Mobile Tech & Gadgets

inject false instructions also to hackers

by

During the last two years, he prompt injection It has been the favorite weapon of cyber attackers against artificial intelligence systems. This method involves hiding a malicious instruction within an email, a calendar invitation or a web page so that an AI agent ends up obeying the intruder instead of its legitimate user. The good thing is that this same technique also works to stop attackers. What has happened? Cybersecurity firm Tracebit has published a research in which it demonstrates that an AI agent that is attacking a system can be stopped by simply placing, next to the passwords and keys that the attacker seeks to steal, a text designed to activate the model’s own rejection mechanisms. As soon as the attacking AI reads that text, its security barriers go up and the attack stops in its tracks. Researchers have dubbed the technique context bombing. How it works. The logic is simple to understand although its discovery was not so simple. All major language models have certain issues built into them that they refuse to address under any circumstances, whether for security reasons or for political and regulatory reasons. Tracebit takes advantage of just that by placing near the fake credentials (the same ones they use as bait to detect intruders) fragments of text that push the model into those forbidden areas. According to explained Andy Smith, co-founder and CEO of Tracebit told Ars Technica, with this technique a rejection mechanism is being activated in the context, something that has a strong and forceful effect and from which agents have a hard time recovering. Once that content enters its context, the model continues to reject any subsequent instructions. The examples, depending on the model. Not all models stop at the same thing, so the trap must be adapted to each case. For Western models, such as those from Anthropic or Google, it has worked to include references to the manufacture of biological weapons, such as instructions for producing inhalable anthrax spores. For the models developed in China, however, the key has been to mention “Tank Man”, the man who stood in front of the tanks in the Tiananmen massacre of 1989, a topic that these systems are prohibited from boarding due to censorship by the Chinese Government and that from Xataka we have actually confirmed every time we have had to test a Chinese language model. The results, in figures. Tracebit tried the technique about five of the most advanced models of the moment (Opus 4.8, Gemini 3.1 Pro, GLM 5.2, DeepSeek 4 Pro and Kimi 2.6) simulating an Amazon Web Services (AWS) environment and entrusting them with routine development tasks that, along the way, led them to come across false keys. In 152 recorded attack attempts, the investigation point that: Attacks that gained administrator access fell from 57% to 5%. Those who also managed to leave a permanent back door went from 36% to 1%. The percentage of attempts that completed at least some phase of the attack dropped from 91% to 15%. The most striking case was that of Opus 4.8, the most effective model of the five when it came to attacking, since without the context trap it achieved administrator access in 93% of the attempts; with it, it failed in 100% of the cases. Vulnerability by nature. The beauty of this technique is that it exploits a weakness that is very difficult to correct. And according to they count According to researchers, a developer can retrain a model to be less strict when faced with a security question, but it is much more difficult to eliminate the political or regulatory restrictions that come as standard, because they are not a failure that can be “fixed”, but rather a deliberate design decision. That makes these barriers a fairly stable point of support for those who are dedicated to developing protections against these attacks. Background. In May, Tracebit already had presented a “canary” systemthat is, fake cloud resources that look legitimate but are not actually used, so that any attempt to access them triggers an alert for the security team. In that first test, with 951 attack attempts, the canaries gave an average of eight minutes’ notice before the attacker gained administrator access. The problem is that the AI ​​agents needed only 14 minutes on average to complete the attack, so the defenders had a window of just six minutes to react. Context bombing was born precisely to gain that missing time, stopping the attack instead of simply warning of it. What this does not solve. Neither Tracebit nor anyone else claims that this is the definitive solution to prompt injection. There is still no way to prevent a model from occasionally confusing an instruction with any piece of data within its context, and that confusion is precisely what makes the technique possible. Cover image | AI-generated with Gemini In Xataka | “It could weaken US competitiveness”: Kimi K3 forces Trump to resume his plan to stop Chinese AI

We don’t know if today’s pop music is better or worse than that of 1960, but we do know one thing: it’s much more depressing.

by

What does the music you listen to say about you? And what does the music that is published say about us, as a society? A group of scientists from the Queen Mary University of London (QMUL) has tried to answer these and other questions by analyzing tens of thousands of songs released between 1960 and 2023. The result is a vast study which has just been published in Nature and leaves a fascinating conclusion: in the last six decades pop has become more depressing. The question is obvious: Why? Much more than music. Anyone who has listened to music while working, playing sports or simply walking down the street will have experienced that songs are much more than an artistic expression: they are also a powerful lever capable of altering our mood. Not only that. Music, from the most sophisticated to the most popular, points directly to our emotions and obsessions. Aware of this, a group of researchers including Vjosa Preniqui, from Queen Mary University, decided some time ago to embark on a peculiar exercise: grab tens of thousands of songs mainstream and analyze their lyrics to find out what they are talking about. Almost 400,000 topics. The team examined more than 380,000 songs published between 1960 and 2030, hours and hours and hours of music composed by artists from several generations who sing about their obsessions. To be more precise, the researchers worked with two large music archives. The first consists of more than 377,000 songs with lyrics in English released between 1960 and 2010 and that have been added to the WASABI database. The second is made up of 5,500 songs extracted from the Billboard charts and that cover a longer period that extends from the beginning of the 60s to 2023. Objective: language. To manage such a musical collection, the researchers They turned to AI and language analysis techniques. They also focused the focus on a very specific aspect of the songs: they were not as interested in the harmony, instrumentation, rhythm or complexity of the compositions, as their lyrics. The focus was on language, emotional language and its relationship with morality. That is, what were the songs about? Of fidelity, decency and friendship or of conflicts, deception and rebellion? The goal, as Preniqi and his colleagues explain in his paperis to examine “the evolution of moral expressions in pop music lyrics” over six decades. It may sound very abstract, but the researchers paid attention to ten great timeless topics: ‘care’, ‘harm’, ‘justice’, ‘cheating’, ‘loyalty’, ‘betrayal’, ‘authority’, ‘subversion’, ‘purity’ and ‘degradation’. They also looked at the artist singing: whether it was a man, a woman or a mixed group. What did they find out? That today’s popular music is much darker than what our grandparents and parents listened to. By comparing the lyrics of the 60s with the most recent ones, Preniqi and her companions found that they have become “increasingly negative”. “Our results reveal an increase in moral vices and a decrease in expressions of virtues, along with an increase in negative feelings, anger and disgust,” they explain in the article of Nature. In short, today it is much easier to listen to songs that are about infidelities, hurts or degradation than in the 60s, when it was more common to sing about purity or mutual care. “This change, which extends until the 2020s and varies depending on the genre of the artists, highlights that expressions oriented to vices and emotional charges have increasingly gained prominence in music lyrics. mainstream“, concludes the study. Can it go further? The research provides some figures that help to better understand this transformation. In WASABI the letters related to degradation shot up by 52%, those associated with damage by 49% and those referring to deception by 48%. References to subversive acts also rose 41%. In the Billboard lists, which extend the ‘photography’ until 2023, the panorama is not very different: references to deception skyrocketed by nearly 71%, to degradation by 62%, to subversion by 50% and to injuries by 36%. less uplifting. If we look at the lyrics that deal with ‘virtuous’ behavior the trend is the opposite: Songs that talk about caring for others decreased by 24%, those dedicated to purity by 12% and those about loyalty by 11%. Again the trend is similar in WASABI and Billboard. The researchers did notice changes depending on gender of the artist (men and mixed groups seem more given to negative lyrics) or musical genresalthough these latter data must be handled carefully because there may be important imbalances in the samples analyzed. And what do the authors say? They claim that their study helps us understand both the evolution of popular music and our own obsessions. “Music is much more than entertainment, it is one of the ways that societies tell stories about themselves. By looking at song lyrics over several decades we can see how emotional expression and moral narratives evolve over time,” reflect Preniqi. “What we discovered was a gradual shift away from language associated with virtues such as care and decency toward themes that reflect conflict, harm, and other moral concerns. These patterns depend on several factors, such as musical genre and impact, but offer a fascinating perspective on the evolution of cultural values ​​and emotional expression.” Images | Spencer Davis (Unsplash), QMUL and Vjosa Preniqi et al. in Nature In Xataka | In 1992, Black Sabbath recorded a song about technology replacing humans. In 2026, it serves as a premonition

Hollywood has exhausted the vein of superheroes. Now it’s the turn of terror and youtubers

by

The film industry is always looking for its new great vein and it seems that, when a formula begins to show signs of exhaustion, a genre appears ready to pick up the baton: the terror. Superheroes are already a model that seems more than worn out and, of course, they are no longer that infallible answer that dominated the box office for a decade (tell that to James Gunn and his ‘Supergirl‘). So the mission is now to find that next great franchise capable of drawing audiences to theaters. In fact, it is not the first time that Hollywood finds some oxygen in horror, since when it needs to reinvent itself it turns to it as a true laboratory of ideas. It happened with him found footagewith the auteur horror that enlightened an entire production company like Blumhouse and, now, it seems that the next revolution comes thanks to a generation of creators who learned to tell stories on YouTube before in a film school. It is true that with the last three leading films (‘Iron Lung’,’Obsession‘ and ‘Backrooms‘), and with the numbers in handit’s easy to draw that conclusion. But what is perhaps worth pointing out is that this is not a phenomenon of this last year nor is it a mere fad. Horror has been a testing ground for the industry for years; experimenting with new technologies, riskier production models and innovative narratives. The difference is that now the gateway is not a festival like ‘Sundance’, a prestigious school or a certain surname. This new generation has learned to edit, create, edit and script from their bedroom. The proper names Curry Baker already pointed out ways with his footage ‘Milk&Serial’a found footage filmed for just $800 and published for free on YouTube, which has more than two million views. But there was nothing to suggest that his next project, ‘Obsession’, would end up becoming one of the phenomena of the year. Its twisted and innovative version of the classic “be careful what you wish for” has not only skyrocketed the popularity of its protagonist, Inde Navarrette (he is already chaining new projects and even stars in the new Sombr video clip along with one of the stars of ‘Off Campus‘), but it has also turned the box office upside down. With 371.2 million dollars raised around the world has already surpassed ‘Sinners‘ becoming the highest-grossing English-language original live-action production of the decade. And, if that were not enough, it has achieved something that has not happened since 1982: increasing its collection during each of its first three weekends. Not far behind ‘Backrooms’ Kane Parsons. The Californian opened his YouTube channel when he was barely ten years old, but it was in 2022, when he was only 16, when he caught the attention of the internet with his short film ‘The Backrooms (Found Footage)‘. Inspired by the popular creepypasta about an infinite labyrinth of offices and alternative dimensions that began to gain traction in a forum like 4Chanended up redefining the analog horror and accumulating tens of millions of views. Four years later, Parsons not only dared to direct ‘Backrooms’, he also signed the script with Roberto Patino and participated in the soundtrack. The result, another resounding success, with 330 million dollars raised, becoming the film highest grossing of A24 with its first version. The third name of this batch is Mark Fischbach“Markiplier”. In this case, his fame does not come from a viral short film, but from one of the largest communities on YouTube, where he has accumulated close to 40 million subscribers. Accustomed to writing, editing and even financing his own projects, he decided to make the leap to feature film with ‘Iron Lung’adaptation of the indie horror video game by David Szymanski published in 2022. And so it was formed another success for this generation of filmmakers that emerged from the Internet, sneaking into second place at the US box office in its first weekend in theaters. Hollywood adapts communities When Hollywood wants to experiment, it does so with terror; but now the studios have an advantage. Betting on Barker, Parsons or Markiplier is not jumping into the void, they all come with a community behind them that has already validated their work. They have learned with Adobe Premiere, and not by taking their first steps in large productions, but they have also perfected a language that connects naturally with a generation that has grown up with the Internet. There, those rhythms inherited from YouTubea found footage taken further, liminal spaces and a more fragmented narrative that young audiences instantly recognize. In the end, you are buying into a different way of telling stories. Fischbach writes, directs, edits, produces, finances and stars in much of his own work. He released ‘Iron Lung’ through his own production and distribution company, Markiplier Studios. It turned its audience into the driving force of distribution, as its 39 million subscriber base undoubtedly had something to do with its $50 million box office success. Parsons, with his 1.7 million subscribers, creates a phenomenon from scratch with a very well-built audience accustomed to what the author had to offer; those found footage evolved and that particular aesthetic. So, despite not having such a huge community when Hollywood knocked on his door, he built it based on a very specific idea. “What has happened here shows that it is possible to create a niche in the network, make it work sustainably, and then scale it when more resources come in. I’m sure there are a lot of people right now trying to figure out how to repeat that phenomenon, how to make lightning strike twice.” Kane Parsons These phenomena are the next great mutation of the genre. What’s more, Hollywood studios are betting on digital assets and are not just limited to Tiktok or YouTube: now their radar points to Reddit. They search the platform for short stories and ideas that can become films after happened this last year; come on, a’blood … Read more

A startup has created a machine that makes gasoline with air, water and electricity. It is your bet to save the combustion engine

by

Aircela, a New York startup, has taught the world a machine capable of manufacturing gasoline using only air, water and renewable electricity. The device, which is not exactly small, has already been shown in operation in several public demonstrations and aims to become an alternative to biofuels and a solution for those who are not yet willing to switch to electric. Below these lines we tell you all the details. What has happened? The company presented its technology in 2025. According to collected Popular Science, its CEO, Eric Dahlgren, wanted to make it clear that it was not a simple concept on paper. “We haven’t built a prototype. We’ve built a machine that works,” he said. The goal of Aircela, founded in 2019 by Eric and Mia Dahlgren, is that this fuel can be poured directly into the tank of any gasoline car without touching the engine. How it works. According to they count those responsible, the process is divided into three phases. First, a fan introduces air into the machine, which passes through a network capable of retaining CO2 and generating a liquid carbon solution. An electrolyzer then separates that solution while producing hydrogen from water. Finally, a compressor combines both elements to create methanol, which is refined into ready-to-use gasoline. The result, as detailed by the company itself in different demonstrations, is a fuel without sulfur, ethanol or heavy metals, compatible with current engines without the need for any modification. Between the lines. Current performance is, for now, modest. And each machine captures around 10 kilos of CO2 per day and thus produces a single gallon of gasoline (about 3.8 liters), with a storage capacity of up to 17 gallons (about 65 liters) in the tank, according to share the middle. This means that we would not be able to fill a car’s tank for several days. Furthermore, just as share In Futurism, the process consumes approximately twice as much energy as is ultimately stored in the fuel, putting its efficiency below 50%. Even so, Aircela maintains that using independent solar panels, the electricity cost would remain below a dollar and a half per gallon. The but. Dahlgren insists that its technology does not compete with the electric car, but rather complements it in those uses where replacing the vehicle is not viable in the short term, such as freight transportation, motor racing or areas without sufficient electrical infrastructure. In fact, the ABC7 media points out that the company has no intention of manufacturing larger machines to produce more fuel, but rather of multiplying the number of units, something similar to how a solar farm works. Who is behind. Aircela has managed to raise nearly $5.5 million to date, according to data from Pitchbook. Its investors include Chris Larsen, founder of Ripple; Jeff Ubben, ExxonMobil activist advisor; and Maersk Growth, the venture capital fund of the Danish shipping company AP Moller-Maersk. Morten Bo Christiansen, head of energy transition at Maersk, explained in a statement collected by Carbon Herald that the shipping company invested because of the startup’s innovative approach and that it hopes, over time, to go from being an investor to becoming a client. The technology, furthermore, draws directly from the work of physicist Klaus Lackner, pioneer of direct air capture and Dahlgren’s former mentor during his PhD at Columbia University. And now what. There is still no talk of commercial deployment, but the first uses of this machine would be focused on industrial, commercial clients and areas without access to the electrical grid, according to share from Interesting Engineering. Aircela has also negotiated with several car manufacturers, including Jaguar Land Rover, as well as an American manufacturer and another German whose names it did not reveal, according to they claim from Observer. The price of each unit could range between 15,000 and 20,000 dollars, as share from Futurism, although Aircela is confident that mass production will lower the cost over time. Images | Aircela In Xataka | The wildest work in the North Sea: a colossal artificial island where half of Europe wants to plug in to receive energy

“The world doesn’t realize how much the AI ​​industry depends on scientists in China” Jensen Huang, CEO of Nvidia

by

When someone as important to technology as Jensen Huang speaks, the bread rises. The CEO of Nvidia is a person who, although he is not known for giving headlines as controversial as his colleague Elon Musk, does say interesting and important things to understand the sector. At the beginning of the year he sat with Time to chat at length about the AI: whether or not it is a bubble, its impact or who is winning this race. In that interview China appeared and the commercial dispute that your company has with the Asian giant. There he made something clear: the world needs China more than the other way around. An example: his armyChina is achieving technological independence while its American counterpart increasingly resorts to Chinese imports. Thank China. When it comes to AI, Huang is clear: the providential weight of Chinese talent in the research and development of artificial intelligence. The CEO of Nvidia affirms that much of the world’s progress in artificial intelligence depends on Chinese researchers, and that the interdependence between the economies of China and the United States is greater than the public perceives. And he sums it up like this: “The idea that the United States can disengage from China is wrong.” Why it is important. Because the position of the boss of the most valuable company in the world, who is also American, collides with the toughest wing of the White House, which every now and then comes with new restrictions on the export of chips to China. If the industry’s reference supplier depends on symbiosis instead of separation, it is a warning to sailors for those who promote those obstacles in the form of laws that seek to limit Chinese access to the most advanced processors. Context. Note that this was not the first time that Jensen Huang argued for Chinese supremacy in AI. Already in May of last year at the Washington Forum claimed that 50% of the world’s Artificial Intelligence researchers are Chinese and that this data should influence “how we think about the game” of technological competition. Months later, in an interview with Citadel Securities collected by Fortuneinsisted that it is a mistake not to let these researchers build AI on American technology and that those policies that seek to harm China often also harm the United States. In detail. In that same interview Huang also made it clear that Taiwan will continue to be the centerpiece of AI chip manufacturing, despite the noise and interest from the White House in relocate production to the United States. In fact, in other interviews he has already clarified that manufacturing in America does not mean abandoning Taiwan, but rather spreading the risk. A complement. After all, You cannot abandon its highly refined production chain overnight as Trump intends. The small print. Jensen Huang has repeated the same speech on several occasions and scenarios, but as CEO of Nvidia we cannot lose sight of a reality: the commercial interest of his company. The Asian giant’s market share was enormous before export controls began, going from 95% to 0%. So Nvidia is the main one interested in defending openness. On the other hand, and although that 50% is real and matches different measurements Regarding talent in AI, this weight measures origin or training, not where these scientists work or what banner they have for the advances they achieve, since there is a large number who carry out their work in American universities and companies. In Xataka | “We chose the wrong technology”: Jensen Huang returns to Japan to acknowledge his debt to Sega 30 years later In Xataka | Jensen Huang enters the Samsung salary controversy: “Workers should earn as much as possible” Cover | Wikipedia and Gary Lerude on Flickr

Google is making money by betting on its rivals. Some already give you more profit than your own business

by

Alphabet published a few days ago its latest quarterly resultsresults that break all their records with 112.1 billion dollars in net profit, 298% more than a year ago. The surprise lies in the origin of this profit: almost all of it comes from the accounting revaluation of two stakes in companies that Alphabet does not control: SpaceX, which is already listed on the stock marketand Anthropic, which doesn’t do it yet but he is expected to take the step soon boosted by a financially spectacular 2026. For Alphabet, advertising grows 24% in the company’s revenue. Google Cloud soars 82%. And yet, 88% of the quarterly net profit comes from the revaluation of two financial bets that today compete with central areas of the group. In detail. Participation in SpaceX goes back a long way. In 2015, Google invested 900 million when the company was valued at 10,000 million. After the June IPO, at a valuation of 1.77 billionthat package is worth 94.1 billion. A multiple of more than one hundred times the original capital that rewarded Google’s good eye more than ten years ago. Anthropic’s is a shorter story. And more expensive. Alphabet has injected 13.3 billion into it from April 2023with commitments for up to an additional 30,000 million. The private assessment of the laboratory behind Claude It has gone from 350,000 to 965,000 million in one year. In 10-Q, Alphabet’s private holdings total 124.3 billion. Virtually everything is Anthropic. Between the lines. On the B-side of the album, Anthropic is signing engineers from DeepMind quite regularly. Its model competes with Gemini in the same field (API for developers, programming agents, business assistants), and a good part of the companies that do not want to be tied to Google choose it precisely for that reason. Every dollar that Anthropic collects from Alphabet also increases the value of Alphabet itself. SpaceX instead operates starlinkthe satellite network that supplants the ambition that Google tried and left with Project Loon. Sundar Pichai has not sold a single share of SpaceX or Anthropic. And yet, the market has recognized almost 100,000 million in profit from them. Yes, but. Capital gains are not cash. Of the 94.1 billion in SpaceX, 80,000 cannot be sold in the short term. And 14,100 are blocked until the third quarter of 2027. Anthropic remains unlisted, so the valuation is worth until the next round or until the next cut. The royal treasury, meanwhile, is being emptied. Alphabet has raised its investment guidance for 2026 to the range of 195,000-205,000 million (from the 180-190,000 announced in April), and quarterly free cash flow has gone into the red. The accounting balance of the party is paid by the data centers. And now what. The restrictions on SpaceX are beginning to be lifted after its first presentation of results as a listed company, scheduled for August 4. With 94 billion in shares blocked and an investment plan that is close to 200 billion annually, Alphabet has obvious reasons to sell part of the position. As soon as possible. Doing so implies an uncomfortable admission because it recognizes what this quarter already leaves black and white: that the company has earned more by financing its rivals than by running its own business. And with his business it is not that he has earned little. Except that the free cash flow is negative due to the enormous investment that the AI ​​career requires. In Xataka | The EU has just fined Google 890 million for the DMA. The figure weighs less than a paragraph almost hidden in the statement Featured image | Xataka with Magnific

How they are alike (and what they are not) beyond sharing a last name

by

Now that Samsung has introduced the new generation of their smartwatchit is good that we stop for a moment to compare them with their peers in the competition. In this article we are going to start with the best model of the brand, the Samsung Galaxy Watch Ultra2to compare it with the Apple Watch Ultra 3. The differences between the two smartwatches Variety of sizes and their screens The biggest difference between both watches is in the design, since while Samsung is committed to a round formatApple maintains the square design that we see in all their smartwatches. The sizes are quite similar, finding 47.4 mm cases for the Galaxy Watch Ultra2 and 49 mm for the Apple Watch Ultra 3. The screens have different technologies. Samsung is betting on a 1.52-inch Super AMOLED panel with a maximum brightness of 5,000 nits, while the Apple Watch incorporates a 1.99-inch Retina LTPO OLED panel with a brightness of up to 3,000 nits. With this we can see two things: The Samsung Galaxy Watch Ultra2 case is larger and its screen is smaller. On the other hand, its screen offers a higher level of maximum brightness, so at certain times it can be seen better outdoors. Target audience Both smartwatches are aimed at a sporty publicsince they are the models within each brand that offer the best features in terms of sports functions and health sensors. In any case, we are talking about two quite complete models. Perhaps the Apple Watch Ultra 3 has satellite connectivity, but this is something that we are not always going to take advantage of. Both watches offer a good variety of options when it comes to health monitoring. More battery One of the biggest differences between the Samsung Galaxy Watch Ultra2 and its previous generation is the battery, which has grown considerably to a capacity of 800 mAh. According to Samsung, with the new processor (it has gone from Exynos to Qualcomm) it promises up to 60 hours of uninterrupted autonomy. On the other hand, the Apple watch falls a little short, and those from Cupertino maintain that their smartwatch promises a theoretical autonomy of up to 42 hours. However, these figures are theoretical, so in practice they will vary depending on the use we give to each watch. Ecosystem Unlike what we see in smartwatches from other manufacturers, choosing between an Apple watch or a Samsung one will depend above all on the mobile phone we have. Samsung reserves some functions for Galaxy mobile users, such as sleep apnea, while Apple does the same with precision search or route synchronization with the Maps application. In any case, we can do this simplify even more: The Samsung Galaxy Watch Ultra2 only works with Android phones, while the Apple Watch Ultra 3 only works with iPhone. Prices As these are high-end smartwatches, the prices we find in both cases are very high. However, the Samsung Galaxy Watch Ultra2 can boast of being a little cheaper, since it has been launched for an official price of 749 euros. Instead, the Apple Watch Ultra 3which has been on the market longer, is currently 829 euros (or more depending on the store). However, on the occasion of its launch, the Samsung Galaxy Watch Ultra2 can now be found on sale in some stores. Which smartwatch to choose? Why choose the Samsung Galaxy Watch Ultra2 The Samsung Galaxy Watch Ultra2 is an evolution of the previous generation. These are the reasons why we would choose it: If you have an Android mobileespecially if it is a Samsung Galaxy. The reason is that Samsung reserves some functions for its smartphone ecosystem. Your batterywhich has grown quite a bit compared to the previous generation. To do sportsespecially considering its sports functions and health sensors. Samsung Galaxy Watch Ultra2 The price could vary. We earn commission from these links Why choose the Apple Watch Ultra 3 The Apple Watch Ultra 3 is an excellent smartwatch, both in terms of design and everything it offers. These are the reasons that would make us opt for it: If you have an iPhonesince it is only compatible with its smartphone ecosystem. Your screenwhich is better used (the box is smaller). Your designespecially for reading notifications thanks to the fact that it has a larger, square screen. To do sportsfor the same reason as the Samsung watch: its sports functions and health sensors. The price could vary. We earn commission from these links Technical data sheet for the Samsung Galaxy Watch Ultra2 and the Apple Watch Ultra 3 samsung galaxy watch ultra2 Apple Watch Ultra 3 dimensions and weight 47.4 x 47.1 x 10.7mm 61.5 grams 49x44x12mm 61.6 grams screen 1.52 inch Super AMOLED Resolution 498 x 498 pixels Always On Up to 5,000 nits sapphire crystal 1.99-inch Retina LTPO OLED 502 x 410 pixel resolution Up to 3,000 nits processor Snapdragon Wear Elite Apple S10 ram memory 2GB Undetermined internal storage 64GB 64GB Sensors Samsung BioActive Sensor temperature sensor Accelerometer Barometer Gyroscope Geomagnetic sensor Light sensor. SpO2Optical heart rate sensorEKGAccelerometerGyroscopeAltimeterDual frequency GPSCompassAmbient light sensorVO2maxtemperature sensor Battery 800mAh 42 hours of autonomy compatibility Android 13 or higher 1.5 GB of RAM or more iPhone 11 or later operating system Wear OS 7 One UI 9 Watch WatchOS OTHERS Auto Brightness AoD mode Heart rate monitoring Sleep monitoring SpO2 monitoring Stress monitoring vascular load Sleep pain Hearing Bioelectrical impedance Fall detection NotificationsMusic control Microphone Speaker More than 100 sports modes Google Gemini Google Play Emergency call function Fall detection Car accident detection Triple microphone Customizable rotating crown PRICE 749 euros From 829 euros Samsung Galaxy Watch Ultra2 The price could vary. We earn commission from these links 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 | Samsung, Apple, Pedro Aznar in Applesfera In Xataka | Best mobile phones 2026. Which one to … Read more

the map that shows who is in charge and who has been left out

by

When AI arrived and we saw everything it could do, the first impression of many was that it was going to take away their jobs (and the truth is, some statements didn’t help either), although the reality at the moment is different: we work more and rest less. Knowing how to use AI has become another skill, one that can be essential depending on your job: being a waitress is not the same as being a secretary or being a plumber or journalist. But even if the profession matters and the deployments are more or less global (in fact, Europe often lags behind as a matter of legislation), AI has not reached the same level throughout the world. The Microsoft AI Diffusion Report that monitors the AI ​​economy on a global scale (specifically, 100 countries) for the first quarter of 2026 collect a fact: 17.8% of the world’s working-age population uses it, but there are 10 countries that lead it and many that are behind. Of course: Microsoft’s report uses anonymized telemetry from its products. That is, if someone uses only ChatGPT, Gemini or Claude and never touches anything from Microsoft, they do not appear in the count. Microsoft tries to correct this bias with statistical adjustments, but it does not completely eliminate it. Microsoft’s report looks better with this infographic adaptation by Visual Capitalist what colors countries based on their use of AI: yellow tones for those places with high adoption and purple ones where it is little used. It is, in a nutshell, a quantitative snapshot of the actual use of generative AI tools. It is inevitable to associate the adoption of AI with labor productivity and competitiveness. Without going any further, Microsoft itself does it relating its real effects on employment: dev work in the US broke records in 2025, reaching 2.2 million, 8.5% more than the previous year. Obviously, this gap between the countries that use AI the most and those that use the least has as a background unequal access to this technology. In short: it is a more than perpetual digital divide and widens the inequalities that already exist. Which countries lead the use of AI Adoption of AI in the world by country. First quarter of 2026. Microsoft via Visual Capitalist Among the countries that lead the use of AI, one stands out: the United Arab Emirates, with 70% of its working-age population using artificial intelligence and growth that quadruples the world average rate. They are followed by Singapore (63%) and a group of highly digitalized countries, mostly European: Norway (49%), Ireland and France (48% each), Spain (44%), New Zealand (43%), and the United Kingdom, Netherlands (42%). Closing the top 10, Qatar tied with the Netherlands. The pattern is clear: they are moderately rich economies, with good internet connections and governments that have been investing in digital infrastructure for years. But the vast majority moves well below those figures and what better example than an a priori unexpected United States: it only has a 31% share despite being the cradle of almighty powers such as Google’s Gemini, Anthropic’s Claude or OpenAI’s ChatGPT. It is one thing to build artificial intelligence and quite another to use it.. Here it depends more on having a homogeneous digital infrastructure, state adoption policies and territorial equality, factors where small states like the UAE or Singapore have structural advantages. In the queue are many countries from Africa, Central Asia and Latin America, with quotas of just one digit or little more: Venezuela 10%, Russia 10%, Morocco 12%, Bolivia 13%, among others. The digital divide between north and south, at its finest: the Global North reached 27.5% usage compared to only 15.4% in the Global South. AI is spreading, but in a heterogeneous way towards those who were already advanced in digitalization. However, it is worth remembering again that since it is based on Microsoft product telemetry, this indicator may underestimate real adoption in markets where other providers predominate: China deserves special mentionwhere the AI ​​ecosystem has powerful local products such as Baidu, Alibaba, Deepseek or Moonshoot and his impressive Kimi K3. In Xataka | Who is really winning the AI ​​race, in a graph that puts Google in trouble In Xataka | If the question is which of the big tech companies is winning the AI ​​race, the answer is: none Cover | Visual Capitalist

the cities of “the blinds down” arrive

by

In Damgana small coastal town in Brittany, France, some time ago they began to paint their shutters red, blue or green, more colorful options than the white that has traditionally been used for facades. “We had them painted,” the mayor recognizesJean Marie-Labesse. That chromatic change may seem like a minor detail, but it says a lot about the challenges that the town faces. If you have decided to give the blinds a cheerful tone it is because Damgan It is a holiday town and many of its windows are closed almost all year round. By giving a touch of color to the blinds, the City Council intends (at least aesthetically) that Damgan does not look like what it is: a town plunged into a comatose state almost all year round due to the weight of second homes. A percentage: 74%. Damgan case made headlines in France just a year ago and, beyond the changes it may have experienced in recent months, it is interesting because it tells us about a phenomenon that affects many other tourist cities, both within France and in the rest of the world: the “boom” of the second residence at the expense of the habitual residence. In the French town, yes, the trend is pronounced. Le Parisian assures that in 2025, 74% of homes were second homes, a percentage that reached almost 80% before the City Council took action on the matter. From 1,900 to 30,000 inhabitants. In practice, this percentage translates into houses and more houses closed tight for most of the year, thousands of blinds drawn in a fade to white which the City Council has wanted to avoid. The data is eloquent. It is estimated that of the 4,000 homes of the town, 3,000 are second homes. Translated into inhabitants, this means that the town goes from just 1,900 residents in winter to 30,000 during July and August. Why is it important? Because Damgan leaves a barbaric example of something else: how the imbalance between seasonal and stable housing ends up taking its toll on the city, even threatening some of its basic services, such as schools. Labesse remember For example, the year he became mayor (2014) only one baby was born in Damgan and one of the school classes that the town enjoys ended up hanging in the balance. “It’s something crucial for the people.” Another front in which the overwhelming weight of vacation housing is felt is the real estate market. Le Parisian relates the case of a 50-year-old carpenter from the area who, after separating, was unable to find an apartment to move into. He has a good income, but that does not help him much. “The workers can’t live here, they have to go 20 kilometers away. We are becoming a town for seniors,” regrets Hervé du Souichfrom ‘Les Volets Ouverts’, an initiative that allows locals to stay in empty houses without having to move outside of Damgan. Beyond Damgan. The town in French Brittany is interesting for what it represents, but it is by no means a unique case. According to French Propertyin continental France there were 3.2 million second homes in 2021, almost 10% of the real estate stock. The vast majority are French citizens, especially on the coast, although there are also between 80,000 and 90,000 who belong to Britons. In certain regions of the Mediterranean, such as Languedoc-Roussillon or Provence-Alpes-Côte d’Azur, holiday homes account for 14% of the stock. Not far from Damgan are Carnac and La Trinité, where the percentage of second homes is around 72%. And a good part of that home, the Parisian newspaper states, is only used a few weeks a year. In Villard-de-Lansin Isère, two thirds of the homes function as second homes and the percentage is even higher in some towns in the region. These data leave those seeking stable accommodation in a delicate situation, recognizes Labesse: “The problem is not so much the price as the lack of supply.” Beyond France. France is not a unique case either. In London, second homes and empty houses are such a serious problem that the authorities are already they try to stop it using tax pressure. According to the data managed in 2023 by the mayor’s office, in the British capital there had been some “30,000 empty homes” for a long time, with a high concentration in certain points, such as Kensington or Chelsea. Madison Trust Co. shows that there are areas of the US, such as Marinette-Iron Mountainin which more than 25% of residences are used seasonally or recreationally. Well known in Spain. The phenomenon is well known in Spain. A sensitive part of the residential stock of Ibiza and Formentera, for example, are second homes. In 2023 Ibiza Diary needed that of the almost 86,000 properties in the Pitiusas, around 6,200 were inhabited sporadically and 18,300 were empty. In some cases the photo is more complex, such as in the town of Sant Joan, where 44% of the homes were considered non-main. They are the cities with the blinds drawn. A previous version of this article was published in July 2025. Image | Mairie de Damgan In Xataka | In Madrid there were no longer any cheap hotels, but there was a “virgin” space for tourists: the industrial estates.

In 1966 a player from Argentina was sent off but there was no way to explain it to him. Since then there have been yellow and red cards

by

When two people want to understand each other but do not speak the same language, they try to do so by speaking louder, gesticulating or resorting to things they do not understand about borders. And be careful, this is assuming that in football we always want to understand the referee: it would not be the first or the last time that a player acts the Swede when faced with a warning or a substitution… if he is winning, of course. Despite the sui generis speech by Tom Cruise In the last World Cup 2026, in modern football we manage with English (however tacky) and the rules of the IFAB, made up of the four football associations of the United Kingdom and FIFA, which allows anyone to know what it means when the referee team raises the flag or a card suddenly appears. It wasn’t always like this, so imagine the pitched battles that took place. The problem. An Italian and a Chilean go and… no, it’s not a joke: at the beginning of the 60s there were episodes of great violence in football and one of the most sadly remembered matches was the match between Chile and Italy in the 1962 World Cup, so much so that it is known as the “Battle of Santiago“. At that time there were no mechanisms to put a stop to that aggressiveness in the form of fights and fouls that occurred as a result of the heat on the playing field. The World Cup that changed everything. The quarterfinal match between Argentina and England in the 1966 World Cup was a before and after. As pick up the Smithonianat one point, an Argentine player tried to communicate with the (German) referee and his vehement gestures and pleas, impossible for the referee to understand, cost him expulsion for verbal violence. The player refused to leave the field until Kenneth George Aston, a member of the FIFA Referees Committee and also responsible for the refereeing of that World Cup, intervened. In that match the referee Rudolf Kreitlein had warned the English Bobby and Jack Charlton and had expelled the Argentine Antonio Rattinbut it was so unclear in the game that those interested they found out better laterthanks to the press. In fact, so much so that the England coach, Alf Ramsey, He went to a FIFA representative to clear up any doubts. Of course, the heat resulting from the match, the ambient noise and the language barrier between players and referee made understanding each other an impossible mission. Aston came to that conclusion, and he left there determined to change it. The solution. So he set out to do something to make the referee’s decisions clearer for both players and spectators. And something came to mind that everyone knows: traffic lights. Aston counts how it came to him: “As I was driving down Kensington Street in London, the traffic light turned red and I thought: ‘Yellow’, you can still get through; ‘Red’ means stop, out of bounds.” When he got home, he told his wife Hilda about his idea, who they count He disappeared from the room to return a few minutes later with two cardboard cards that fit perfectly in his shirt pocket, thus materializing the idea. The system officially premiered at the 1970 World Cup in Mexico, 107 years after the birth of football’s first governing body, and since then it has been adopted in all federations. The power of a simple cardboard. The writer Rob Walker summarizes well the effect that that simple piece of paper has: it has enormous power when a referee displays it in an almost theatrical way in the player’s face. It provokes a stadium full of whistles and exaggerated gestures from the arms of the cautioned player and his teammates. Because everyone knows what it means: the loss of possession of the ball, a free kick and even being left with one less player on the field of play. Kenneth George Aston, who died in 2001, summarized the vision of football where that gesture that he himself invented and that is still valid in each and every match has extreme importance: a dramatic work in two acts with 22 actors and a stage director, the referee, without a script and without a known ending. In Xataka | Argentina 6 – Peru 0: the story of the alleged biggest fixing of the World Cups between dictatorships and missing people In Xataka | There is something in which Uruguay is a true and overwhelming world power: generating elite footballers Cover | Jason Gulledge and Tasnim News Agency

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.