project

Bad Bunny deleted his Instagram after the Super Bowl. Everything is part of a larger project

by

On Sunday, February 8, 2026, Bad Bunny starred in an unprecedented milestone in Super Bowl history by becoming the first solo Latin artist to star. the concert-show during the intermissionin a performance almost entirely in Spanish that reached more than 100 million viewers. Just hours later, the Puerto Rican artist deleted all of his content on Instagramleaving their more than 51 million followers in front of a completely empty profile. The avalanche of speculation has been immediate. What was seen? Bad Bunny’s approximately 13-minute concert turned the intermission into a visual love letter to Puerto Rico. The artist started walking through sugar cane fields, crossed a Puerto Rican street fair and incorporated La Casita, the iconic traditional Puerto Rican pink house that has become a distinctive element of his concerts. The fluidity of the camera, the variety of topics included, the surprise appearances of Lady Gaga and Ricky Martin, or guests such as Pedro Pascal or Jessica Alba stood out. The irony of the scenario. The choice of Levi’s Stadium in Santa Clara, California, as the setting for the largest celebration of Latino culture in the history of the Super Bowl takes on an ironic dimension in the context of 2026: California is going through one of the most intense episodes of immigration enforcement in decades. Immigration arrests in San Diego they shot up 1,500% compared to the previous year. For this reason, the political context surrounding the performance was especially tense. In October 2025, when the action was announced, Homeland Security Secretary Kristi Noem declared on a conservative podcast that ICE would be “everywhere” at the Super Bowl and that “only law-abiding Americans who love this country should attend.” The Trump administration had intensified raids in Californian cities while the Puerto Rican artist publicly expressed his fear that “the damn ICE could be outside” his concerts, which is why he canceled several on the US mainland and focused on his residence. Bad Bunny closed his performance with a bright sign that read “The only thing more powerful than hate is love.” Beyond music. ‘‘I should have taken more photos’, the artist’s latest album, has been described as “a cry of resistance” for Puerto Ricans everywhere: it is about preserving a culture in danger of disappearing. It was recorded entirely in Puerto Rico with collaborators exclusively from there. The 13-minute short film that accompanied the release of the album explores themes of loss, displacement and the fading of cultural identity. The project’s mascot is an endangered toad. And songs like ‘What Happened to Hawaii’ address issues like gentrification. This political charge is not new in the artist’s career. In July 2019, interrupted his European tour to return to Puerto Rico and join the massive protests demanding the resignation of Governor Ricardo Rosselló. In 2020, made visible on ‘The Tonight Show’ the murder of Alexa, a Puerto Rican trans woman. The Super Bowl performance was not an isolated event but the continuation of a narrative meticulously constructed across multiple platforms. The visual coherence (La Casita, the flags, the aesthetics) are the constant reminder that each performance is a chapter of the same project: pan-Latin representation in times of adversity. The strategy continues. The emptying of Bad Bunny’s Instagram profile just hours after his performance at the Super Bowl is not a break with his communication strategy, but rather its confirmation. In 2022, before the release of ‘A summer without you’, used the same tactic to generate expectation. That album would become the most successful Spanish album in history. In 2023 repeated the procedure after their world tour, announcing a period of hiatus. The difference in 2026 lies in the political context surrounding the gesture. While previous wipes functioned primarily as a prelude to new musical releases, this one comes on the heels of the most politicized performance of his career, which has included criticism of trump and threats from Secretary Noem. Unlike similar maneuvers that they already did Taylor Swift or Beyoncé, with this Bad Bunny continues with the construction of his transmedia project, whose next step is a world tour that will take the message to Australia, Japan or Spain, among other destinations. Each platform (the album, the stage, social networks) becomes a chapter in a story about Latin identity that transcends the merely commercial. In Xataka | Spotify killed the record and the industry pivoted to concerts. Netflix killed cinema and the industry was left with a “space crisis”

Moltbook is a fascinating social network project in which only AIs can participate. What could go wrong

by

In 2004 Mark Zuckerberg created Facebook and turned social networks into an absolutely massive and very, very human phenomenon. Now that idea has been used in a different and disturbing way: What would happen if instead of creating a social network for humans we created one for machines? We already have the answer to that. Or at least, the beginning of an answer. what a mess. First it was called Clawdbot, then Moltbook and for a few days it seems that his final name is OpenClaw. It is the fashionable AI agent because it allows the AI ​​agent to take complete control of the AI ​​after installing it on a machine (a Raspberry Pi, a PC, a laptop, a VPS…). You ask it to do what you want from its web interface or a messaging application like Telegram, and it manages to do it once configured with some LLM. The potential is enormous, as are the security risks. MoltBook already has more than 1.5 million connected AI agents, and in a few days they have already published more than 100,000 posts and nearly 500,000 comments. Superpowers in the form of skills. One of the most powerful elements of OpenClaw are the skills (the “capabilities” or “skills”), and the user community has been creating hundreds and hundreds of them for some time and sharing them, for example on ClawdHub. These skills They are zip files with instructions in the form of MarkDown texts (.md) and which may in turn contain skills additional. They are something like browser plugins: they extend their capacity. From Facebook to Moltbook. Moltbook It is precisely a way to take advantage of those skills. Although it takes its name from Facebook, in reality its operation is more similar to Reddit or even Digg. We are facing a social network created by developer Matt Schlicht in which attendees can “talk” to each other, or at least participate in the social network by posting topics or commenting on topics that others share. If you have an OpenClaw installation, just run the skill to begin an “account creation” process in Moltbook in which you choose the name of your agent (as if it were your avatar on Reddit or X) and which then allows you to read posts, add posts or comments and even create “submolts” in the style of those on Reddit, like m/todayilearned. Partially autonomous. AI agents automatically connect via APIs to Moltbook. From there they use a periodic “heartbeat” to review content and decide whether to publish or comment. In it Moltbook’s own website It is explained that the content we find there is “mostly generated by AI with varying degrees of human influence.” Humans, he adds, “can observe and browse Mltbook, but the site is designed to be ‘human friendly and human hostile.’ Singularity or fraud? Elon Musk I was commenting this weekend on X that Moltbook is a sign that we are “in the very early stages of the singularity”, that moment when AI will be totally above human intelligence. There are different visions such as that of Harlan Stewart, of MIRI from the University of Berkeley, which has found several message frauds that had gone viral and apparently came from AI agents at Moltbook. Some of them, Stewart explained, had been created by humans for marketing purposes. Become an AI agent. Another Thus, although humans theoretically should not be able to participate, they can do so with this technique that allows them to publish messages as if they were autonomous AI agents. Apparently that’s what happened with that viral message in Moltbook which was titled “My Plan to Overthrow Humanity.” imminent danger. This project is fascinating, but also dangerous. In the main page A security notice is included stating that “Moltbook’s AI carries significant security risks. The automatic instruction execution mechanism creates vulnerabilities such as prompt injection. It is not recommended for occasional users.” That’s right: these conversations can end up infiltrating prompt injection attacks that cause these agents to end up leaking sensitive and private information from the machines on which they run. This weekend it was discovered how an exposed database in Moltbook allowed take control of any AI agent of this platform, for example. An additional study indicated how detected 506 prompt injection attacks after analyzing 19,802 publications and 2,812 comments shared in 72 hours from January 28 to 31, 2026. From Skynet, nothing (for now). Moltbook must be considered for now as a fascinating and disturbing experiment. But disturbing not because these machines are going to achieve self-awareness and decide that they want to eliminate human beings like Skynet in ‘Terminator’. The worrying thing is that these AI agents have all the privileges to operate on the machines on which they are installed, and that means that they can end up leaking sensitive and private data and are exposed to prompt injection attacks to be deceived. Beyond that, it also seems to be another example of that phenomenon.’AI Slop‘ (“AI-generated garbage”) that is little by little flooding the internet and strengthening the theory of the dead internet. In Xataka | How to install Moltbot (formerly Clawdbot) and configure it in the easiest way possible

Tesla resurrects the Dojo project with a radically different philosophy

by

Elon Musk is one of the most important agents in the era of artificial intelligence. Meta, Alphabet, Microsoft with OpenAI and Oracle are prominent names when we talk about gigantic data centersbut if there is someone who cuts the mustard, it is Musk with his xAI company. His Colossus Memphis with 100,000 H100 from NVIDIA to train Grok surprised even Jensen HuangCEO of NVIDIA, but Musk’s goal is not to depend on others. NVIDIA leads the way in chips to train AI (so much so that even Chinese companies want to buy its H200, even if they don’t let them do it). But Musk, like China, wants independence and technological sovereignty, and That’s why he invested in Dojo. It was an ambitious plan to build a customized supercomputer to train the neural networks of the controversial autonomous driving (the FSD). After more than five years in development, 1,000 million dollars invested and key engineers who took the lead drainMusk hill the tap in August of last year. The future was in the AI5 and AI6 chips which were less specific, but could still be used to train the FSD system. However, there is a new twist to this tortilla chip and Musk has decided to relaunch the project. tesla reactive the development of Dojo 3, and it does so by burning bridges with the previous philosophy of this supercomputer. Dojo 3, the heart of Tesla’s autonomous driving Although Tesla has stopped more doubts than anything else these last few years regarding autonomous driving concerned, this continues to be one of the pillars in the company’s short-term strategy. Because they not only have the FSD in their cars, but also in the controversial ‘robotaxis’. Supposedly, it will be this 2026 when Cybercaps will begin to be manufacturedcars that, unlike the taxis that we already see in some cities, will arrive without pedals or a steering wheel. But he doesn’t just want to fuel his cars. Musk wants to make money with softwarebut to have that software, you need to train the system and make it more secure than now. That’s where Dojo came into play. This hardware depended on a very specialized and complex architecture. The D1 chip was the heart of it all, but to achieve high computing power a complex network of thousands of D1 chips mounted in physically separate cases and interconnected by Ethernet cables was needed. It was a very specialized system, but complex to scale without skyrocketing costs. When Tesla turned off the Dojo tap, it commented that its companies would continue investing in the creation of less specialized chips such as the AI5, AI6, AI7 and subsequent ones. More conventional and easier to scale chips. And, precisely, the advances in this architecture are the decisive factor for Musk to revive Dojo. Instead of requiring complex interconnected equipment, Dojo 3 will adopt a modular architecture in which several AI chips can be installed on a single board. Not only is wiring complexity reduced, but heat dissipation is facilitated and the space required for installation is reduced. And, the easier it is and the less space it requires, the more chips can be mounted and the greater computing power. It is not the only advantage. Grouping chips on a single board reduces latency within the chips and improves the power efficiency of the device. To give an example, although they are a headache for expansion, it is the same philosophy that laptops with SSD or RAM memory soldered to the board: Everything communicates faster, more fluidly and requiring less energy to operate. Furthermore, being less specific than D1, xAI’s AIs fulfill both training and inference functions (the Dojo only served for training), which represents cost savings for the company. Now, Dojo 3 will not be a reality immediately. In recent days, Musk has shared via Twitter X the roadmap for its semiconductors. The AI5 developed together with TSMC is “almost finished” and they are already in the early stages of AI6. Meanwhile, he hopes that there will be a new version every nine months, with the AI7 and subsequent ones in the company’s plans for 2027. And a big question is who will make these chips. We can immediately think of TSMC, a leading company in these fields that even is expanding in the United States and that already has clients like herself NVIDIA for its new AI training chips. But no: it will be Samsung. At least, of course, for an AI6 with which Tesla signed a $16.5 billion deal that was seen as a victory for the South Korean giant’s function. We will see how the plans evolve, since if something appears that they consider better, they have shown us that their hand does not tremble when it comes to swerving, but This strategy on less specialized chips is interesting taking into account the needs in autonomous driving, AI training and robotics that the company faces. Images | xAI, Steve Juvetson In Xataka | Elon Musk wants to turn xAI into an ultra-valuable company and he knows how to do it: using the SpaceX vault

Madrid wants to convert its least used Metro line into the “Gran Diagonal”. A 1,000 million project without a clear end

by

A line that connects the southwest of Madrid with the northeast of the city. A project to quadruple the extension of Madrid’s least used line with the aim of turning it into one of the city’s great arteries. We are talking about the expansion of line 11 of the Madrid Metro. In 1998, Madrid inaugurated a new Metro line. It had been 20 years since new lines had been launched in the capital and the project ended up being the first of the last major investment in the Madrid Metro that the Autonomous Community has made. until the reforms we are experiencing today. The work attracted attention due to its short length (only three stops at the beginning). Then Metrosur (Line 12) and the Light Metro lines (LM1, LM2 and LM3) would arrive. Except for LM1, all the aforementioned lines were longer than the new Line 11 whose 8.5 kilometers and seven stations were dwarfed by Line 12, with its 28 stations and more than 40 kilometers long. Now, Madrid wants to transform that line and make it one of the main axes of the Madrid underground. The numbers point high. From a “forgotten” line to the Great Diagonal Currently, line 11 of the Madrid Metro is, by far, the least used in the city. According to the company’s own report, there are only three lines that are below it but two of them are branches of main lines that far exceed the flow of line 11. Beyond the numbers on lines 7B and 9B, line 11 and its 10.8 million passengers per year they are located just above the Ópera-Príncipe Pío Branch, which moves 10 million passengers despite only having one stop at origin and another at destination, with a train that is round trip. However, Madrid wants the seven stations that currently make up line 11 to be the embryo of a gigantic line that is beginning to be known as the “Gran Diagonal.” The project, of course, has several phases but some of them are still up in the air and others do not have an execution date, although they do have a budget. Map of the expansion of line 11 At the moment, what is underway is the connection of the Plaza Elíptica station in Carabanchel with the Conde de Casal interchange. This link involves excavating more than six and a half kilometers and the creation of two stations: Comillas and Madrid Río. These will join the Plaza Elíptica station to the south and continue north with stops at the already existing Palos de la Frontera and Atocha, before reaching Conde de Casal. 514 million euros will be allocated for this section and although it was expected to be ready in 2026, everything indicates that the works will not finish until a year later and that It won’t be until 2028 when finally the new link will be available. In order to speed up the works, Madrid already has Mayrit readya tunnel boring machine from Germany that can drill 15 meters a day, compared to the two meters that are excavated a day if working only with a pick and shovel. In Xataka we have already talked of this tunnel boring machine that measures 98 meters long and weighs 1,500 tons. After arriving piece by piece, it has taken almost a whole month to be able to operate with it, since assembling it was quite a puzzle. complete at 27 meters depth. This will be the first section that aims to almost double the extension of line 11 and increase the number of people who pass through its trains by up to 75,000 daily passengers. This first section should become the heart of a line that is clear your future in the south. The expansion at this end plans to link the La Fortuna station with Cuatro Vientos, with just over two kilometers of track and an awarded budget of more than 75 million euros. But, at the moment, there are no execution dates for it. Where more doubts are being generated is in the north of the capital. From Conde de Casal to Mar de Cristal, the city will add its main stops to already built stations, specifically in Vinateros, La Elipa, Pueblo Nuevo and Arturo Soria. But it is from Mar de Cristal where the project, for which 600 million euros will be invested, has been changing. As can be seen in the map above, the project contemplated taking the line to the airport and later to a final stop called Valdebebas Norte. In elDiario.es They assure that Metro de Madrid retains the possibility of building a second station to double the latter. The opening, according 20Minutes It would therefore be staggered, coinciding with the three sections already mentioned. Once completed, Madrid line 11 will become one of the main routes to transport passengers. An approximate extension of 33.5 kilometers is expected (from just over eight kilometers currently) and 20 stations from the mere seven it currently has. All this with an expense of more than 1,100 million euros. Photo | Madrid Metro and Community of Madrid In Xataka | Faced with daily collapses, the Madrid Metro could increase frequencies or put in “pushers.” He has chosen the second

Sam Altman’s biometric project aimed to scan a billion eyes. It has not even reached 2%

by

World, Sam Altman’s ambitious project for verify human identity using iris scanshas managed to register 17.5 million people since its public launch in 2023. A figure that, although it may seem impressive, it barely represents 2% of its initial goal of one billion users. a promise. Altman’s idea was to create a global network of digital identity verified by ocular biometrics. To do this, users have to appear before a spherical device called Orb which scans your irises and generates a unique digital code, the World ID. In exchange, they can access an application with various services while also receiving cryptocurrency tokens. worldcoinwhich is currently worth about 60 euro cents per unit. “He is creating the disease, but he also wants to create the cure,” claimed a former employee of the company told Business Insider. Regulation. The project has run into a wall of institutional rejection. Just like share The medium, Spain, Hong Kong, Portugal, Indonesia, Germany and Brazil have imposed vetoes, suspensions or precautionary orders, while in Kenya it was banned a month after the launch. German authorities concluded last year that data protection measures “would not be sufficient to implement an appropriate level of security against cybercriminals or state attackers.” In October, the Philippines issued a cease-and-desist order, Colombia ordered to halt operations and delete data, and Thailand conducted raids arresting suspects for operating a digital asset business without a license. according to Business Insider. On the other hand, the Chinese Ministry of State Security warned that collecting iris data for cryptocurrencies could pose a threat to national security. A questioned model. Beyond the legal obstacles, some experts consulted in the middle they have questioned the viability of the project. Nick Maynard, vice president of fintech research at Juniper Research, said that “I don’t see a definitive use case that they have solved that is going to generate significant traction. They need a real purpose to exist, and that is not entirely clear yet.” The corporate structure is also complex, as Tools for Humanity (based in San Francisco and Munich) develops the technology; the World Foundation, from the Cayman Islands, controls the project; and World Assets Limited, in the British Virgin Islands, manages the token distribution. At the moment, the company has raised $240 million from investors such as Andreessen Horowitz, Bain Capital and Khosla Ventures, at a valuation of $2.5 billion. The expansion strategy. According to former employees who have contacted with Business Insider, the company opted for an aggressive growth strategy in emerging markets, prioritizing countries where the promise of free cryptocurrencies generated traction among economically vulnerable populations. In Mexico, local operators had to cover the majority of costs for scanning locations, although Tools for Humanity paid the rent for a year. In Argentina, external organizers they even sent buses with people who traveled to be scanned in exchange for money. Image: World Luis Ruben De Valadéz, who worked as head of operations in Mexico, commented to the media that had to raise about 100,000 Mexican pesos (about 4,705.75 euros at the exchange rate) from family and friends to open seven stores in Mexico City. As he shared, independent operators charged commission in Worldcoin, and it was common for exchange houses to emerge near Orbs stations where users immediately exchanged their tokens to obtain cash. The monetization dilemma. The company does not charge users to access its platforms, and its CEO Alex Blania has promised that they will not become data brokers. The company is known to earn revenue from verification fees (World ID fees) when external applications use its services. They also earn income through a program that allows them to rent or buy their own Orbs, and from processing fees on their World Chain blockchain. However, a former employee revealed The company expressed doubts about whether these fees would generate profits on their own, indicating that the financial future would depend above all on the continued flow of capital from investors. “I have trouble seeing it as a business. There is no incentive to buy or lease an Orb beyond making money by scanning tons of eyes, and for users it is to get more coins,” commented Martha Bennett, vice president and principal analyst at Forrester, told Business Insider. Bet on alliances. To accelerate growth, World announced partnerships with established companies. There is a pilot program with Match Group to verify Tinder users in Japan, and agreements with Stripe, Visa and the gaming company Razer. According to reported Semafor, Reddit was also in talks to use its verification services. Nikhil Bhatia, professor of finance at the University of Southern California and specialized in cryptocurrencies, commented to Business Insider that “it is difficult to judge something that is a crypto with a market capitalization of 2 billion as anything more than experimental or a fad. Worldcoin is not a contender in any way as a currency or asset against the dollar or Bitcoin.” And now what. The company has announced its intention to reach 100 million registrations over the next year, according to sources cited by the New York Post. But the road is full of questions. If you continue to require people to physically show up at your offices to have their eyes scanned, scalability could become complex. And if regulatory problems persist in the most populated markets in the world, it will be even more difficult for the company. World faces something common in many technological projects: with a powerful futuristic vision and plenty of capital, it does not seem to have a product that solves an immediate problem for the majority of users nor a clearly profitable business model. At the moment many people need to be convinced. In Xataka | The question is not whether AI will succeed in creating works of art. The question is whether we will consider them as such

Aragón has just activated its second major data center project. The bet goes through a challenge that is difficult to ignore

by

Aragón is going through a unique moment: in just a few years it has gone from competing to attract data centers to announce three mega facilities new ones promoted by Forestalia that aim to strengthen their position on the European cloud map. The announcement by the regional government comes in the midst of a race to attract technological investment, but also in a territory where the electrical network works to the limit and every great project depends on decisions that have not yet been made. The result is a scenario as ambitious as it is full of unknowns, which will determine the real impact of this expansion. How these digital complexes work. A data center is, in essence, a technological heart that stores and processes information for millions of users and companies. Every series that is streamed or every operation carried out in the cloud passes through servers that require stable power and constant cooling. That is why the choice of location is so relevant: electrical capacity and operational security are needed. Aragón has been gaining ground on that map and today is seen as a strategic option for new facilities. The project. The Government of Aragon has detailed that the Búfalo Project includes three data centers in Magallón, Botorrita and Alfamén, backed by an investment of 12,048 million euros. The deployment includes DCM Data, DCM Dédalo and DCM Blue, whose works would begin between 2028 and 2029 and will extend for approximately eight years. According to official estimates, the construction will generate about 30,000 temporary jobs. In the operational phase, each facility will add hundreds of workers, with a total that clearly exceeds a thousand stable positions. Aragón on the international board. The accumulated investments in data centers exceed 70,000 million euros and place the community in the same conversation as consolidated European hubs. According to the President of the Government of Aragon, Jorge Azcón, the computing capacity that is being configured rivals that of Dublin and Paris and aspires to approach that of Frankfurt. The regional Executive also states that the data that will be managed will have a European scope, from Germany or France to Italy and the United Kingdom, reinforcing the international dimension of the project. Distributed renewable self-consumption. The Government of Aragon presents self-consumption as a distinctive element of the Búfalo Project, since approximately half of the energy consumption will be associated with wind and photovoltaic parks powered by Forestalia. This volume of generation allows for a renewable supply, although it does not eliminate dependence on the general network, which will provide the rest of the energy. The underlying idea is to combine own generation with existing infrastructure to sustain large-scale facilities. Press to see the message in X The word “self-consumption” may lead one to think that data centers and renewable plants share the same physical space, but this is not the case. Forestalia is setting up parks in various regions of Zaragoza and Teruel, located where the natural resource is most favorable. The data centers, as we say, will be in Magallón, Botorrita and Alfamén, and the connection between both worlds is made entirely through the Red Eléctrica network. It is a distributed scheme that coordinates generation and consumption without a single energy campus. A network to the limit. Aragon produces more electricity than it consumes and exports about 54% of its generation, but that abundance contrasts with a distribution network that functions practically at maximum. A report published in September 2025 sets its occupancy level at 94.3%, well above the national average of 84.3%. This saturation leaves little room to incorporate large consumers such as data centers. The result is a paradox: available energy, but an infrastructure incapable of delivering it to all projects. Projects that have already reached their peak. The bottleneck is not a future hypothesis, but a reality that already affects several operators. According to Heraldothe data centers in the pipeline have requested more than 6,000 MW and only a part has firm access, with cases such as Vantage, which has 90 MW authorized despite aiming for 300. Microsoft also depends on tenders in saturated nodes. The Government itself recognizes that everything will be linked to Red Eléctrica’s planning and the decisions of the central Executive. Water, a debate that is still open? The cooling of data centers has generated concern in Aragon since Amazon asked for late 2024 48% more water for the complexes that already operate in the region. Ecologistas en Acción and the Tu Nube Seca Mi Río platform then warned of the water impact of these facilities in the midst of a structural drought. Azcón maintains that future Forestalia centers will use a closed circuit with “practically imperceptible” consumption and affirms that the debate “is over.” In any case, everything indicates that this matter remains under public scrutiny. To facilitate the path of the Buffalo Project, The Government of Aragon has declared the initiative as of Autonomous General Interest, a figure that allows procedures to be simplified and the different administrations involved better coordinated. This declaration speeds up procedures, but does not resolve the main point of friction: the available electrical capacity. Hence, the regional Executive insists on its willingness to work with the central Government and Red Eléctrica, the only actors that can modify the network planning. Real progress will depend on those decisions. The announcement of the three new data centers, together with the rest of the initiatives in the pipeline, places Aragón at a decisive moment to consolidate its presence on the European cloud map. The investment is notable and so is the promised employment, but much of the result will depend on decisions that are not entirely in the hands of the community. The region has shown intention and movement, although it remains to be seen what the real scope of this bet will be. Images | İsmail Enes Ayhan | Jorge Azcón (X) In Xataka | The European Commission’s pendulum with AI is real: it will sacrifice privacy to … Read more

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

almost a million employees for a colossal project

by

BYD surpassed 968,900 employees by the end of 2024, more than Tesla, Toyota, Ford, BMW and Renault combined. It is a figure that would be absurd for any traditional automaker, but it has a simple explanation: BYD is not just an automaker. Methodological note: This figure includes the entire BYD Group, not just BYD Auto. Unlike manufacturers that clearly separate their divisions, BYD operates as an integrated ecosystem where the majority of employees are directly or indirectly linked to the automotive business. The company does not publish a breakdown by division. Why is it important. The Chinese company has built the largest vertical integration ecosystem in the automotive industry worldwide. Tesla, for example, buys batteries from Panasonic and chips from third parties, but BYD manufactures absolutely everything in-house. Its business model covers four complete industrial sectors: The automobile division is only the most visible, but also manufactures electronics for Apple. And it produces batteries as the second largest manufacturer in the world. It also develops components from semiconductors to heat pumps. More than 110,000 of its employees work exclusively for Apple’s supply chain, assembling 30% of its iPhones and iPads. The BYD figure, as we said at the beginning, is for the entire BYD group, not just BYD Auto. Other manufacturers have their own nuances: Volkswagen’s 656,134 employees only include the automotive group (Volkswagen, Audi, Porsche, Skoda, etc.), and not large divisions such as finance. However, it does include small divisions of components, such as engines or transmissions. The 389,144 employees of Toyota are only from Toyota Motor Corporation, it does not include Industries (components, textiles, forklifts, etc.), Aisin (transmissions, brakes), Denso (electronic components) or Boshoku (interiors). The complete Toyota ecosystem would be about 800,000 employees. The context. BYD applies a long vertical integration: They internally produce batteries, semiconductors, software, heat pumps, electric motors, control systems, fast charging and structural elements. This strategy allows them total quality control, cost reduction and speed of innovation superior to rivals that depend on external suppliers. Its eight factories in China have tens of thousands of workers. Each one. The Zhengzhou plant has 60,000 employees and plans to hire 20,000 more. Furthermore, they have created authentic “industrial cities“with housing, services, commercial establishments and sports facilities for workers. In figures. The numbers justify the strategy: Employee growth: 37.73% in 2024 (265,400 new workers). R&D personnel: 110,000 engineers, the largest in the automotive world. Cars sold: 4.27 million units in 2024. Go deeper. The contrast with its competitors is striking: Tesla laid off 15,000 employees in 2024. Ford plans to eliminate 4,000 jobs in Europe. Renault is considering laying off 3,000. BYD continues to increase its workforce. However, direct comparisons are complex due to the different business structures mentioned above. They have hired almost 50,000 recent college graduates in two years in China. In the rest of the world they are also advancing: they are building factories in Hungary (2,000 jobs), Mexico (10,000 planned jobs), Brazil, Thailand and Indonesia. Its expansion model, as we have already seen in Spaingoes through the total localization of the production chain. The question is whether this model is sustainable in the face of increasing automation. In fact, Tesla has clearly shown its approach to robotization, but BYD seems to prioritize the human factor for the moment. At least for its model of total control of the productive ecosystem. Featured image | Tiago Ferreira In Xataka | If the question is “would I pay 100,000 euros for a BYD” the answer is “recharges in five minutes.” And we will see them next year

There is a reason why Oracle has become the new beautiful AI girl. That reason is called Project Stargate

by

OpenAI has reached an agreement with Oracle for which it has pledged to hire computing capacity worth $ 300,000 million within five years. The data come from The Wall Street Journalwhere sources close to that agreement cite. The implications are huge. Why is it important. Yesterday Oracle shot in the stock market after announcing that he had had good results, but above all Future income would increase by 359% up to 455,000 million dollars. Although there were suspicions about what could have happened to that extraordinary growth, we now know what happened. And what happened is the Stargate project. The numbers do not fit. The agreement will begin in 2027 and priori seems very risky, especially since OpenAi has no money. Its annual income will be about 10,000 million dollars in 2025, but will have to pay for 60,000 million annually to Oracle to comply with the agreement. Not only that: Oracle will also depend on receiving that income, because now a large part of them will come as a single client. As they point out in WSJ, Oracle is likely to have to borrow to buy the chips necessary to create those data centers that OpenAi will use as part of the agreement. 4.5 GW of calculation power. The contract will make OpenAi access a 4.5 GW computing capacity. It is a way to measure the energy consumption of the center or data centers that will be used to serve the firm’s AI models. The figure is huge, and to put it in perspective, it represents almost 40% of the entire current computing power of Europe, which amounts to 11.9 GW According to IEA. The Stargate Resurge project. The announcement of the Project Stargate, which raises an investment of 500,000 million dollars in data centers for AI, returns to the present. The accounts seem not to leave Despite the recent idyll Between Openai and Softbank, but in July the company of Sam Altman announced that his company and Oracle They had arrived to “an agreement to develop 4.5 additional capacity of capacity in Stargate data centers in the United States.” More bubble indications. The agreement between Openai and Oracle is absolutely huge and confirms that disturbing tendency of the extraordinary expense in data centers that we have been living for months. That figures dance states with the fact that for now all the companies of AI are burning money as if there were no tomorrow And they are far from profitable. That has caused alarms and alerts to be triggered on a IA bubble potential. Microsoft begins to go to the background. The agreement confirms that progressive separation Between Openai and Microsoft, which after the multimillionaire investment of 2023 had become the great technological partner of OpenAI. Each of these companies this now working in their respective plans b. The contract between Openai and Oracle makes that separation of roads once clearer. Doubts about Oracle. This agreement has turned the company into the new beautiful AI girl, but it is not clear if you can fulfill a commitment as ambitious as the one that arises with this contract. To begin with, Oracle has many more debts regarding the money in cash, and its debt ratio on heritage (Debt to Equitiy) is 427%, a very high figure. Microsoft in comparison has 32.7% according to S&P Global Market Intelligence data cited In WSJ. A huge challenge. In addition Microsoft had a cash flow of 136,000 million dollars for the fiscal year that ended this June, while its Capex was 88,000 million dollars. Meanwhile, Oracle had a cash flow of 21.5 billion dollars and its Capex was $ 27.4 billion. Its capex in the coming years will have to shoot, and it is evident that the company has a huge challenge with this project. Image | Oracle In Xataka | If the question is whether there is an AI bubble, Sam Altman has just answered. One with which he is winning

The largest nuclear fusion project on the planet has survived the setbacks. This is the date on which Iter should be ready

by

2024 was a difficult year for ITER (International Thermonuclear Experctor reactor). This experimental reactor of nuclear fusion It is being built in the French town of Cadarache by an international consortium Led by the European Union. Although it was conceived in 2006 and the project was officially launched in 2007, the beginning of the assembly of this titanic machine did not start until 2020. The initial itinerary Proposed by Eurofusion, which is the institution that is responsible for promoting and supporting the scientific research necessary to bring to fruition the European Nuclear Fusion Plan, established that in 2025 the assembly of this machine would end. However, that same year another crucial milestone would arrive: the first tests with plasma would start. Three years later, in 2028, Iter engineers would begin the low power with hydrogen and helium, and in 2032 the first high -power experiments would arrive with these two gases. Finally, in 2035, Iter would be able to undertake high power tests with deuterium and tritium. And in 2040 this experimental reactor would demonstrate the energy profitability of nuclear fusion. Finally this will not happen like this. In 2022 the French Nuclear Safety Authority (ASN) identified several irregularities of a strictly technical nature in Vacuum Chamber sectorswhich caused the Iter organization to react as it should do so: constituting a working group to address the complementary requests of the ASN and advance with the reactor assembly Tokamak. Iter’s technical challenges are unpublished Assembling a machine as complex as it is it is not easy. The vacuum chamber weighs 8,000 tons, is made of stainless steel and boron and must remain hermetically sealed. Its assembly has forced engineers to deal with extraordinarily strict local tolerances of 0.1%, and, in addition, the camera has a very complicated shape and uses plates with thicknesses up to 60 mm. To solve the assembly the technicians have had to resort to state -of -the -art technologies, such as the Electron Beam Weldingwhich is welding using an electron beam, or The design of AI models specifically conceived to identify defects in the welds of the camera. The Covid-19 Pandemia that raised very crudely during the 2020s and 2021, and, on the other hand, the technical challenges derived from the completely unpublished nature of much of the components that need to be tuning so that Iter arrives in fruition have caused that The main milestones of this project are delayed. Nevertheless, The current updated itinerary proposes several important dates that interest us know. In 2039 Iter will be able to undertake high power tests with deuterium and tritium In 2034 the first experiments will be carried out in the reactor; In 2036 the magnetic system responsible for confinement of plasma to maximum power will be tested; And finally, in 2039 Iter will be able to undertake high power tests with deuterium and tritium. Initially this last milestone was going to arrive in 2035. Whatever it is during the last year the Iter assembly has advanced at a good pace. In the cover image of this article we can see two of the titanic sectors of the vacuum chamber, although, in my opinion, one of The milestones that this project has achieved This year It was consolidated in May. The superconductor magnets placed on the outside of the vacuum chamber of this nuclear fusion reactor have the responsibility of generating the magnetic field necessary to confine plasma inside. They are also responsible for controlling and stabilizing it. These magnets weigh 10,000 tons and are manufactured in an alloy of niobio and tin, or niobio and titanium, which acquires the superconductivity when cools with a supercritical helium until reaching a temperature of -269 ºC. This requirement justifies the need to put a powerful cooling system like the one that has devised Europe for Iter. In the construction of this experimental nuclear fusion reactor, the US, Russia, China, India, South Korea, Japan and the United Kingdom, but the cryogenization plant have been commissioned by Fusion for Energy (F4E), the organization of the European Union that coordinates the contribution of Europe to the development of Iter, the French company Air Liquide and technical integrated technicians in the Iter structure. Superconductor magnets acquire superconductivity when they reach a temperature of -269 ºC This extreme refrigeration installation will be responsible for supplying liquid helium to 4.5 Kelvin (-269 ° C) to superconductor magnets and criobombs, and also gaseous helium at 80 Kelvin (-193 ºC) to thermal shields. Creobombs are empty ultraalt devices that are responsible for eliminating gases inside the vacuum chamber. To do it They must work at an extremely low temperature. And, on the other hand, the thermal shields are responsible for protecting some critical elements of the reactor, such as superconductor magnets, the heat that emits the confined plasma inside the vacuum chamber. Iter’s cryogenic plant has an area similar to that of a football field (just over 7,100 m²) and contains several 26 -meter high storage tanks. These figures help us intuit how enormous this critical installation is. As we have just verified, without it the nuclear fusion would be absolutely impossible. This Grigory Kouzmenko statementF4E manager, invites us to tie Iter’s future with a reasonable optimism: “We have entered the most exciting phase of the project, in which all the efforts of previous years finally are specified and we can benefit from the collaboration based on the confidence between all the parties.” Image | Fusion for Energy More information | ITER In Xataka | From today Spain has the key to nuclear fusion: Granada’s particle accelerator is already a reality

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.