Mobile Tech & Gadgets

China no longer plays in its favor

by

Apple has closed its fiscal fourth quarter of 2025 with $102.5 billion in revenue, surpassing the psychological barrier of $100 billion in a quarter for the first time. Earnings per share have reached $1.85, 13% more than a year ago. Wall Street expected less, so the stock is up 4% outside market hours It is the best quarter in Apple’s history. It is also the one that best exposes its dependence on China. Why is it important. Apple is already worth more than $4 trillion, the third company to reach that valuation after NVIDIA and Microsoft. Its results affect hundreds of suppliers in its production chain. But the growth of the iPhone, which still accounts for half of its revenue, has slowed. China is both a threat and an opportunity: If you regain traction there, the rally continue. If not, services will have to compensate more and more. And they are not infinite. Yes, but. ‘Greater China’ (a region that includes mainland China, Macau, Taiwan and Hong Kong) is the only region that has fallen compared to the previous year. Revenues in that market have been $14.5 billion, 4% less year-on-year and well below the $16.4 billion expected by analysts. Tim Cook has tried to soften the blow by promising that they will grow again in the first fiscal quarter thanks to the iPhone 17but the numbers sing: Apple is losing ground where it hurts most. Besides, Chinese brands are winning the battle of prestige on their own territory. Manufacturers like Huawei, Xiaomi or Vivo are no longer cheap alternatives and have started to position themselves as premium options, with special emphasis on the former. Apple is no longer the only status symbol in a market that manufactures many of its products. The money trail. The Services division has reached $28.75 billion this quarter, 15% more than last year. It is a historical maximum and the figure that really sustains Apple’s growth. In the full fiscal year, Services have exceeded $109 billion, another record. iPhone: 49 billion (+6%). Services: 28,750 million (+15%). Mac: 8,726 million (+13%). iPad: 6,952 million (practically flat). Home, wearables and accessories: 9,013 million (-0.3%). In this last division are Apple Watch, AirPods, HomePod, Apple TV… Services already represent 28% of total revenues but their very high margin compared to hardware means that they generate close to 50% of operating profit. Services, after all, do not require complex supply chains or rely on product cycles. In detail. The tariffs have cost $1.1 billion in the quarter and are expected to reach $1.4 billion in the next. Kevan Parekh, the chief financial officer who has replaced Luca Maestri, has projected revenue growth of 10% to 12% for the December quarter — the first of Apple’s fiscal year — with iPhone sales growing by double digits. Analysts expected only 6%. Cook has highlighted the “very strong demand” for the iPhone 17, launched in September alongside the iPhone Air. They have also mentioned supply constraints, suggesting that they could have sold more if they had been able to make more. The backdrop. Apple depends on China in two directions: As a consumer market. And as a production center. This double dependence is a geopolitical vulnerability that has become more evident with the trade war. The company has tried to diversify its manufacturing towards India and Vietnambut China remains irreplaceable in the short term. Meanwhile, in China, Apple is no longer perceived as the only aspirational brand. Local manufacturers have improved a lot in design, cameras and software, which leads to an improvement in perceived value. And they’ve done it while Apple navigated years of incremental iPhone updates. Featured image | apple, Li Yang In Xataka | Ode to rounded corners, the visual element that has proven Steve Jobs right once again

We already know how to retrieve the exact prompts that people use in AI models. It’s terrifying news

by

A group of researchers has published a study that once again raises alarm bells regarding privacy when using AI. What they have managed to demonstrate is that it is possible to know the exact prompt that a user used when asking a chatbot something, and that puts AI companies in a delicate position. They can, more than ever, know everything about us. A terrifying study. If you are told that ‘Linguistic models are injective and, therefore, invertible’ you will probably be shocked. That’s the title from the study carried out by European researchers in which they explain that large language models (LLM) have a major privacy problem. And it has it because the transformer architecture is designed that way: each different prompt corresponds to a different “embedding” in the latent space of the model. A sneaky algorithm. During the development of their theory, the researchers created an algorithm called SIPIT (Sequential Inverse Prompt via ITerative updates). Such an algorithm reconstructs the exact input text from the hidden activations/states with a guarantee that it will do so in linear time. Or what is the same: you can make the model “snap” easily and quickly. What does this mean. What all this means is that the answer you got when using that AI model allows you to find out exactly what you asked it. In reality, it is not the answer that gives away, but the hidden states or embeddings that the AI ​​models use to end up giving the final answer. That’s a problem, because AI companies keep these states hidden, which would theoretically allow them to know the input prompt with absolute accuracy. But many companies already saved the prompts. That’s true, but that “injectivity” creates an additional privacy risk. Many embeddings or internal states are stored for caching, for monitoring or diagnosis, and for customization. If a company only deletes the plain text conversation but does not delete the embeddings file, the prompt is still recoverable from that file. The study shows that any system that stores hidden states is effectively handling the input text itself. Legal impact. There is also a dangerous legal component here. Until now, regulators and companies argued that internal states were not considered “recoverable personal data,” but that invertibility changes the rules of the game. If an AI company tells you that “don’t worry, I don’t save the prompts” but it does save the hidden states, it’s as if that theoretical privacy guarantee is of no use. Possible data leaks. A priori it does not seem easy for a potential attacker to do something like this because they would first have to have access to those embeddings. A security breach that results in the leak of a database of those internal/hidden states (embeddings) would no longer be considered an exposure of “abstract” or “encrypted” data, but rather a plain text source from which, for example, financial data or passwords that a company or user has used when asking the AI ​​model could be obtained. Right to be forgotten. This injectivity of LLM also complicates the requirements of regulatory compliance for the protection of personal data, such as the GDPR or the “right to be forgotten.” If a user requests complete deletion of their data from a company like OpenAI, they must ensure that they delete not only visible chat logs, but also all internal representations (embeddings). If any hidden state persists in any register or cache, the original prompt would still be potentially recoverable. Image | Levart Photographer In Xataka | OpenAI is making the tech industry unite its destiny with yours. For the sake of the global economy, it better work

what is this subscription, features and differences with ChatGPT Plus

by

Let’s tell you what are the features of ChatGPT Gothe new subscription of ChatGPT available in Spain. It is a cheaper alternative that gives you access to some advanced functions, but with which you will not be able to use some other extra functions such as creating videos. It is a new subscription that is now beginning to arrive in our country. Still, it is a subscription to consider if you do not want to pay as much for greater access to GPT-5 or advanced imaging, since It costs 10 euros a month. We are going to tell you its characteristics and the main differences with other subscriptions. What is ChatGPT Go ChatGPT Go is the cheapest subscription plan to access advanced features of the artificial intelligence from OpenAI. This is an intermediate plan between the little that the free version offers with its numerous limits, and what the Plus plan offers, which is the “normal” AI plan. Being in this middle place, ChatGPT Go offers expanded access to GPT-5 and any new versions of the artificial intelligence model that come out in the future. It also widens file upload margins and speeds up image creation. But it is missing some things that the Plus version does have, as we will tell you later. So, it becomes an option for those who prefer to pay less and have access to fewer functions that ChatGPT offers. It has a price of 9.99 euros per monthless than half of the 23 euros per month that the Plus version costs, being a more popular price for common users. Features of ChatGPT Go vs. Plus Feature Free ChatGPt chatgpt go ChatGPt plus chatgpt pro Price €0/month €9.99/month €23/month €229/month access to gpt-5 Access to GPT-5 Expanded access for GPT-5 GPT-5 with advanced reasoning GPT-5 with professional reasoning Messages and uploads Limited file uploads Expanded messages and uploads Expanded messages and uploads Unlimited messages and uploads Images Limited and slow imaging Expanded and faster imaging Expanded and faster imaging Unlimited and faster image generation Memory Limited memory and context Expanded memory and context Expanded memory and context Maximum memory and context Deep investigation Limited depth research Limited depth research Expanded in-depth investigation, and agent mode Maximum in-depth investigation and agent mode Projects, tasks, gpts No access Custom Projects, Tasks, and GPT Custom Projects, Tasks, and GPT Expanded Custom Projects, Tasks, and GPT sora No access No access Video generation with Sora Video generation with extended Sora codex agent No access No access With Codex agent With expanded Codex agent Video preview No access No access No access New Feature Research Preview As you can see, the ChatGPT Go version has expanded access to GPT-5, but without advanced reasoning of the plus version. Its in-depth research is limited as in the free version, although hopefully not as much, and it gives you access to the GPTs personalized, and the creation of projects and tasks. ChatGPT Go it doesn’t give you access to Sora and its video creation, nor to the Codex programming agent. It shares expanded memory and context, better image generation, and more file uploads with the Plus version, although OpenAI does not specify the differences that possibly there is. The Plus version is still the standard plan that gives you access to the main features of ChatGPT, from video creation to greater capabilities and advanced reasoning in GPT-5. There is a higher subscription, ChatGPT Pro, but its price is still so high that it is not worth it unless you use AI professionally and always need the highest limits. But hey, at least the Go version gives you a more reasonably priced gateway to features that go beyond the limitations of the free version. This is good news for those who want to use ChatGPT on a daily basis and take advantage of various functions, but without needing it so much as to pay 23 euros per month. In Xataka Basics | Get the most out of GPT-5: 11 features and tricks to get the most out of ChatGPT’s new features

Character.AI has just shown ChatGPT the way to tackle the problem of AI and minors

by

The suicide of a teenager and other cases of delusions have put AI chatbots in the spotlight for their effects on mental healthespecially that of the youngest. ChatGPT has already implemented parental controls in response to this growing concern and a few days ago they admitted to being aware that ChatGPT was causing serious mental health problems for some users. . Now, Character.ai, one of the chatbots that is also at the center of the controversyhas made a drastic decision. +18. United States already is legislating to regulate the use of so-called ‘AI companions’ in minors and Character.ai has gone ahead by taking a more drastic measure than parental control. As they say in TechCrunchStarting on November 25, the app will begin limiting the usage time to two hours for those under 18. Little by little, that time will be reduced until it is zero. Age verification. To ensure that those under 18 can no longer use the app, Character.ai is going to deploy an age verification system that analyzes user behavior. If it fails, the app will use additional identity verification and facial recognition to block users who are not of legal age. ChatGPT, warm up as you go out. OpenAI has recently taken steps to prevent cases like that of Adam Rainemainly the integration of parental control in ChatGPT. The measure contrasts with Altman’s statements days later, when he said that ChatGPT would allow us to have conversations with erotic contentbut only for adults. Altman said that the app would require age verification for anyone who wanted to use it, but he did not say anything about blocking it completely to minors as Character.ai has done. Shot in the foot? Making ChatGPT an app for people over 18 years of age would be a strong measure to end the problem or, at least, comply with what the regulations ask for. first laws that want to regulate ‘AI companions’. However, it could be shooting themselves in the foot because it would cause them to lose a lot of users, something that cannot be afforded in the current climate. OpenAI recently published a in-depth report on your usersbut did not share the data of users under 18 years of age, according to them for privacy reasons. We know that almost half (46%) of their total user base are between 18 and 26 years old. That is, a large part of them are very young users. Minors and AI. OpenAI does not want to tell us how many minors use its chatbot, but thanks to other studies we know that the use of generative AI is very popular among this age segment. According to a International Plan studyin Spain 86% of young people between 12 and 21 years old have recently used AI. In turn, 18% of girls and 12% of boys admit to using it “to talk and tell their things.” Image | Character.ai, Pexels In Xataka | The great paradox of cell phones in adolescence: they are bad for performance and mental health, but banning them at school does not help

Discounts on Xiaomi and Apple mobile phones, an Android projector on sale and more in the last Hunting Bargains in October

by

We return with a new Hunting Bargains loaded with offers on various types of products. Although this week we have seen many sales on mobile phones, many other devices have also received good discounts to end the month, and as a good preview of what is to come on the next Black Friday. In this article we are going to review what they are five of the best deals we’ve seen this week. Wanbo Mozart 1 Pro by 349 eurosan Android projector that comes with a powerful speaker system. Xiaomi 15T by 489 eurosa good price for one of the latest phones launched by Xiaomi. Samsung U8075F by 279 eurosa very reasonable price for this 43-inch TV. Honor 400 Pro by 494.10 euros with coupon, a great price for this mobile phone that comes with headphones and charger. iPhone Air by 1,129 eurosone of the best prices that Apple’s mobile phone has had to date. Wanbo Mozart 1 Pro PcComponentes is celebrating its Halloween campaign and has lowered the price of numerous devices. Among them, we have the Wanbo Mozart 1 Pro by 349 eurosa home projector with an elegant design whose most interesting point is that it comes with an operating system Androidwhich allows you to download apps from streaming platforms. It can be configured to have a projection of 55 to 120 inches, offers Full HD resolution and has a powerful speaker system that offers 16W audio power. The price could vary. We earn commission from these links Xiaomi 15T That Xiaomi lowers the price of its mobile phones when launching them on the market is already common, and fortunately weeks later they can be found even cheaper in supplier stores. Powerplanet, for example, has the Xiaomi 15T by 489 euros in its 256 GB configuration. It is a good mobile phone whose screen offers a 1.5K resolution and compatibility with Dolby Vision. Besides, Their cameras are signed by Leica. The price could vary. We earn commission from these links Samsung U8075F If you are looking for a good television and you don’t want to spend a lot of money, be careful because Amazon has the smart TV Samsung U8075F for only 279 euros. It is a television from its entry range that comes in this case with a 43-inch screen. It is compatible with HDR10+incorporates the Filmmaker mode to watch movies and series and works with both Alexa and Google Assistant. Samsung U8075F (43 inches) The price could vary. We earn commission from these links Honor 400 Pro Honor usually has very good offers in its official store; Most of the time, there are discounts, coupons and even packs – even a combo of all of this. It is the case of Honor 400 Pro which right now is found 494.10 euros and comes with Honor Headphones Pro headphones and a 100W Honor SuperCharge charger. He Honor 400 Pro It is a mid-range mobile that comes with 512 GB internal storageit has 100W fast charging and wireless charging and its main sensor is 200 MP. To access this offer, you must select the coupon that appears in the payment panel. If it does not appear, you can write directly by adding “ADTO10“. The price could vary. We earn commission from these links iPhone Air Although we have not yet found offers on the iPhone 17 – at least in the main supplier stores -, we cannot say the same for the iPhone Air. Apple’s mobile phone has received its second (and best) discount at Powerplanet, whose price has fallen to 1,129 euros. He iPhone Air It is the thinnest mobile phone that Apple has launched to date (0.56 cm), but that is not the only thing that stands out. Incorporates the A19 Pro chip (the same one that has the iPhone 17 Pro Max), comes with 256 GB of internal storage and has a good battery that offers, according to Apple, a autonomy of up to 27 hours video playback. 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 | Wanbo, Xiaomi, Samsung, Honor, Apple In Xataka | The best mobile phones (2025), we have tested them and here are their analyzes In Xataka | Best televisions in quality price. Which one to buy and seven recommended 4K smart TVs

features, price and technical sheet

by

7,246 units sold in 2025 at the end of September. That is the data that raises the Omoda 5 as the best-selling car from Chery (Omoda, Jaecoo and Ebro) in our country. The compact SUV is one of the great cars of the moment, betting on the design-price-equipment combo as its great strength. The data is so good that, in fact, the SUV has the Top 10 best-selling gasoline cars in Spain within reach. Month by month, they have already sneaked in on occasion. They told us from Omoda It has only been a few days since they hope that this is just the beginning and that the arrival of a hybrid version should give them a good boost to sales. Nowadays the customer demands ECO sticker and the Omoda 5 SHS-H has it at a groundbreaking price, as is usual for the brand. Technical sheet Omoda 5 BODY TYPE. five-seater SUV MEASUREMENTS AND WEIGHT. 4,373 meters long, 1,824 meters wide, 1,588 meters high. Wheelbase of 2,610 meters and weight of 1,480 kg. TRUNK. 370 liters MAXIMUM POWER. 224 HP WLTP CONSUMPTION. To approve ENVIRONMENTAL DISTINCTIVE. C, ECO and Zero Emissions DRIVING AIDS (ADAS). 13 ADAS aids, including those required by the European Union and adaptive cruise control with lane centering. 360º HD camera for parking assistance OTHERS. Own software compatible with Android Auto and Apple CarPlay, via Bluetooth. Wireless charging for mobile phone. USB type A and C ports for the front and rear seats. Heated and ventilated seats as standard. ELECTRIC HYBRID. Yes, with a 105 kW (143 HP) 1.5 TGDI engine and a 150 kW (204 HP) electric motor. Consumption to be approved. Plug-in HYBRID. No. electric Yes. With a 150 kW (204 HP) engine and a 61.05 kW battery to approve 430 km. price and release Now available: Gasoline version from 27,900 euros. Hybrid version from 28,795 euros. Electrical from 37,900 euros before aid from the MOVES III Plan. Winning formula, now with ECO sticker There are no major changes in the Omoda 5 hybrid in relation to its gasoline versions but, in reality, it is not that it is needed. For now, the commitment of this Chinese brand seems to be working with an approach that focuses on design and a different image but, above all, on a cabin that doubles down. Now, with the Omoda 5 SHS-Hthe hybrid version of the model, the company adds a combination in which we find a 105 kW (143 HP) 1.5 TGDI combustion engine and a 150 kW (204 HP) electric motor that can together deliver up to 165 kW (224 HP) with a maximum torque of 294 Nm. Its acceleration remains in 7.9 seconds to cover 0 to 100 km/h. The company highlights that the battery for this hybrid is a 1.83 kWh LFP with “one of the highest instantaneous discharge powers in the sector.” They point out that in ideal conditions (25ºC and more than 50% battery charge) it can deliver up to 85 kW (114 HP) for 10 seconds, improving acceleration and recoveries. At Omoda they talk about “one car, two worlds” so, in the absence of testing it and seeing the technical specifications of the vehicle, we should expect great prominence from its electric side. It will have, to vary the power delivery and play with both engines, an ECO and a Sport driving mode. The accompanying equipment will continue to be really extensive. Among the highlighted elements, the double 12.3-inch screen is maintained, compatibility with Apple CarPlay and Android Auto or the dual-zone air conditioning. The customer will have a multifunction leather steering wheel, adaptive cruise control with emergency braking, lane maintenance and a rear camera as standard. Regarding its prices, the Omoda 5 SHS-H will be sold from 23,990 euros: Photo | Omoda In Xataka | “The hybrid is what people ask for”: Shawn Xu, global CEO of Omoda and Jaecoo, tells us about his plans for Spain and Europe

move 400,000 people a day without a driver

by

Line 6 of the Madrid metro is about to become the first line in the network to operate with fully automatic trains. There will be no driver, but more frequency, more capacity and less electricity consumption are expected. The project, valued at 300 million euros, will end in 2027 and has the objective of transforming the most used Madrid suburban line into one of the most advanced in Europe. A technical achievement. The Circular transports daily about 400,000 peoplemaking it the busiest line on the entire network. Automating it entails both a technical and logistical challenge, since they must maintain the service while the electrical infrastructure is renovated and security systems, such as platform doors, are installed. Madrid Metro describe as “a new technical achievement in the history of the suburban”, and no wonder. It will be the largest renovation since 2008 and will affect more than 11% of the current fleet. In detail. Automation requires profound changes to the infrastructure. The most important one is increasing catenary tensionfrom 600 to 1,500 volts, a jump necessary to power the new trains and reduce electrical consumption by up to 30% on L6. Across the entire network, the savings will be 3.6%, according to inform the organism. The works include the comprehensive renovation of seven electrical substations and the installation of automatic doors on the platforms. The Community of Madrid has invested more than 450 million euros in the purchase of 40 new convoys manufactured by CAF, designed to coincide with the openings of the current carriages and facilitate the transition. How it affects the user. The new trains promise a passing frequency of two minutes33% more speed and 17% more capacity by eliminating the driver’s cabins. All this translates into less waiting time and more space for travelers. But the process has not been easy. The works, which began in May 2024 and will last until the end of the year, have forced complete sections of the line to close. The buses replaced the service, although it was insufficient given the demand, forming collapses in stations like Moncloa. A first step. Madrid is betting on automation not only to improve its service, but as its particular strategy to aspire to sustainability. Reducing electricity consumption and increasing energy efficiency fit with European decarbonisation targets of public transportation. Furthermore, the project places Madrid in the select group of cities with fully automatic metro lines. We can see examples of these in Copenhagen or Lille, among others. Now it remains to be seen if the L6 project will be replicated on the rest of the lines. And now what. The works will continue until the end of the yearand it is expected that starting January 1 all stations will be open to the public during their usual hours. However, full automation of the network will not come until 2027, when the new trains enter service. Until then, the Madrid metro will continue to be a testing ground. The good thing is that if it works, the L6 will be the perfect showcase for what awaits the future of public transportation in the city. Cover image | Madrid Metro In Xataka | In 2018 it was a countryside on the outskirts of Chongqing. In 2025 it will be the largest train station in the world

The United States is offering millions of dollars to quantum companies. In exchange, he wants to keep a piece of each

by

The United States has opened a new stage in its industrial policy. This time it is not about aid without return or simple soft loans: Washington is offering millions of dollars to quantum companies in exchange for a share in its capital. The information comes from the Wall Street Journalwhich points out that the agreements seek more than just supporting promising companies. The message is clear: the Government wants to ensure a seat at the table for a technology that can reconfigure the economy and global power for decades to come. The initiative fits into a chain of recent decisions in which Washington has been deepening its presence in sectors considered strategic. The Government transformed almost 9,000 million dollars in previous aid to Intel in a participation close to 9.9% and obtained special rights in US Steel to oversee sensitive corporate decisions. He also supported MP Materials in the critical mineral chain. The signal is clear: when the sector is considered vital, Donald Trump’s White House seeks to stay on board. When public money also buys influence Conversations affect some of the most visible names of the American quantum ecosystem. According to the newspaper, companies such as IonQ, Rigetti Computing and D-Wave Quantum They are negotiating with the Department of Commerce the entry of the State into their capital. Other firms, including Quantum Computing Inc. and Atom Computing, are studying similar deals. Operations would start from a minimum of 10 million dollars per company in this initial phase, with the possibility of more applicants joining as the program progresses. The conditions are not limited to a mere public investment. The Commerce Department is studying formulas ranging from equity stakes to intellectual property licenses, royalties or revenue sharing schemes. The conversations are led by Paul Dabbarformer executive of the quantum sector and current number two in the department, according to published information. At this stage there are no closed agreements, but the approach indicates that the State seeks a tangible return and supervision tools. Washington’s interest is not explained only by financial reasons. Quantum computing is emerging as one of the technologies with the greatest capacity for industrial transformation. These machines promise to solve calculations that would take eons to current systemswith potential applications in fields such as drug design, advanced materials or highly complex chemistry. Adding to this momentum is international competition, with companies like IBM, Microsoft and Google involved and China advancing its own quantum race. The security dimension adds another layer of urgency. Quantum algorithms are projected to They may violate traditional encryption systemsincluding RSA and ECC, exposing both sensitive communications and critical infrastructure. The risk is not limited to the future: the strategy known as harvest now, decrypt later suggests that malicious actors are already collecting encrypted data for decryption when this capability becomes available. Given this scenario, Fortinet highlights the need to move towards post-quantum cryptography and strengthen networks and systems. The practical potential of this technology is well illustrated by the pharmaceutical sector. McKinsey highlights that quantum can transform drug development by enabling precise molecular simulations, something that classical calculus and pure AI fail to always capture. Large companies are already testing these systems to study proteins, evaluate chemical reactions or reduce experimental steps. This ability to model complex structures from scratch promises to accelerate research, improve the success rate in trials and shorten times to market for new therapies. The implementation of this approach is not limited to companies. According to the Wall Street Journal, the Commerce Department reorganized the office responsible for the scientific side of the CHIPS program and recovered several billion dollars that had been allocated to previous technology initiatives. The political message is transparent: the Executive wants public investments to be measurable and for the State to have mechanisms to benefit when the funded projects mature, especially in sectors with high strategic involvement. The shift raises dilemmas typical of a more interventionist model. Public participation can facilitate stability in strategic sectors, but it also opens the door to conflicts between technological, industrial or political priorities. The central doubt is to what extent the presence of the State will affect the pace of decision and the flexibility that the most competitive sectors demand. There are still relevant unknowns. The final percentages that the State could reach or the exact conditions that would accompany the participations are not known. According to the information available, the agreements are still in the negotiation phase and could be modified before being closed. It also remains to be seen what commitments will be required of companies and whether there will be associated performance or governance criteria. At this point, the process is moving forward, but a definitive schedule for awards or formalization of agreements has not yet been announced. Images | Dynamic Wang | D-Wave Quantum | Xataka with Gemini 2.5 In Xataka | The United States and China have finally met to resolve the trade war: one will give in on tariffs, the other on rare earths

If the question is how spacious the Starship will be, the answer is yes

by

Last week, NASA’s acting administrator proposed study alternatives to SpaceX’s Starship to send astronauts to the Moon before China does. SpaceX has just published a blunt response. A paradigm shift. The self-imposed moon race against China has made the United States forget the real reason why NASA chose the SpaceX’s gigantic Starship for his return to the Moon. As SpaceX itself has been responsible for remembering in a long publication Loaded with images, technical details and advances that we were unaware of, its Starship HLS (Human Landing System) is not a lunar landing module like that of the Apollo missions: it is a paradigm shift designed to build a permanent lunar base. Size comparison between Starship HLS and the Apollo lunar module This is Starship HLS. The comparison is almost comical. While the Apollo lunar module, that took the first humans to the Moonmeasured seven meters high, Starship HLS will rise vertically to 52 meters. To put it in terms of room to stretch your legs: The Apollo lunar module had the habitable volume of a wardrobe (4.5 cubic meters). The Lanyue spacecraft that the Chinese astronauts will use has twice the volume. Starship, according to SpaceX itself, will have two-thirds of the pressurized volume of the entire International Space Station (613 cubic meters). What’s more, the SpaceX ship will have two airlocks for exits to the surface. Each of them will have a habitable volume of 13 cubic meters, which means that a single Starship airlock is more spacious than the Chinese lunar landing module that NASA is so concerned about. Render of the Starship HLS cone inside A luxury apartment. If the size comparison wasn’t enough, new renders of Starship’s interior show a level of comfort that no spaceship has ever had outside of sci-fi movies. Forget the image of astronauts crammed into an aluminum can. What we see is a spacious, multi-story interior, with a clean and futuristic aesthetic. There is a spiral staircase, a control area with multiple seats and a bay window offering panoramic views of the lunar surface. Astronauts inside Starship HLS A beast of burden. Starship is not designed to carry new American flags to the Moon. As SpaceX has taken care to remember, it is designed to fulfill the initial promise of the NASA’s Artemis program: create a “permanent and sustainable presence on the Moon”, building a lunar base. Starship cargo variants will be able to land up to 100 tons directly on the lunar surface. This includes pressurized and non-pressurized rovers, nuclear reactors for power generation like the one NASA wants to install before China, and prefabricated lunar habitats. 2026 will be the moment of truth. SpaceX says it has completed 49 key milestones in Starship’s development, including demonstrations of life support systems, testing of landing legs, qualification of the docking adapter, and demonstrations of the elevator and airlock. However, the big obstacle remains refueling ships in orbit to compensate for the evaporation of cryogenic fuel, something that SpaceX hopes to achieve in 2026 with the new Starship V3. Without fuel transfer in orbit, Starship cannot reach the Moon with its crew and its 100 tons of cargo. Images | SpaceX In Xataka | The enormous size of Starship, in images that give an idea of ​​its scale In Xataka | A genius named Tom Mueller designed the engines for the Falcon 9. And now that genius wants to beat SpaceX on its own turf

YouTube is ready to launch the largest video reconstruction project in history. And yes, it will use AI

by

YouTube was born in 2005 and, since then, it has become the largest audiovisual archive in recent history. For years, millions of users uploaded videos in 240p or 480p because that was what the cameras, connections and devices of the moment allowed. This material does not lose value because it has low resolution: there are extraordinary pieces that continue to be a reference. But today the screens are better, the sound matters more and that difference is noticeable. So now comes an attempt to update that experience without erasing the past. An ocean of videos. The YouTube catalog is not large: it is huge. The figures published by electroiq put the total at around 4.3 billion videos in 2025, after a stage in which the Shorts format It pushed the increases to levels never seen before. About 800 million were added in 2023 alone. That momentum has tempered, in part because of controls on repetitive content and a lower craze for short clips, but the trend remains. If the current pace continues, the service could surpass 10 billion videos before 2030. YouTube begins to “reconstruct” its videos. YouTube has announced that will begin to automatically improve videos uploaded in resolutions between 240p and 720p, raising them to HD quality using artificial intelligence. The process does not delete the original files or modify the base video: it is an alternative version visible under the “super resolution” label. Creators will be able to decide if they want it to be applied and viewers will retain the option to view the content in its original resolution. It is a measure that seeks to modernize the archive without altering its authenticity. Lens: 4K. The roadmap is clear. After starting with videos below 1080p, YouTube wants automatic enhancement also reach 4K resolutions “in the short term,” according to its announcement. To support that jump, heavier uploads are already being tested with some creators and thumbnails will also be able to reach 4K, thanks to the extension of the file limit from 2 MB to 50 MB. Everything points to an attempt by the platform to ensure that both the content and its presentation are at the level of current panels. The audio also goes up a notch. The modernization of the catalog does not stop at the visual. YouTube has also introduced automatic audio enhancements that adjust the mix and maintain a more consistent volume between different scenes. These automatic improvements are grouped under the “Stable volume“, which the viewer can activate or deactivate according to their preference. With this, the company seeks to prevent sound jumps from breaking the experience, something common in old videos or recordings with basic equipment. It will not be for all videos. YouTube clarifies that these improvements will not be applied indiscriminately. They will only affect videos uploaded in low resolution that have not been previously remastered to 1080p or higher. Additionally, as we already mentioned, creators can decide from YouTube Studio whether they want the platform to apply visual or audio enhancements to their future uploads. It is a measure that seeks to avoid unwanted distortions and give room to those who prefer to keep their content exactly as it was published. If these enhancements are disabled for the channel, viewers may not be able to use features such as “Stable volume” or “Super resolution” on that content. Upscaling no longer lives on your TV. Many televisions include their own systems to improve the image, but YouTube’s approach is different. Instead of applying upscaling on the device, the platform does it in the cloud, allowing it to process millions of videos consistently without depending on the user’s hardware. Additionally, the viewer can choose between the original playback or the enhanced version from the quality menu, with an option visible and explicit in the interface. Catalog, discovery, legacy. For creators, this update has an immediate benefit: they don’t need to re-upload old videos to make them look better on current screens. The file remains intact and the enhancement is applied as an additional layer, respecting the original. This can help valuable pieces from years ago gain presence, without altering their essence. The viewer, for their part, receives a more homogeneous experience and the possibility of choosing how to view each content. Images | Xataka with Gemini 2.5 | CardMapr In Xataka | NVIDIA has risen to the top for its AI data centers. Your next big leap: cars

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.