I don’t lie if I say that More than once I have made me want to launch the mobile. On the main screen there are already a few ads spread there, but that is not the worst. Some of the ads that appear when I open the app or when I try to see one of the cameras. The worst are the Full screen ads. They cannot be skipped until they pass a few seconds, the button to close them is tiny and sometimes it does not work, causing it to end in the App Store, or wherever the announcement in question link. So every time I open the app. But the thing is not there, when entering each camera, advertisements also leave full screen. When you finally manage to see a camera, a banner comes out that covers the controls to be able to move it *Chef’s Kiss*. Few apps I remember that they have bombarded me both with advertising and Yi Home, although it is not the only one. A current example is Capcut, the Tiktok videos editing app. Every time I open the app, I get a full screen advertisement, but it is also that if you leave a moment and enter again, another appears. Not to mention that The entire app is a mines field with ‘Pro’ functions And they don’t stop insisting to join. Pay or suffer Yi Home has never been a super clean app. I already had some Banner type ads and has always shown me pop-ups to join the payment plan. The thing now has no name. In addition to full-screen advertising, those pop-ups continue to come out and remind me that, If I pay, advertising will disappear. On top with picn. The summum of despair. (The controls can be relocated, but was there no other place?) The cheapest payment plan if you have several cameras costs 79.99 euros a year. It seems expensive, but I think that although it would cost a tenth, I wouldn’t pay either. The thing about this app is A manual of how to lose customers. At least I know I am not the only user discontent, In Tustpilot is full of negative reviews on the same subject and in This Reddit thread There are also several angry users. In the same thread they also comment on several ways to eliminate ads such as installing previous versions of the app or configure a DNS with Ad-Block. In my case I take another way to eliminate ads. I change my chambers I had been wanting to buy a while Another 360 camera And obviously I was not going to buy another of the same brand with the problems I was having with advertising in the app. One of those days when I almost crashed my mobile, gave me to get into Amazon to see other cameras and make the decision. TP Link Tapo C210 After reading a few reviews and making sure that his app did not seem like an online casino, I opted for the Top-Link tapo. I found a Offer pack with two cameras 360 And I bought it. When I set them up, some notifications came out within the app to go to the Premium version, but I have been using it for a week and they have not reappeared. Come on, what was Yi Home’s app a long ago and that It should be normal in an app of a security camera. Images | Amparo Babyloni, Xataka In Xataka | It is not you, YouTube is filling with more and more ads. Especially if you see it on a smart TV

While giving us mobiles with better cameras, more power and a flood of AI, I just want them not to die in summer

by

In recent years we have seen huge advances in smartphones cameras, in His brilliant screenshis fast charge, Your powerand what they manage to do thanks to the progress that artificial intelligence has brought to the sector (now, not when it began to promote itself eight years ago). And yet, in a summer as hot as we are happening, what I miss most is being able to take advantage of all those functions. Because as temperatures are taking place since May in Andalusia, the reality is that I can’t. The problem. When the heat arrives, the girls fall in love and Solar panels have a fatal. It is also the worst enemy of modern smartphone, whose current functioning completely modifies. With high ambient temperature, smartphone begin to suffer thermal strangulation or Thermal Throttling. It is a technique that reduces the performance of the device so that it does not continue to warm up. What do we notice? In which everything is slow, getting to cause Lagand slowdown accused even in high -end mobiles that do not suffer with normal ambient temperatures even when we demand a lot. It is also common for smartphones to begin to deactivate functions that require greater consumption and therefore heating, such as high rates of soda (90 Hz, 120 Hz 0 more), fast charge or high brightness on the screens. The latter is especially unfortunate, because it is on sunny summer days (when we have more hours of sunshine) when we need the record shine peaks that manufacturers promote with each recent launch. The situation on extreme heat days. The last great heat of these weeks has been the perfect example of how smartphones suffer. In an encounter with friends in the field where there were smartphones of several major brands, the smartphones became unusable in the hottest hours. Fully attenuated and impossible screens to see the sun, abysmal performance, extremely hot bodies and worrying warnings of two types: impossibility of continuing to load to protect the battery, and complete smartphones blockages with a temperature warning as The one Apple details on its website. The first caused that we had to wait until sunset to load the smartphones (with the consequent complete download of several of them for hours) every day. The second, even losing calls: c on the very hot mobile, a relative called me, and the call was abruptly cut by showing a notice like the next. The situation is also improvable with more moderate temperatures. A heat wave in Andalusia is an exceptional situation and that is not lived in countless territories. However, in May, I was at an outdoor wedding about 25 degrees of temperature and decided to take photos and videos with an iPhone 16 Pro and a Samsung Galaxy S25 Ultra. The iPhone responded with Throttling, strongly attenuating the screen and slowing down any interaction with the interface. The Ultra Galaxy S25 responded something better, without being slow but also presenting little shine in its panel. The situation was paradoxical: I was using two of the smartphone with the best camera in the market, but I could barely record and take pictures, for the slowness with which they worked and because I almost did not see what I was shooting. Not because of the reflexes of the sun, but because of the little maximum shine of the screens once the terminals felt some heat. Another similar situation, changing camera for GPS, is what is lived in cars in summer: the mobile attenuates its screen and there are cases in which it costs to see the navigation interface, especially if at the same time we are carrying the terminal. And it is not trivial, we load them using the GPS because the consumption shoots and on a long trip the battery does not endure. Keys to protect them from heat. The solutions we have as users are, unfortunately, limited. But There are things we can do. For example, minimize its use in the sun. In summer, as we pointed out, it will be difficult, but there will be multiple situations in which we can place ourselves in the shade. Or what is easier: always leave the mobile in the shade, instead of at very hot points from a place in the sun, on the beach or in the car. They are not few occasions in which I have remembered acquaintances that his mobile would be burning because they had forgotten the sun. We can reduce camera use, especially when necessary. Sometimes, doing tourism, we tend to leave the screen on with the camera in use, preparing for a shot that sometimes does not arrive. It is in those situations when the terminal turns off until we have a shot to shoot can save a lot of heat. House and car tricks. My maximum advice is to use the terminal as GPS subject with a support for air conditioning ventilation grilles, so that it will be cool. It will load at normal speed and work as a browser without major problem. It is a advice that applies even if we use something like Carplay or Android Auto, as the vehicle will continue to resort to the terminal GPS, with the consequent heating. At home, two other tips that I have put into practice: use Cooling with integrated fan or even Cold accumulators covered with a trap that also greatly reduces the temperature of the smartphone when they come into contact with them. Thus they are trying to mitigate the problem. Manufacturers lead a lifetime fighting heat. The problem is that, at the same time, they are integrating components that, although they are more efficient than in previous generations, also have higher final consumption, especially when we think that more and more cameras, more bright screens and chips capable of Move AAA games (although The best is pulling streaming). He Xiaomi 15 Ultrafor example, he convinced us for how his cooling system with Steam camera three -dimensional manages to tame the Snapdragon … Read more

The mobile industry has been desperate for years to improve its cameras. Honor believes having found the key

by

A few weeks ago I had the opportunity to try the Honor 400 Pro. It is a mid-high-end phone with the focus on mobile photography. So far, there is nothing especially surprising. But if I tell you that this phone sets the bases of a Honor who does not believe that the future of photography is in the hardwarebut in cloud computing, things may change. In Xataka we have been able to talk to Luo Wei, Chief Imaging Architect of Honor. He is the one in charge of leading the photography part, applied to it and super zoom in honor, and ex-huawei. And someone who is clear that to win the smartphone career you have to win the AI ​​race. If marketing sounds like you can be calm. It is not, it is the vision of the future of a giant who aspires to conquer Europe. Honor wants to lead in Europe, and wants to do it with ia To understand why AI for honor in phones is so important and not misunderstand it as a marketing exercise, it is necessary to know the relevance that the company has with its Alpha plan. In March 2025, the company revealed a investment of 10,000 million dollars In five years to promote artificial intelligence on their devices and display your first AI agent, in association with Qualcomm and Google. The objective is clear: lead the AI ​​agents in the mobile ecosystem and be the main reference in this race. At the moment, he will do it hand in hand with these two partners in Europe. Honor has different strategies for China and Europe. In our territory, Google Cloud is the main partner It is also relevant to understand that this plan has a lot to do with The position of honor in Europe and how it aspires to conquer our market, more than a proposal centered purely on your native country. With counterpoint data in your hand, Honor is the brand that grows most in shipments in Europe in an interannual wayand wants to continue pushing leading the functions of AI. In China, its AI approach is different, deeper and more local. There the partner is not Google Cloud, they are national giants like Alibaba (Qwen + Wanxiang), Baidu and Tencent Clouddeepening much more in your Yoyo assistant and hand with Depseek. All of AI One of the pillars of this plan, in addition to the creation of the first agriculture for smartphones, is Aimage. Landing the term to the tangible: the high -end phones of honor process the photographs in the cloud to “improve them”. And about this there are many questions. The first is why. Here Luo Wei states that, historically, Mobile photography has been limited by hardware. A limitation that according to Honor can only be overcome using cloud computing, especially for tasks such as image restoration, provide movement (turn into short video) static images or improvements in the zoom. Being more concrete, 12,500 million processable parameters for its cloud model vs. 3,000 million parameters with which current hardware can deal. Adding to this point the possibility that stresses honor of updating its algorithms via Ota. It’s something we have talked about in Xataka: We have been watching a stagnant industry at the photographic level and with a clear physical limit when improving your hardware. Currently, there are no larger sensors. The second point is in the democratization of functions such as Super Zoom (Zoom improved with AI) on devices beyond its flagship. Creating a cloud platform from which you can drink lower ranges models allows you to honor an experience similar to that of a high range in cheaper models. “The algorithms have been increasingly improving since December last year, when this function had just left. And if you compare the future with what is now, five months later, you can see that the cloud -based algorithm is updated every two weeks is becoming remarkably more real.” For honor it is also especially important The improvement margin through updates that allows AI. Honor takes it seriously: they update it every two weeks. The key here is not alone in time, it is in how. System updates can be installed or not, depend on storage space and hardware requirements. Luo says that, using the cloud, they could update the algorithm daily if they wanted and access to it would be immediate. The second great unknown has to do with profitability. Using Google’s cloud sounds good, But someone has to pay it. Here is surprising the transparency of Luo, who explains that honor is still debating what the business model will be and if they will end up charging users or not to use these services in the cloud. Here, Luo explains that honor does not pay for users or requests, but for QPS (consultations per second). In other words, for the ability of users to access the cloud service, and thus their costs do not depend on user behavior. It is, currently, its way of making this business sustainable, although it makes it clear that it is something that is being worked on and discussing internally. Betting everything to the cloud brings an inevitable question, What happens to the privacy of users. Honor explains that its AI cannot be used without prior consent of the user (it appears from the configuration assistant), and secondly, that all photographs that are uploaded to the cloud to be processed are eliminated after being treated. On the future of photography, Luo closes with a metaphor. At the beginning of the history of automotive, the gearbox was manual. Subsequently, it went to automatic changes and, today, We are already talking about autonomous driving. And yet, we still need drivers. For honor, the AI ​​will allow any user, regardless of their degree of skill, obtain the best photograph in any scenario. For expert users, it will be a co-pilot in editing functions, elimination of elements and post-foto work. Image | Honor In Xataka | … Read more

In the war cameras vs lidar, Tesla has a lot to learn from an unexpected product: the Chinese aspiring robot

by

Light detection and ranging. Or, what is the same, detection and measurement of light. These are the words behind Lidar. This technology uses light pulses to map the environment and discover each and every corner of a stay or an open space almost in real time, as well as to recreate 3D environments with enormous precision. To understand how a lidar radar works, I recommend watching Mark Rober’s video in which it tells the differences between a car equipped with this system and its tesla, which exclusively uses cameras to detect the obstacles that you can find along its path. Beyond the controversy arising in relation to the tests that Rober does, the video explains well Why Lidar is such a complete system. In a simple way, the system takes advantage of the speed of light to emit infrared pulses. These bounce in the object in question and the system calculates how far that object is using the time that the pulse of light has taken to go and return. As light pulses have a very small size and, as we said, it takes advantage of that very high speed of light transmission, can map the objects almost in real time. In recent days, after Mark Rober’s video, the controversy of whether it is better to use an LIDAR or one system that exclusively use cameras and recreations by software is better or worse to guarantee the best behavior in terms of autonomous driving. A controversy that has left people trying launch your tesla against a wall With a road painted in the purest shyesty style. And it is a long time since it made it clear that I would bet everything on the use of cameras for manage your driving aid systems Or, in the future, completely autonomous driving with the robotaxis that wants to put in the streets. At the time, home vacuum robots also lived this moment of indecision. And along the way, Irobot who was the leading market leader has ended up giving millionaire losses and has seen how Chinese manufacturers have eaten much of the market. The secret of the latter: they use lidar instead of trusting everything to the cameras. What can you learn from a vacuum cleaner “We have substantial doubts”, with these words Irobot owners responded to the question of whether they could move forward with their operations. My partner Javier Pastor explained A few months ago when Amazon raised the purchase of Irobot in 2022 (which fell into regulatory terms) The company had a value of 1.2 billion dollars. In the third quarter of 2024, the calculation was about 200 million dollars. Shortly after, The situation has not improved a lot. In 2024, Irobot lost more than 145 million dollars. However, they had reduced their losses by 52%. The company faced a perfect storm. After spending the worst of pandemic, spending on home robots and other products (such as computers) They collapsed. That coincided with the launch of new Chinese products that were faster and faster than Irobot’s: they used the lidar sensor. This system is ideal for improving the capacities of a vacuum robot. It detects better obstacles, mapping the house more precisely and that translates into faster and more effective work. The result is especially good if it is combined with the cameras and artificial intelligence systems to discover smaller obstacles that can go unnoticed or create confusion. In recent years, that has been the trend of the market in cleaning the home. The combination of this system with a more attractive price He has triggered the sales of Chinese aspiring robots, to the detriment of an Irobot that accumulated more than 60% market share. The comparison with Tesla and the electricity car market is evident. The latter are using Lidar sensors in their cars and offer a product equal to or better than Europeans and Elon Musk’s for a fraction of their price. He Xiaomi Su7 and its wide reception He is being a good example of this. Tesla has long since defends that Lidar are not necessary to improve autonomous driving systems but studies say that, as in the case of vacuum robots, the best performance is signed When both technologies are combinedLidar and cameras. In the case of Xiaomi, the videos shown from How your total autonomous driving system operates They are impressive, with very human behavior and managing the smallest spaces very well. This is possible because, among other things, A lidar is better when calculating distances And suffer less when the light falls. Neither does the risk of the camera be dazzled, At least not intentionallygenerating Ghostly brakes. So far, everything indicates that Tesla does not seem willing to return to anything other than the “All Chamber” To save costs. At the moment it has worked but you have to demonstrate that it is as effective as more advanced systems. And there is also the largest electric car market in the world. In China, where foreigners are suffering from the local product, Tesla is hitting a good batacazo in 2025. Photo | Tesla and Irobot In Xataka | “It’s like living millions of lives”: Tesla trusts their own drivers to advance Waymo and Cruise and shoot their value

If you are going to buy a mobile this year, attentive to the cameras of honor Magic7 Pro, but also to your first offer

by

It is increasingly common for mobiles to launch already reduced – sometimes we have to wait about two or three days to find them on offer. And the Magic7 Pro Honor It has not been any exception. The brand’s new mobile has arrived under an official price of 1,299 euros, but it can already be found in stores like Amazon for 1,099 euros. * Some price may have changed from the last review A mobile with cameras that promise a lot He Magic7 Pro Honor It is a mobile that we cannot lose sight of. The most interesting thing is that we talk about a powerful model, which comes Equipped with the best in processors for android mobiles: the Snapdragon 8 Qualcomm Elite. But, in addition, we also find other characteristics and specifications to take into account. First, its 6.8 -inch screen offers a good resolution of 2,800 x 1,800 pixels and a maximum brightness of 5,000 nits. To this we must add that it is compatible with the Dolby Vision image format. On the other hand, the Magic7 Pro honor is one of the mobiles who already have a Small battery, since in this case we talk about 5,850 mAh. In addition, this battery supports both fast charging of 100W and a wireless load of 80W. But If there is something that can highlight a lot in the honor Magic7 Pro is its cameras section. In this case, on the front we have a 50 MP camera, while we find a module that consists of a 50 MP sinor (variable opening f/1.4 YF/2.0), another 200 200), another 200 200 MP (F/2.6, X3 optical zoom) and a sensor more than 50 MP (f/2.0 ultra wide angle). Other mobiles that may interest you Samsung Galaxy S25 256GB Mobile phone with AI, Galaxy AI, 12GB RAM, 50MP camera, Dual SIM, IP68, Manufacturer Guarantee 3 years + 1 extra year, gray color (Spanish version) * Some price may have changed from the last review Xiaomi 14t Pro – 12+512GB smartphone, Leica Summilux Lenses, MediaTak Dimensity 9300+, Amoled screen 6.67 ” 144Hz, Hypercharge 120W, Black Titanium (ES version) * Some price may have changed from the last review Some of the links of this article are affiliated and can report a benefit to Xataka. In case of non -availability, offers may vary. Images | Honor In Xataka | The best mobiles, we have tried them and here are their analysis In Xataka | The best price quality price. Your analysis and videos are here

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.