longer

Earning $180,000 a year is no longer enough to live in San Francisco. The AI ​​boom has eaten up the technological middle class

by

In San Francisco you can earn $180,000 a year and not make ends meet easily. The AI ​​boom It is making what would be a huge salary anywhere else become mere pocket change, preventing entry into an impossible real estate market and a standard of living unattainable for the majority. Silicon Valley’s dream is to move. In it New York Times They tell the story of Katrine Razniak, a LinkedIn recruiter, and her partner Adam Woodbury, a software engineer. They have been in the city for five years and earn 180,000 and 185,000 dollars respectively, but despite this they are not able to get something as basic as an apartment for two. They spent months looking for a one-bedroom apartment to move in together on a budget of $5,000 a month, but it was impossible. They say they saw one for $5,200 and the waiting list already had 30 interested parties an hour after the open house. Buying a house is not an option either: The median home price is $1.7 million.. Given this situation, the couple is considering moving to another state where their salaries would allow them to live as an upper class. The new elite. While Katrine Razniak and her partner have to continue living in shared apartments, there is a new technological upper class that can afford to live comfortably in the city. They are the new millionaires of the AI ​​boom, employees of companies like OpenAI and Anthropic, with stratospheric valuations and stock options. In October of last year, when OpenAI was worth $500 billionwe already talked about that the city’s real estate market was exploding. Employees began selling stocks and buying homes like there was no tomorrow, causing prices to skyrocket. There is no improvement in sight. Not even a year has passed and Sam Altman’s company worth 852 billion dollars and Anthropic aims for 900 billion. To put it in context, when Uber went public in 2019 it was worth 82 billion, we are talking about ten times that amount. The imminent IPO of the two companies, added to the recent IPO of SpaceX, aims to create new fortunes, until 20 new billionairesaccording to estimates. Working in technology is no longer a guarantee of success in San Francisco, now you have to work in AI (and own stocks). It’s not just the rent. The median salary in San Francisco is also higher than the rest of the country and has been growing over the past six years, going from $153,000 to $196,000 today. However, everything else has also gone up. Aside from housing, life in general isn’t exactly cheap either. According to data from Cost of Living Indexthe cost of living in San Francisco is 65.6% higher than the rest of the United States. Katrine Razniak says her monthly spending has grown by $1,000 without changing her lifestyle, causing her and her friends to replace restaurant outings with home-cooked dinners. The housing problem in San Francisco. Spain is going through a housing crisis unprecedented, but we are not the only country facing this situation; Canada, Germany, Mexico, China o The United States are also suffering from it. In San Francisco specifically, the housing market is absolutely broken. As we said, the average price of a home is 1.7 million dollars, more than triple the national average of $450,000, and the rent is already the most expensive in the entire country, even ahead of New York. The demand is very high and the construction of new buildings is stagnant and the little that there is at a decent price ($5,000 a month is decent) flies in a matter of seconds. Image | photoholgic in Unsplash In Xataka | Sleeping capsules for 800 euros a month: the housing shortage in San Francisco is reaching the extreme

DeepSeek no longer wants to compete only with models. Its new front aims directly at NVIDIA’s business, according to Reuters

by

In just over a year, DeepSeek has stopped sounding like a rarity in the Chinese industry to become one of those names that already appear every time we talk about the global race in artificial intelligence. First we look at it for its models, for its efficiency and for the shock it caused beyond China. Now the question begins to move to another terrain: what happens when a company that competes in software understands that the next advantage may be in the chips that make it possible to execute that AI on a large scale. The jump to hardware. The information that opens this new front comes from Reuters. The agency assureda, citing three people familiar with the matter, that DeepSeek is developing its own artificial intelligence chip, aimed at inference tasks and not training new models. We will see the technical nuance immediately, because it changes the reading of the movement quite a bit. For now, caution is mandatory: DeepSeek has not publicly confirmed the project would be in an early phase and the company did not respond to the agency’s request for comment. The key is in inference. The easiest way to understand this is to think about what happens after training. Once the model is built, every question we ask and every answer we receive requires putting it to work again. It is not an isolated operation, but a routine that is repeated millions of times if the product works. That is why a chip designed for that phase does not aim so much at technical prestige as at something more earthly: making using AI cheaper, faster and less dependent on third parties. The move is best understood if we look at what DeepSeek has depended on so far. The company has used chips from NVIDIA and Huawei to train and run its models, including the base that held R1, trained on NVIDIA H800a chip designed for the Chinese market whose export to China was banned by Washington at the end of 2023. Since then, DeepSeek has increasingly relied on Huawei: In April it launched its V4 model adapted to Ascend and Huawei said its processors were used in part of V4-Flash training. DeepSeek is no longer a footnote: Until not so long ago, the global debate on AI seemed to revolve almost entirely around American companies such as OpenAI, Google, Microsoft, Meta or Anthropic. DeepSeek changed part of that conversation by demonstrating that China could also produce models capable of circulating outside its domestic market and force the industry to look towards Hangzhou. Recall that the company was widely celebrated in China as a national AI champion. The trend is already seen in a good part of the sector. Google has been developing its TPUs for years, Amazon has Inferentia for inference payloads, Microsoft has Maia and Meta works at MTIA. Reuters also cites two recent movements especially close to the case: OpenAI announced its Jalapeño chip with Broadcom in Junealso oriented to inference, and Anthropic was considering designing its own chips. The pattern is quite clear: large AI companies want to rely less on third-party providers and better control the cost, performance and availability of the computing that powers their services. The big obstacle is manufacturing it. Designing a competitive chip is not the same as wanting to have it. Developing an AI accelerator typically requires years, a lot of capital, and a network of design, foundry, and memory partners. For a Chinese company, furthermore, the problem does not end at the technical level: US export controls limit access to the most advanced foreign factories and also to high-bandwidth memory, a key component for this type of chips. Times change. NVIDIA arrived at the AI ​​boom with an advantage built over decades: in 1999 it launched the GeForce 256, presented by the company itself as the industry’s first GPU, and in 2006 launched CUDAthe architecture that helped take the parallel processing of its chips beyond graphics. When the models started requiring massive amounts of compute, I already had the hardware and ecosystem in place. For years, for much of the industry, competing in AI meant going through its chips. What the DeepSeek case suggests, with all caution, is that this dependency is beginning to have cracks. Images | Xataka with Nano Banana In Xataka | Samsung earns 19 times more than a year ago. Investors have reacted by sinking the stock 7%

Selecting all the traffic lights is no longer enough to prove that you are not a robot. Now you have to scan QR codes

by

There are those who think that There are no people on the internet and everything is bots.is what is known as dead internet theory. For years, we have used captcha technology to prove that we are not robots, but visual tests and puzzles are no longer reliable proof because AI can skip them without problem. The solution is to make us scan QR codes, but there is an ulterior motive behind it. Scan to verify you are a person. This is what I found this morning when trying to enter a website. My first instinct was to check the URL and all the details in case I was facing some kind of scam, but no. I took out my phone, opened the Google app and scanned the QR to access it. It is the new reCAPTCHA method to verify that we are not a bot. Although this novelty was introduced last monthI had not come across one of these QR until today. An anti-AI test. As we said, AI has managed to solve the most common captcha tests, such as the classic visual tests that ask us to select elements from an image or write the letters and numbers in an image. The solution that Google, owner of reCAPTCHA, has found is to make us scan a QR code with our mobile phone. The function is called Mobile verification” and They describe it like this: “In addition to visual challenges, it includes AI-resistant controls that use a mobile device to scan a QR code and verify human presence and device integrity.” At the moment this function is in the experimental phase and we can still use the visual challenges by clicking on the eye icon, but everything indicates that it will become the verification standard. The real reason. Being resistant to AI is not the only reason for this new method, there is something else that has consequences for many users. With mobile verification, Google has linked reCAPTCHA to Google Play Services making it only work with devices that have a current version of Google services. In practice, this means that users of Android phones without Google, such as Fairphone, GrapheneOS or Huawei phones, will not be able to verify that they are human. In other words: either you use our services or you will not be able to browse freely. No problems on iOS. As they point out in this Reddit threadGoogle could argue that having Play services updated is a security issue, but at the same time they allow the system to work on iOS where you are not forced to install any software. I have scanned the code myself from an iPhone and the only requirement is that the device has iOS 16.4 or a later version, something that almost all iPhones will comply with without problems. This shows that it is not about security, but about controlling the ecosystem and penalizing those who try to escape from it. The danger of QR. During the pandemic, QR code scanning had a golden age thanks to restaurants eliminating physical menus. This caused the QRishing scams, which basically consists of falsifying the QR codes so that they direct us to a fraudulent website. If Google normalizes QR scanning to browse the web, it’s a matter of time before fake codes posing as reCAPTCHA verification start appearing and scammers have a field day. Image | reddit In Xataka | You thought you were solving a CAPTCHA to enter a website. You were actually training an AI

‘Backrooms’ is going to return to theaters with a 15-minute longer version. If successful, we will have a dangerous precedent

by

‘Backrooms‘ has raised a surprising 330 million in theaters with its first version in theaters. Wait: first? Are there different versions of ‘Backrooms’? Now it is: five weeks after the premiere, its production company A24 is bringing it to theaters again, this time with 15 minutes of exclusive material. Obviously it is not the first time that the industry has made moves of this type, but it is the first in which there is so little time between versions. And it sets a precedent that could give us more than one disappointment in the future if it is successful. The return A24 to relaunch ‘Backrooms’ on July 3 under the title Backrooms: Everything Must Go Edition, just five weeks after its original premiere on May 29. They are 15 minutes of “unreleased content in theaters”, and with special features: they will be included in block after the final credits, they have been directed by the film’s director Kane Parsons and it will be exclusive to this re-release and will not be seen in home editions or streaming. Less and less. The director’s cut or director’s cuts (something that this ‘Everything Must Go Edition’ is not exactly, since the new footage will not modify the existing one, but is extra material after the credits, like a “deleted scenes” extra on a DVD) are not new in the industry. They have existed since 1942, when Chaplin’s ‘The Gold Rush’, from 1925, was re-released with new material. The modern concept was born in the seventies with Peckimpah’s ‘The Wild Bunch’, when a restored version was released in cinemas. The most popular was the Director’s Cut of ‘Blade Runner’, ten years after the release of the original, with changes that modified the meaning of the first film. In all of them and others equally popular (and often released in cinemas), such as ‘Close Encounters of the Third Kind’ or ‘The Gates of Heaven’, there is a coincidental pattern, and that is that the rerun comes years after the original premiere. And often, accompanied by a discussion about the emasculating power of the industry, which prevents the full expression of the creator’s version. He director’s cut It is an advertising tool, yes, but it is also a vindication of the films without interference from the studios. This time, the studio itself releases an extended version when no one has complained that the first cut is mutilated. The danger. It’s clear what’s at stake here. If the move is successful, the next step is to release intentionally “incomplete” versions to generate two box office cycles. First the base audience is reached, and then, weeks later and when the hypethe film is re-released with “exclusive” material. Here specifically, the move is clear: that the production remains profitable in the juicy weekend of July 4, very important at the US box office. The difference with very popular extended versions such as those of Peter Jackson’s ‘The Lord of the Rings’ seen on DVD, is obvious. A24, in question. Although for a time A24 has been the pretty girl of the Holywood production companies for his ease in taking risks, for cultivating undervalued genres and for supporting directors with clear auteur visions (and still get blockbusters and sweep the Oscars), for a few months now his image has been questioned. To the open adoption of one totally pro-AI policy Now this very tricky marketing move is added. In a couple of years we will see if what today sounds like an extravagant maneuver becomes a standard tactic: launch “version 1.0”, wait for it to hit the box office, re-release a “complete vision” after a month and a half. Before the director’s cut It took months to arrive, now even those small islands of creative integrity have been subject to the extremely accelerated calendar set by social networks and the streaming. In Xataka | The most terrifying photo on the internet has no monster, no story and no one knows who took it: Backrooms comes to the cinema

the cities are no longer yellow

by

The astronauts who have been lucky enough to travel to space more than once in the last decade are privileged witnesses of a chromatic change on a planetary scale. From their vantage point 400 kilometers high, they have been able to see that the cities, previously faint spots of amber, now shine with an intense white light. It’s not a metaphor. It is the visible trace of one of the most rapid and widespread infrastructure transformations in recent history: the great replacement of public lighting. We have retired the old sodium vapor streetlights and massively embraced LED. This change, driven by regulation in favor of energy efficiency, has redrawn the night map of the Earth, a phenomenon that can be seen more clearly from space. The invention that earned a Nobel Prize in Physics. Old sodium vapor lamps, especially low-pressure ones, were monochromatic in nature. They emitted light in a very narrow band of the spectrum, resulting in that characteristic and ubiquitous yellow-orange hue that tinted our streets and skies. LED lights work in a completely different way. His breakthrough, which earned Isamu Akasaki, Hiroshi Amano and Shuji Nakamura the 2014 Nobel Prize in Physicswas the invention of the high-efficiency blue LED. By combining this blue LED with a phosphor coating, it was finally possible to generate a bright and affordable white light. This diode is not only more efficient (exceeding 300 lumens per watt, compared to 16 for an incandescent bulb), but it offers a much broader spectrum. Southern Europe in 2025 from the International Space Station. Image: Don Pettit The cities changed color. To the eyes of a night observer in space, cities have gone from being yellow to glowing bluish white. Milan is the paradigmatic case: it completed its transition to LED in 2015, and appears in an ESA comparison with before and after photos taken by astronauts André Kuipers and Samantha Cristoforetti. But it is by no means the only case. Los Angeles was a pioneering city: it ordered the replacement of 140,000 streetlights in 2009. Buenos Aires modernized its lighting with smart LED streetlights between 2013 and 2016. New York finished replacing 500,000 bulbs in 2023. Barcelona plans total remote management of public lighting by 2028. But India is the country that carries out the largest replacement in the world, with more than 13 million LED streetlights already installed. The b side of this transformation. Like any revolution, the LED has a dark side. Light is cheaper, so cities are not only replacing old streetlights, but also increasing the number of light points or their intensity. The result is that we are leaving a brighter planet, where it is most difficult to escape of light pollution. The statistics indicated otherwise, but it must be taken into account that light pollution is measured by satellites, and satellites are partially blind to blue light. This means that the actual increase in light pollution, especially that perceived by human beingsis much higher than official figures indicate. To make matters worse, blue light is the one that interferes the most with our biological clock, and can affect the quality of sleep, in the same way that disorients migratory birds and the moths. The future is adjustable. The solution is not to go back to sodium. The efficiency of the LED is indisputable. The key, as with any technology, is in its application. The next phase of this transition is not about changing light bulbs, but about installing smart streetlights. It is estimated that almost one in four streetlights will be smart by 2030. When connected, they can regulate their intensity depending on the time or traffic, detect faults in real time and collect environmental data. This remote management will allow one of the new lighting maxims to be applied: using only the necessary light, when and where it is needed. In parallel, other solutions have emerged to protect biodiversity, such as red light streetlights being tested in Nordic cities so as not to disturb the bats. and the idea of ​​bioluminescence as an organic way of generating light without any electrical consumption and with minimal environmental impact. Image | The Iberian Peninsula in 2012, by astronaut Don Pettit In Xataka | Why sunlight doesn’t illuminate space: solving the question a child sent to the ISS

the continent has just realized that its infrastructure lives in a world that no longer exists

by

a tugboat approaching a dutch drawbridge and watering it with their hoses; stopped trams in Leipzig, British supermarkets without chilled products, melted roads… They seem like a bunch of curious anecdotes about how Europeans survive one of their first real ‘heat waves’. But they are not. Each of these failures is the symptom of a problem that, despite bombastic speeches and sickeningly detailed plans, we have persisted in forgetting: that climate change is serious. And here we are. A Europe that does not exist. For practical purposes, the event these days is the second heat wave that the continent has faced so far this year. We have seen incredible things: 37.3 in the United Kingdom, 37 in Denmark, 41.7 in Germany, 39.5 in Slovakia, 39.4 in the Netherlands… it is not only that the June highs have fallen in almost all the countries of Western and Central Europe, it is that absolute records have been broken (i.e. also July and August) in four countries. According to World Weather Attributionis the most severe episode ever measured in the region studied: in 1976, such heat would have been “virtually impossible” in June. And that is perhaps the most important lesson of these days: that the Europe of 1976 no longer exists. And we have begun to notice it in the worst possible way. Although we can make distinctions between what has happened these days (In Leipzig, the problem is that the sealant between the lane and the road surface softened to dangerous levels; while in Holland the bridges began to be refreshed by protocol without any problem being detected), the truth is that these are all signs that the European infrastructure is outdated. Many of Europe’s roads, bridges and highways were designed for maximums between 32-35. Before, exceeding that limit was something anecdotal (in the 119 years between 1881 and 2000 There was only one day in Germany that measured 40° or more), today is the ‘new normal’ (last week there were 4 days like this). It is important to note that so far I have not said anything about mortality. It will take some more time to have the complete data, but suffice it to say that France has already registered around 1000 deaths attributable to the heat wave. The obvious question is… what have we been doing all along? While all this is happening, no one can claim ignorance or surprise: in March 2024, the European Union itself recognized in its first European Climate Risk Assessment that “Europe was not prepared” for what was coming, that policies “were not keeping pace with the increase in risks” and that incremental adaptation “was not going to be enough.” I don’t want to say that nothing has been done. There are analyzes that say that without the adaptation of this century (things like heat plans, surveillance or the alert system) mortality would have been a 80% older. However, the data is there: no matter how much we have done, the deficit grows with each passing day. And that means we’re not doing enough. What can we expect? It seems like little or nothing. In recent years, public support for climate policies appears to have tempered. And there is a lot to do: we must not forget that estimates tell us that Europe’s air conditioning fleet will go from less than seven million devices in 1990 to more than one hundred million in 2030. That requires radical changes: an enormous reconversion that, given what we have seen, we do not know if we are going to want to undertake. Europe knows what is coming and knows what it has to do. You have it planned, signed and approved. The question is whether he will do it before this stops being an anecdote and begins to become an unmanageable crisis. Image | Bill Iliot In Xataka | ENT doctors agree: “Sleeping with air conditioning forces the nose to work excessively”

“In three or four years it is possible that there will no longer be anyone developing software here”

by

SAP is one of the largest software companies in the world, so when its CEO states that in a very short time there will be no one programmingthings are going to get really complicated for software engineers. Christian Klein, CEO of SAP, assured in an interview for Financial Review that: “Software development is the function most affected by AI, and there is a possibility that in three or four years there will no longer be anyone developing software at SAP.” Klein’s statement is not a pressure maneuver on his employees, but is part of a strategy that the manager has been building for months, and which now has a name: the Autonomous Company. A company that wants to reinvent itself from within. SAP has more than 110,000 employees worldwide and is the largest software company in Europe. Of those 110,000 employees, more than 30,000 hold software development positions. That means Klein hopes be able to automate work which today employs more than 27% of its workforce with AI agents. The CEO shares the idea that the vibe coding It allows someone without technical training to generate software from natural language instructions. Klein, just like Like many other senior technology managers, he sees this as the beginning of the end for programmers. as we know it currently. That is, the role of the software developer will no longer be linked to the generation of the code, but to your supervision. SAP does not fire, it replaces. However, what Klein wanted to make very clear is that, although SAP will tend to reduce the number of software developers on its staff, that does not mean that it will lay off its employees. They will simply go to develop other roles within the company to meet new development needs. Klein explains that the workforce will change and product managers, who previously barely programmed, will start working alongside experts from different sectors to create new AI agents. “We need product managers who know how to read code and understand business. Although the demand for developers is low, we need more data scientists,” says the manager. Looking for AI trainers, not programmers. According to data from the latest study On global trends on the future of work prepared by ManpowerGroup, AI skills have surpassed engineering and computer science as the most difficult to find, with 72% of companies having difficulty filling these positions. The profiles Klein describes, product managers who understand how AI works and design AI agents, do not yet exist in quantity. what the industry needs. Developer confidence in the code generated by AI fell from 40% to 29% in a single year. 46% of professionals actively distrust these tools. 45% of the code generated by AI contains errors serious safety issues. Klein talks about a future without programmers, but today’s data shows that human supervision is still the difference between a reliable product and one with holes. In Xataka | Richard Liu, CEO of JD.com: “When robots deliver packages, the day will come when delivery drivers will no longer be needed Image | SAP, Unsplash (Becomes Co)

Experts agree that opening windows at night and closing them during the day is no longer the best strategy against heat.

by

For decades, long before air conditioners and the fans will take Every corner of its geography, Spain knew how to handle the heat: during the day, everything was closed tight; at night, everything open and ventilate. It sounds rudimentary, but in 2026 that is still the strategy we recommends the Ministry of Health. And yet, there is a problem. The Spain of the blinds game did not sleep in cities at 26 degrees at four in the morning. Not 26, not 28, not more than 30, as has been happening these days in Almería. The arrival of the ‘hellish nights’ challenges everything we thought we knew about domestic heat management. Therefore, we have asked ourselves… and now what? What we were doing until now. That popular knowledge I was talking about goes viral every time the heat makes an appearance. And, like I say, it makes sense. In a world where it cools down at night, the winning strategy is night cooling and daytime insulation. What he says Health Plan against high temperatures. The central issue here is that that world has ceased to exist. Because the key is not the time, it is the thermometer. In reality, what we are looking for is to open when it is cool and close when it is hot. It’s usually cool at night, but… What if he stops doing it? Summer has stretched on for five weeks and Spain is about two degrees warmer than at the beginning of the eighties. The torrid nights (with minimums of 25 degrees or more) have multiplied by ten since 1984 in the ten most populated capitals and, according to AEMET data, this affects about nine million people. Even the tropical ones (minimums of 20 or more) today add a dozen more a year than decades ago. We can’t sleep… although researchers have not agreed on the ideal temperature for sleeping (some point which is about 18.3ºC, but there is no consensus), they have done so on a fundamental idea: sleeping in the heat is objectively a bad idea. For whatever reason, it is true that our temperature changes between wakefulness and sleep. In fact, “thermal regulation is a significant factor” in sleep control, explained the teacher Cameron Van Den Heuvelfrom the University of Adelaide. “About an hour to thirty minutes before sleep, the body begins to lose body heat. This increases feelings of tiredness in normal healthy adults.” People with insomnia, without going any further, “show that they have a higher basal temperature just before sleeping than people who do not have sleep problems.” Ambient heat does not help this thermal reduction and it seems to be proven that when the temperature is very high, it is more difficult to fall asleep and, when it is achieved, it is of very poor quality. And then what do we do? In the interior of the peninsula, as the thermal amplitude remains high, opening at dawn works. There is nothing to change. The problem is for those who live on the coast or in the big city. When the night exceeds 20 degrees, the strategy is reversed. Ventilating at night is no longer the main technique and the battle is won during the day: you have to seal the house as soon as the street heats up, squeeze out the thermal inertia and use the air conditioning wisely. However, as the nights get warmer the household tricks come to an end. If the trend is confirmed, we will have to assume that the housing stock must be transformed: sun protection, rehabilitation and climate shelters will be the buzzwords in a few years. Image | Fernando Rosado In Xataka | ENT doctors agree: “Sleeping with air conditioning forces the nose to work excessively”

Ukraine no longer wants to send soldiers to the front. His answer is an “unpublished Frankenstein”: robot-tanks with machine guns

by

The British inventor Ernest Swinton first proposed using “land ships” to cross trenches and avoid sending infantry directly into fire. That idea gave rise to first modern tank a year later. Now, more than a century later, the logic is the same again: find a machine that can go where a soldier no longer can. A front that is too deadly. The war in Ukraine has entered a point where the simple act of sending men towards the line of contact begins to be too expensive a luxury. The saturation of drones has turned wide swaths of the front into real death zones where any movement is detected and punished almost instantly. This has changed the logic of the battle because it is no longer just about who has more armor or more artillery, but about who can continue hitting without exposing bodies. And here Ukraine is beginning to embrace a radical idea: if the front devours soldiers, perhaps the time has come to stop sending soldiers. The birth of a new “Frankenstein”. The Ukrainian response is as pragmatic as it is disturbing: take weapons systems designed for fixed positions and assemble them about ground robots. In fact, companies like Frontline Robotics They have converted their turret autonomous Buria (a kind of metal arm capable of firing machine guns or grenade launchers) into something new: small armed unmanned vehicles that act as micro-tanks. They are not tanks in the classic sense, but they fulfill part of their function. They move, shoot, seek cover and attack without a human being inside. It is a kind of improvised hybrid, a creature assembled with different parts to survive a battlefield that no longer forgives. The war against Russian infiltrations. The immediate objective of these robots is not major offensives, but something much more specific: hunt small groups of Russian infiltration. Moscow has been exploiting a simple but effective tactic for months: send small teams of infantry that slip between lines, forests and trenches to avoid aerial surveillance and penetrate Ukrainian defenses. They are small movements, difficult to detect and cheap to execute. Ukraine has understood that responding by sending more infantry only feeds that crusher. So now he sends armed robots, controlled from dozens of kilometers away, to intercept them before they go deeper into the line. The classic tank is getting old. There is a compelling reason why this idea is gaining traction: traditional tanks They are suffering greatly. Both Ukraine and Russia have discovered that a multi-million euro armored vehicle can be destroyed by a cheap drone in a matter of minutes. The economic equation has been broken. In that context, an armed terrestrial robot is a brutally logical solution: it costs less, it is manufactured faster and if it is destroyed no one dies. It is a mutation of the armored concept. Less armor, less power, less glamour… but more expendable. And in an industrial war, the expendable is often more valuable than the perfect. Evolution at the speed of war. The most striking thing is the speed with which all this evolves. Frontline Robotics It ensures that it introduces small changes up to twenty times a month and major updates every six months. The battlefield works like real time laboratory. Soldiers send constant feedback, and companies adapt their machines almost on the fly. It is a brutal advantage compared to the bureaucratic slowness of many Western industries. What today is a robot with a machine gunwithin a few months it may be a much more sophisticated system. Put another way, Ukraine is learning faster because it has no choice. Defend without humans. The underlying idea is more than a simple tactical innovation. Ukraine has already used ground robots to evacuate woundedcarry supplies, place mines and clear routes. This year they have made more than 50,000 missionsa gigantic jump compared to just 2,000 the previous semester. But what is truly new is the offensive leap: Russian positions captured only with aerial drones and ground robots, without a single infantry entering first. There has even been surrenders to machines. This paints a disturbing future: entire sectors defended and attacked by unmanned systems, where humans are increasingly distant from the point of impact. The ground war is mutating. What Ukraine is doing with these robot tanks may seem improvised, even almost artisanal, but it contains a transformation a lot more fat. For centuries, conquering land meant sending men into fire. Now start mean send machines. Not because they are better at everything, but because they are sacrificial. And on a front where every meter costs blood, the priority is no longer to advance bravely, but rather preserve lives. For this reason, Ukraine is not simply building new robots: it is testing a new way of waging ground warfare where the soldier is no longer the first piece to move. Image | Frontline Robotics, Oleksandr Klymenko/Ukrinform In Xataka | Satellite images revealed that Russia covered a building with an anti-drone cage. Ukraine turned it into an action movie set In Xataka | The drone war in Ukraine just solved an old military obsession: how to blow up a bridge without tons of bombs

The world’s leading expert in identifying deepfakes has a big problem. No longer able to identify them

by

Is called Hany Farid and is considered one of the world’s leading experts on deepfake videos. This digital forensics expert was capable of uncovering videos manipulated by governments, for example, but now he has decided to leave Silicon Valley for one simple reason: It is no longer able to differentiate the real ones from those that are being generated with AI tools. And we are not surprised. Deepfakes indistinguishable from reality. In the last two decades Farid, 60, has specialized in identifying fake videos. This professor at the University of California at Berkeley has confessed that advances in generative AI have made traditional detection methods are no longer of any use. Their conclusions confirm the feeling that we have had for a long time with this type of content: AI has advanced so much that the problem is no longer just deepfakes: it is that we distrust even real photos. Farid’s reputation precedes him. His father worked for 50 years as a chemist at Eastman Kodak, and Farid grew up visiting the dark room often, watching photos become photos as they passed through the different liquids. He ended up designing a “digital fingerprint” system that made it possible to detect cases of child pornography hidden on the Internet. In fact, its technology has led to 30 million cases of potential abuse being reported each year, as well as leading to hundreds of arrests and several rescues. I surrender. Faced with the avalanche of perfect deepfakes generated with AI, Farid has decided to leave his job to take refuge on a farm in Vermont. His surrender is the latest demonstration of a harsh reality: We can no longer trust what we see on networks. Now he is dedicated to working with wood, and has distanced himself from networks and technology. The missile that changed everything. The turning point that demonstrates this crisis of this digital forensic task occurred after the viral spread of a video showing the alleged impact of a US missile on a school in Iran. Farid spent an entire day breaking down the sequence frame by frame: analyzing the geometry of the shadows, the sound delay rate of the explosion according to the laws of physics, or the pixel length of the projectile. Impossible to decide if it is false or not. He found nothing that could prove that the video was fake, and the same thing happened to other specialists. None could issue a clear verdict of authenticity, and that made it clear that AI video generation is currently so advanced that real content is indistinguishable from a deepfake generated with these latest generation models. Verifying is too complicated. There is another problem here: generating a fake video, whether toxic or not, with cloned voices that are perfectly synchronized with the interlocutor is easy, fast and cheap. Carrying out a forensic investigation to try to detect whether the video is real or not takes hours of computational and direct analysis by specialists. Given that deepfakes manage to go viral in just 20 minutes if they are successful, the methods to contain this spread are useless for a simple reason: they arrive late. The biter bit. The researcher himself was a victim of this reality: cybercriminals cloned his phone number and used AI to generate his voice and thus impersonate his identity. With that clone, they called a close contact who was involved in a court case and managed to extract confidential information. Farid and his wife, a vision researcher at Berkeley, they had to create a secret safe word at the beginning of each family call to certify that each interlocutor was who they said they were. The situation generates a disturbing paranoia and mistrust. “I’m going blind”. In the report of The New York TimesFarid explained that his studies show that most people can no longer differentiate a real photo from a digitally created one. “I feel like I’m going blind,” he indicated, showing his concern about an AI that is managing to obscure the truth and distort reality. Watermarks as a solution. Faced with this avalanche of images and videos generated by AI that are indistinguishable from reality, one of the potential ways to mitigate the problem continues to gain strength. It is, of course, the watermarkstotally invisible and which are part of the metadata of those files. Two promising initiatives. There are several initiatives in this regard, although the most notable It is that of the C2PA coalition which includes, for example, Google and OpenAI. AI tools should add those watermarks identifying those contents (“This video has been generated with this AI application, this image has been generated or edited with this other one”), but at the moment that type of option is not applied by default. Another important project in this sense is SynthIDGoogle’s technology to “mark” these contents as created with AI. Image | Bild (CC0) In Xataka | What happened to Technicolor: evolution and death of the company that changed cinema and was overwhelmed by its ambition

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.