wrong

Ozempic came to simplify the relationship with food. Christmas is proving how wrong we were

by

Christmas has always been a delicate territory for the relationship with food. Family reunions, full tables and seemingly harmless comments turn these dates into a kind of silent examination of the body. For years the answer was the express diet before the holidays. Today, in many cases, the conversation revolves around a weekly injection. In the United States, there were cases of people skipping their Ozempic shot to enjoy Thanksgiving, according to The Wall Street Journal. The phenomenon in Spain is less visible, but the question is inevitable: what will happen this Christmas and what consequences could this jump have? Ozempic in Spain: a discreet use on the rise. Ozempic or other drugs based on semaglutide or tirzepatide have restricted access, require a prescription and, in theory, priority for patients with diabetes. Even so, its use for weight loss exists and is increasing, especially in the private sector. “In consultation, doubts begin to appear about what to do with these drugs on special dates such as Christmas,” explains Dr. Víctor Bravoendocrinologist interviewed in Xataka. “It is not always verbalized as ‘I’m going to skip the dose’, but the idea of ​​’stopping a little’, ‘adjusting’ or ‘I’ll see after the holidays’ does appear.” The difference with the United States is one of degree, not nature. There the debate It is public and massive. Here it begins more quietly, but with the same roots: fear of losing control, social pressure and a complex relationship with food that the holidays intensify. Understand well what it is. Ozempic is neither a one-time appetite suppressant nor a cosmetic aid. Its active ingredient is semaglutide, a drug that replicates the action of the hormone GLP-1 that the body releases after meals. This hormone intervenes in the regulation of insulin and satiety signals, so the treatment reduces appetite and prolongs the feeling of being full. “This is important to understand well,” emphasizes Laura Albó, psychologist specialized in eating disorders and EMDR traumawith whom we have chatted in Xataka. “It is not a pill that takes away the desire to eat only that day. It is a treatment that modifies the signaling of hunger and satiety on a continuous basis. Interrupting it is not neutral.” Besides, as recalled by the scientific reviews recently analyzed by the WHOthese drugs work to lose weight, but they are not free of side effects nor is their long-term impact still known with certainty. Nausea, digestive discomfort and, in some cases, abandonment of treatment are part of the real picture. The Christmas dilemma: enjoy without losing control. One of the most repeated promises of GLP-1 is peace of mind: eating without fear of overflowing. Precisely for this reason, Christmas is experienced as a paradox. “What we observed is that many people consider these dates as an exception,” explains Albó. “It’s the same mental scheme of diets as always: now I can relax, now it’s time to enjoy, and then I’ll control myself again. The tool changes, but not the logic.” According to the psychologist, the conflict is not so much in the amount of food as in the meaning attributed to it. “When someone feels like they need to skip treatment to enjoy themselves, they are reinforcing the idea that eating with pleasure and eating with control are incompatible. And that dichotomy is a clear basis for eating discomfort.” From a medical point of view, Dr. Bravo agrees that expectations are often unrealistic. “Some people hope that by stopping the medication for a few days the body will function as before the treatment. But what usually returns is not a ‘normal’ relationship with food, but a sharp increase in hunger and constant worry about eating.” As we have previously mentioned, in the United States, some patients delayed the weekly dose to arrive with a greater appetite for holiday meals. But Laura Albó warns that this approach displaces the problem: “It’s not just physical hunger. It’s mental noise, paying attention to the menu, the dessert, how much is left. Just what many people had managed to silence.” What happens if you skip Ozempic? From a physiological point of view, interrupting or delaying a dose can cause a return of hunger that is more intense than expected. “The body once again receives signals that had been dampened for weeks or months,” says the endocrinologist. “This does not mean that the person eats ‘normal’, but rather that they may experience a sharp increase in appetite and greater difficulty in self-regulation.” But the impact is not just physical. “On an emotional level, the effect is usually a swing,” adds Laura Albó. “First the idea of ​​permission appears—now I can eat—and then, if the person feels that it has gone too far, guilt and shame come. This cycle is well known in consultations.” Scientific evidence supports this risk. Studies on hormonal regulation of appetite show that the body’s adaptations after weight loss do not disappear immediately. Skipping treatment does not eliminate that vulnerability; in some cases it intensifies it. So what do the experts recommend? There is no single answer, but there is clear consensus among the professionals consulted: Do not make impulsive or guilt-based decisions. Do not use medication as punishment or as permission to eat. Maintain basic schedules and routines to avoid arriving extremely hungry. Understand that two or three meals do not determine the success or failure of a treatment. “The important thing is not to turn Christmas into a test,” summarizes psychologist Laura Albó. “Two meals do not change a body, but they can greatly alter the emotional state.” For its part, Dr. Victor Bravo He insists that any adjustment should be discussed with the professional who prescribes the treatment: “Not so much because of the specific meal of a day, but because of what that decision can trigger later.” The role of the family: the noise that cannot be seen. Although the focus is usually on who takes the drug, the environment has a decisive weight. Comments such as “how little you … Read more

We thought we “discovered” fire 50,000 years ago. We didn’t know how wrong we were.

by

For decades paleontology has maintained a clear distinction in history: it is one thing to use fire and quite another to create it at will. Something that seems very silly, but is essential since until now the evidence we had on the table pointed to the ability to light a bonfire from scratch They dated back 50,000 years. But this has changed. A big change. A published study in Nature He told us that we were quite wrong about this. A team of researchers has pointed out that hominids already possessed technology to make fire voluntarily 415,000 years ago. That is, 375,000 years earlier than we thought. Although what is surprising is that it was not even our species, but the early Neanderthals. Something that has been known after studying a site found in Barnham in England that has given the necessary evidence to reach the end of the matter. How do we know? At the moment we do not have a time machine to travel to the past and see what happened in our history. That is why this discovery makes it surprising that they used reverse engineering to reach this conclusion. The elements that were available at the site were not just ashes, but the “ignition kit.” Researchers were able to identify fragments of pyrite and flint axes, which can be used to make fire. Although the key here is that the pyrite It is not native to that area, but hominids had to intentionally transport it to make fire voluntarily. The mechanism is, in essence, the prehistoric version of a modern lighter: striking the pyrite with the flint generates sparks capable of igniting dry tinder. Confirming it. With these indications, anyone could think that it could be a random fire, and that is why advanced techniques such as archeomagnetism, micromorphology and spectroscopy were used. In this case, the results indicated that the sediments had been heated to more than 700 ºC, which suggests that it was a concentrated and fed fire. This is also added to the fact that the flint axes presented specific cracks caused by cycles of heat and cooling, indicating that fires were made repeatedly. A big jump. The importance of this discovery is monumental since until now we assumed that complete control of fire was a late skill. This discovery sets the controlled ignition clock back by 375,000 years compared to previous evidence from French sites. This tells us that the minds of early Neanderthals, who were most likely found in that area, were more developed than thought. In this way, transporting pyrite implies long-term planning, which is not an instinctive reaction to the cold, evidencing a cognitive ability to think about the future. The domain of fire. Making fire at will is considered a great evolutionary advance since fire can lengthen the day for nighttime socialization or even cook food to obtain more energy with less digestive effort. This also represents a great geographical expansion for the species, since 400,000 years ago Europe was going through a very important glacial period, which made the heat of fire essential for the species to perpetuate itself. Images | Mladen Borisov In Xataka | Neither lions nor hyenas: at the top of the food chain 30 million years ago, there was a “pig” weighing more than a thousand kilos

Microsoft knows that something has gone wrong, and promises these changes

by

For years, if you wanted to play on a computer, the answer was almost automatic: you needed Windows. Linux experience was limited and macOS did not offer a competitive catalog. That landscape seemed immovable until Valve decided to really bet on the game on Linux and showed that there was room to shake the board. Steam Deck It came as an experiment that many did not see coming and ended up reconfiguring expectations, to the point that more and more players are talking naturally about switching to Linux. This change of mood has put Windows under a magnifying glass that it did not have before. Windows’ historical strength in gaming was explained by something very simple: it offered the largest catalog, the most mature tools and a fluid relationship with developers. That basis is still real, but its perception has changed. The end of Windows 10 support along with the strict requirements of Windows 11has put teams that were still performing well on the ropes, unless their owners agree to run out of patches or use the extended update program. At the same time, the integration of functions that many interpret as unnecessary additions has generated some wear and tear. Microsoft tries to retain its throne in PC gaming Valve has been preparing the ground for years so that gaming on Linux stops being an experiment and becomes a viable option. Proton has allowed thousands of games designed for Windows work on SteamOS with a level of compatibility that was previously unthinkable, and the Steam Deck has served as a showcase for that progress. The recent announcement of a new Steam Machine for the show consolidates that movement, placing Valve in a position that challenges the idea that Windows is the only natural destination for PC gaming. It is not a frontal assault, but it is increasing strategic pressure. In parallel, far from presenting a laptop with an Xbox seal, Microsoft has opted for a more flexible path: supporting manufacturers that already dominate this segment. Together with Asus and AMD, he has shaped the ROG Xbox Ally and Xbox Ally Xdevices that run Windows 11 but boot into a full-screen interface designed for controller use. This experience unifies games from different launchers and reduces distractions, bringing the console feel closer without giving up the PC nature. It’s a way to compete in portable devices without having to design and maintain new hardware of your own. Besides, Microsoft presented Several internal improvements in Windows 11 are the result of work with the ROG Xbox Ally, which today benefit a large part of the Windows PC ecosystem. They include more efficient power settings, more stable memory management on Ryzen APUs and lower CPU load on tasks that previously affected performance. Still, the company insists that there is still room to cover. “We are committed to making Windows the best place to play, and we will continue to refine the system behaviors that matter most in gaming: background load management, power and scheduling improvements, graphics stack optimizations, and updated drivers.” Several of the technical improvements announced by Microsoft have already reached the desktop. DirectX Raytracing 1.2 is available and provides tools to process complex scenes more efficiently as long as the GPU and drivers are compatible. Advanced Shader Delivery works on select titles and speeds up initial loading when precompiling shaders during installation. Work on neural rendering is advancing cautiously and is only available in preliminary mode for studies. In parallel, Windows 11 has expanded support for LE Audio, which reduces latency and improves the experience in games that depend on sound. The push for Windows on ARM has become another relevant front to expand the reach of the ecosystem. During 2025, devices enrolled in the Insider program have been able to install compatible games from the Xbox PC app, allowing many titles to be played locally. The Prism emulator has added support for AVX and AVX2 instructions, and several anti cheat vendors, such as Easy Anti Cheat and BattlEye, have added specific support for Windows on ARM. From a gamer’s perspective, Windows retains obvious advantages, such as its catalog and the guarantee that almost everything will work without additional tweaks. Even so, the experience in Linux has improved Enough so that some see a more limited system as attractive, with fewer background processes and more predictable behavior. SteamOS solves many historical obstacles, although its popularity does not reach that of Windows, which continues to concentrate around 95% of Steam users compared to Linux still close to 3%. Windows’ journey in gaming has been long and dominant, but its role is no longer automatically sustained. Microsoft’s recent decisions show that the company is aware of this and wants to correct the wear and tear with technical improvements, a clearer roadmap to the future. Even so, Valve’s push has changed expectations and introduced a competitor that did not exist before. What remains to be resolved is whether these movements will be enough for Windows to retain the preferred place that no one discussed for years. Images | Microsoft | Xataka In Xataka | We knew that Valve was betting on Linux, but it was hiding something bigger: a years-long plan to bring Steam to all devices

We all turn on our emergency lights when we get into a traffic jam. The DGT knows that we are doing it wrong

by

It is more than likely that when you got your license They won’t mention it, but get into a traffic jam It is easy to turn on the emergency lights while braking. It is something almost instinctive, a warning for the one that goes 120 km/h behind you realize that you don’t brake for no reason. And if you don’t, you’ve probably seen it. However, the curious thing is that the General Driving Regulations do not contemplate this action. Because we do it to avoid accidents, but with the law in hand, the use of warnings It’s not what we have to do when we get into a traffic jam.. And yet, there are even new cars that activate them automatically if the system notices that we brake repeatedly. The most curious thing is that it is not bad nor is it a fault. Simply put, the law was written for cars from another era. Although current cars have been eliminating more and more buttons, relegating them to the screensthe emergency lights is one that has remained a physical and tactile piece. It is logical because it is a security element and it is one that we have well located in the control panel. When braking in traffic jams, it is almost a reflex for many drivers to use the emergency light button to warn those behind them of the situation. In fact, some new cars activate them automatically if the sensors (the accelerometer or the brake pressure sensor) detect a sudden deceleration or if the ABS comes into play. If the braking is progressive, they are not activated automatically. The use of emergency lights in a traffic jam: yes, but no (and vice versa) But… what does the law say? As our colleagues remember Motorpassionhe section C of article 109 The General Traffic Regulations of 2003 establish that the correct way to notify those behind us about this situation is: “The intention to immobilize the vehicle or to brake its progress considerably, even when such events are imposed by traffic circumstances, must be warned, whenever possible, by repeatedly using the brake lights or by moving the arm alternately up and down with short and quick movements.” The problem is that theory is one thing, but in practice, if we are slamming on the brakes It is difficult to walk by lifting your foot off the brake.. Much less by lowering the window and warning with signs. It is much easier to turn on the emergency lights, and the person in the back will also see them better than if we put our arm out the window. Why does the law say this? Because it is an article written in another era. It is an anachronism resulting from times in which the ABS It was not so present and in which, to avoid the wheels locking and the car skidding, we did have to lift our foot off the brake. In this way, we were automatically alerting the person behind us. Therefore, the law does not say that we put on the emergency lights in a traffic jam, but they are not going to fine us for it because the DGT understands the good intention when it comes to notifying other drivers about an anomaly in traffic. In fact, the fact that the law does not establish it, but the cars do, speaks about the discrepancy between the “strict law” and reality. The DGT itself advertises it: In fact, here comes the technicality of “whenever possible”a legal hole that protects us when turning on the emergency lights. Now, where it is mandatory to give these lights is when we cannot travel at the minimum speed on the road. That is, if we are in a traffic jam on a highway and we do not reach half the speed of the road, we will have to turn on our lights. Section 3 of article 49 says: “When a vehicle cannot reach the minimum required speed and there is a danger of overtaking, direction indicator lights with an emergency signal must be used while driving.” Will the regulations be modified at some point to reflect the current situation in which all cars launched these last 21 years Do they have ABS? It is not known, but since it is a universal code to alert of the situation, I imagine that it will not be one of the Administration’s priorities. Of course, you have probably found someone who has used them excessively, giving you a scare for no reason when you turn them on in a non-critical situation. And that, precisely, is what happened with some models from the 2000s that turned on the emergency lights automatically, even when braking to exit the highway. For example, early models of Citroen C4 either Peugeot 307 who were ahead of the rest with something that wasn’t going entirely well. Images | Kathy, Prithivi Rajan In Xataka | The V-16 beacons are here to stay (whether we like it or not): this is all there is to do in case of a breakdown

There are those who think that the housing crisis can be solved by building. At the Polytechnic University of Catalonia they believe they are wrong

by

Spain has a problem with housing. That is an (almost) objective fact. The CIS says so, which places it as the great concern of the Spanish, but a quick review of the newspaper archive arrives to confirm it. During the last months few topics have generated more political debate or have taken out so many people on the street such as difficulties in accessing a home. What is no longer so clear is how to solve this “crisis” residential area recognized by the Government itself. Should we build more houses? Does Spain suffer from a housing deficit? Do we need more land to build? Usually the answer to those three questions is a strong ‘yes’. Now a new study signed by two professors of the Polytechnic University of Catalonia (UPC) and published in a magazine linked to the Ministry of Housing points out that perhaps we were wrong. What has happened? That two professors from the Higher Technical School of Architecture of Barcelona (ETSAB), Blanca Arellano-Ramos and Josep Roca-Cladera, have published a study about the problems that Spain is facing in terms of housing. The report in question is titled ‘Five theses about housing policy in Spain’ and is included in a monograph of CyTETa magazine published by the Ministry of Housing. So far nothing exceptional. The curious thing is that the text questions many of the ideas rooted in the real estate sector, such as that our country suffers from a housing deficit or needs more land to build. While the Bank of Spain (BE) estimates 700,000 homes the mismatch between supply and demand, the study questions whether there really is a ‘hole’ in the market or that prices will go down if we build more. Is there a housing deficit? As already indicated in its title, the article is structured around five theses. And the first addresses precisely that point: Does Spain suffer from a housing deficit? The question is interesting because it is one of the most deeply rooted ideas in the sector. The Bank of Spain itself has calculated that it would be necessary 700,000 houses to meet residential demand. For Arellano-Ramos and Roca-Cladera the reality is quite different. In his opinion, one cannot talk about a deficit without first taking into account the excess of housing accumulated between 2011 and 2021 and the stock of vacant properties. The researchers remember that between 2011 and 2021 the housing stock exceeded the growth in the number of homes by 959,554 units, generating a considerable pocket. In fact, they assure that in 2021 the “accumulated excess” was close to 8.1 million properties, a “‘cushion’ more than enough to absorb temporary housing deficits such as the one produced during the 2021-2024 period,” recalls the UPC in the statement in which he reports the study. What does that mean? That for researchers it is not so obvious that Spain suffers from a shortage of new housing. In their analysis they also remember that a good part of the excess of houses and apartments corresponds to second homes and empty homes. The INE itself estimates that at least in 2021 there were 3.84 million of uninhabited properties, 14.4% of the real estate stock. That percentage far exceeds what most experts consider “desirable” (5%), but at least in the statement The UPC does not address another fundamental aspect: the distribution of these wasted properties, if they are located in stressed markets, such as Madrid, Barcelona or Malaga, or in centers where demand is minimal or even non-existent, in the case of emptied Spain. What if we build more? That is the second question the researchers address. What if we build more homes? Would prices be reduced? Their response is once again skeptical to say the least: increasing buildings will not lead to greater social equity nor will it serve to soften prices. “On the contrary”, slide the UPC note. “According to the authors of the study, the solution is not to build more new homes so that the laws of the market balance prices. In addition to having serious environmental effects, what favors is the real estate bubble like the one that occurred around 2000.” What happens in other neighboring countries? Among other arguments, Arellano-Ramos and Roca-Cladera recall that the rise in prices is not a problem exclusive to the Spanish market, but rather something widespread on the continent. So the question is obvious: if the increase in prices is due to the imbalance between supply and demand, do the majority of EU countries share that same problem? “Is there simultaneously a restriction of supply in relation to demand occurring throughout Europe in relation to demand that explains the increase in residential prices? It does not seem that this is plausible. Therefore it is not reasonable, prima facieturn to the scarce construction of new housing as the main cause of the price of housing”, they reflect the authors before remembering that Spain has invested a higher percentage of GDP in construction than the European average. Do we need more land? The researchers also question whether in Spain the problem of lack of accessibility to housing can be explained by the scarcity of land. And to prove it, they go to the newspaper archive: between the late 90s and the early 2000s, buildable land was made available in the country, which allowed for “massive construction” of residential housing. This boom was not accompanied, however, by a reduction in the price of the square meter. Quite the opposite: residential prices increased, as in other parts of Europe. If Spain saw housing prices rise between 1996 and 2008, it was not because there was no land on which to build or build new homes. “Spain became more urbanized than ever and the result did not represent a reduction in prices, on the contrary,” underlines the UPC in your statementwhich recalls that between 2000 and 2012 Spain was the European country with the greatest “consumption” of land: more than 2,400 square kilometers (km2), almost as … Read more

We believed that the NVIDIA-killer would be some other chip manufacturer. We were very wrong

by

Yesterday NVIDIA had a stumble in the stock market. The shares lost 7% and then recovered part of the fall. Meanwhile, Google grew by about 4%. Both movements had the same origin: the rumor that Meta is considering using Google’s TPU chips in its data centers in 2027. Why is it important. During the last few years NVIDIA has managed to dominate imperially the AI ​​chip segment. Its accelerator GPUs made the difference, but although other traditional manufacturers such as AMD tried to follow in its wake, the dominance of the company led by Jensen Huang was spectacular. That could change, and the surprise is that the one who threatens that position is Google. Google prefers to throw balls out. A Google spokesperson explained on CNBC that “Google Cloud is experiencing accelerated demand for both our custom TPUs and NVIDIA GPUs; we are committed to supporting both, as we have for years.” But they have been preparing the move for a decade. Sundar Pichai’s company has been working on the development of the Tensor Processing Unit since 2015. They launched the first version in 2018 to take advantage of it in its cloud computing business, but little by little these TPUs have been gaining performance and are now promising alternatives for AI loads, both for training and especially inference, as Ironwood demonstrates. Anthropic already uses them, Meta could do it. Google has already reached a circular financing agreement with Anthropicto which it will supply its TPUs for data centers that work with its model, Claude. The rumors pointed out by The Information make it feasible that Meta reach a similar agreement with Google and use those chips in its data centers. The difference, of course, is the size of Meta versus Anthropic. NVIDIA shows off its chest. In a post on It is a message with two faces: on the one hand, congratulations. On the other hand, the declaration of intent. But you already know what’s coming. The CEO of NVIDIA, Jensen Huang, already warned at the investor conference when presenting results of the current situation: the rivalry with Google’s TPUs is increasing. However, he also insisted that Google remains his client and Gemini – which has just been renewed with a spectacular Gemini 3— can run on NVIDIA technology. Competition is good. All major technology companies try to avoid dependence on NVIDIA, and almost all of them have their own bets. It’s AMDbut also Intel, Microsoft, amazon and of course the aforementioned Google. But apart from them there are proposals such as those from OpenAI, Broadcom or TSMC that with their XPUs they want to end the reign of NVIDIA. But CUDA is still a lot of CUDA. The development of own chips is promising, but as AMD knows wellNVIDIA continues to have a spectacular wild card with CUDA, the industry standard development platform for AI solutions. The network effect that this technology has generated it’s going to be hard to beatbut Google certainly has resources to try. Image | World Economic Forum | Hilel Steinberg In Xataka | That Qualcomm prepares its own AI chips is good news. Whether it has an opportunity in the market is a very different thing.

We believed that the pyramids of Giza did not hide any more secrets. we believed wrong

by

Talk about the pyramids of egypt is to talk about the Great Pyramid of Giza. The one of Cheops is the most colossalthe best preserved, the oldest of the seven wonders of the ancient world and the only one still standing. However, it is accompanied by two other vestiges of the past: that of Mycerinos and that of Khafre. That of Mycerinos, or Menkaura, is the smallest of the three, and for years we thought it had only one entrance. We couldn’t be more wrong. The hypothesis. It seems incredible that three of the most studied monuments of human history continue to keep secrets, but what the ancient Egyptians did with these three pyramids was colossal. They follow us surprising so much on the outside as insideand from time to time, as exploration technologies advance, we discover something new. However, although the large one attracts all attention, an archaeologist had his sights set on the Mycerinos. The reason is that, on the north side, the stones that represent the entrance are perfectly aligned and appear more polished than those of the rest of the structure. However, the researcher Stijn van den Hoven He noticed that there was another set of exceptionally polished granite blocks on another face of the pyramid. And, in 2019, he hypothesized an additional entrance to the structure. Non-destructive exploration. Since science is not done alone and must be done, an international team of archaeologists from Cairo University and the Technical University of Munich they got to work to investigate Stijn’s idea. This group is part of the project ScanPyramidswho analyze the pyramids with non-destructive measurement techniques. To do this, they use techniques that are also being used in other places in the world (for example, to discover the secrets of the Mayans). Thus, the ScanPyramids team, analyzed the pyramid of Mycerinos with the latest technology non-invasive analysis: ERTor Electrical Resistivity Tomography. This detects subsurface resistivity variations that reveal cavities or materials of different density thanks to electrodes that penetrate the structures and are computed using three-dimensional inversion algorithms. GPR or georadars. Through electromagnetic waves that penetrate materials and are reflected, anomalies can be found. U.S.T. or ultrasounds. They measure the reflections of sound waves to examine the interior of structures without causing damage. Excavations have revealed these much more polished stones Air-filled anomalies. It is not the first time that this combination has been used at the site, since in 2023 a hidden corridor in the Great Pyramid of Cheops was confirmed using these techniques. And in Micerinos it has also worked. Specifically, two air-filled cavities located directly behind the area of ​​polished granite that van den Hoven observed. The Anomaly 1 It is a cavity located 1.4 meters deep from the eastern surface and would have dimensions of 1.5 meters wide by one meter high. The Anomaly 2 It starts at 1.13 meters deep and measures about 0.9 meters wide by 0.7 meters high. From the Technical University of Munich, the archaeologist Christian Grosse has stated that these techniques allow developing “very precise conclusions about the nature of the interior of the pyramid,” affirming that “the hypothesis of another entry is very plausible” and commenting that these results go in the direction of confirming said entry. The alleged cameras Next steps. Independent researchers who were not involved in the study have stated that the entrances to the pyramids from this era are located on the north face, but that further exploration will determine whether these voids are an anomaly, part of a second entrance as the ScanPyramids team suspects, or “something else.” Now, although these techniques allow us to recognize the interior very precisely, they cannot determine the extent of the cavities due to limitations in the penetration capacity of the tools. They have suggested that the next steps should go in the direction of using techniques such as infrared thermography or muon tomography with cosmic rays in order to have more precise information. Implications. In the end, all these works are supervised by the Supreme Council of Antiquities of Egypt and by the Ministry of Tourism and Antiquities, very interested in seeing them carried out. these discoveries (due to tourism issues) while complying with the standards of care for heritage research. Grosse comments that technology allows us to make discoveries that were unthinkable not so long ago and, when the analysis is completed and it is determined what those air-filled cavities are inside Mycerinos, it could transform the understanding we had until now of the architectural design of this 4,000-year-old monument. Images | ScienceDirect, TUM In Xataka | Of the seven wonders of the ancient world, there is one that we don’t know where it is. That makes her the most fascinating of all.

We believed that no open model could outperform GPT-5. A Chinese startup proves us wrong

by

A Chinese startup called Moonshot just launched Kimi K2 Thinkinga gigantic open model with a trillion parameters that has done something that seemed almost impossible: surpass the best proprietary models from companies like OpenAI, Google or Anthropic. If we thought that “Open Source” models could never compete with GPT-5, Gemini 2.5 Pro or Claude, we were wrong. what has happened. This “AI laboratory” had already announced Kimi K2 in July with that gigantic size of one trillion parameters, but now they have released the “Thinking” version with that same size (32 billion active parameters, Mixture of Experts architecture). According to those responsible, the model is capable of maintaining stable use of agentic tools over between 200 and 300 sequential calls. Or what is the same: it can chain long sequences of actions autonomously and apparently without error. The best of all is not that: it is that it surpasses GPT-5 or Claude Sonnet 4.5 in various tests and costs much less than those models. The benchmarks. Those responsible for Moonshot explained how Kimi K2 Thinking achieves the highest scores in Humanity’s Last Exam (general knowledge, 44.9%) and BrowserComp (agent browsers, 60.2%). He is almost at Claude’s level in the SWE software development test, and is also almost the best in another of those benchmarks, LiveCodeBench v6. It is true that in some tests still slightly behind of its “western” rivals, but the achievement is spectacular. More benchmarks. Those responsible for Artificial Analysis have shown their first conclusions after evaluating it with various tests. Thus, they highlight its behavior in agentic tasks that simulate that the model is acting as a customer service agent. In this test it obtained 93% of the maximum, surpassing all its competitors by far (GPT-5 Codex High obtained 87%, for example). They will do more tests, but for now the prospects are fantastic. And on top of that, cheap. On CNBC indicate that training the model cost $4.6 million, a ridiculous figure considering that training proprietary models like GPT-5 It cost about 500 million dollars according to estimates. Using the Kimi K2 Thinking API is also very affordable: $0.6 per million tokens in and $2.5 per million tokens out. GPT-5 Chat costs $1.25/10 respectively, while Claude Sonnet 4.5 costs $3/15 respectively. The details. The model makes use of an INT4 quantization to improve its efficiency without compromising the precision and quality of its responses. Its context window—the “size” of the data we can enter when making prompts—is 256k, a relatively modest figure for large models but still notable. And as a good open model, we can download it to use locally… if we have a real monster at our disposal. The model weighs 594 GB, and for example joining two Mac Studio M3 Ultra It is possible to make it work locally relatively smoothly at about 15 t/s. Alibaba is behindyes. Although the model is developed by an independent startup called Moonshot, this firm has been financially supported by Alibaba, which is becoming an absolute powerhouse in this field. Already not only conforms with developing its own models, which are outstanding (Qwen is the clear example), but is also financing the development of other models such as Kimi K2/Thinking. China and its love for open AI models. During the last few months we have seen how China dominated in the field of open AI models —not “Open Source”—. The Asian giant has adopted an overwhelming philosophy with increasingly better models but which until now seemed to be several steps behind the large proprietary models of OpenAI, Anthropic or Google. This is no longer the case. The race is lively. This achievement represents a new vote of confidence for the open models coming from Chinese companies. It is true that they are huge and that makes it very difficult to use them in practice by end users, but they present an interesting alternative for companies. Image | idnaklss with Midjourney In Xataka | There are many “internal” races within the greater AI race. And Alibaba is winning Open Source

When Facebook launched its own Tinder we didn’t think it could succeed. we were wrong

by

It was 2018 when Facebook announced Facebook Datingalthough it was not until 2020 when arrived in Spain. At that time, dating apps like Tinder had experienced a boom caused by the pandemic, but Facebook had been losing users for some time and the idea had already been established that it was a place for older people. Meta recently shared usage data for its dating service and they shut us up. 21.5 million. It is the number of daily active Facebook Dating users in the 52 countries in which the app is available. They count in NYTimeswhich surpasses Hinge in users, a very popular dating app especially in the United States that has 15 million users. It is the first time that Facebook has shared usage data for its dating service since its launch. Popular among young people. A Pew Center study published by TechCrunchconfirmed the exodus of young Facebook users, which went from 71% in 2014 to only 32%. The most surprising thing about Facebook Dating is that it is having success among the youngest people. According to data from Sensor Tower As of last year, Facebook Dating has at least 1.77 million users between the ages of 18 and 29 in the United States, which represented a growth of 24%. Free. Other apps such as Tinder, Bumble or Hinge have adopted subscription models through which users can enjoy advantages such as knowing who liked you before anyone else. This is free on Facebook Dating and is your main asset against your competitors. Tinder is the app that had the most paying users, but for years has been losing subscribers. They don’t need it. That Facebook does not charge us for using its dating service can be interpreted as a generous gesture, but the reality is that Meta’s income is astronomical. In the last quarter they entered 51,240 million, many of them thanks to the advertising they serve in their apps. Image | Gemini/Goal In Xataka | Meta does not have the most advanced AI of all, but it does have something much more important: a business plan

Everyone is developing chips that compete with NVIDIA’s. They are in the wrong race

by

Qualcomm advertisement on Monday that it is working on AI accelerator chips, which means there will be new competition for NVIDIA. The company that dominates the AI ​​hardware landscape is seeing a large group of competitors try to erode that position, but the problem for all of these companies is not the chips, but something else. A CUDA call. what has happened. Qualcomm has announced the AI200 chip, which will begin selling in 2026, and the AI250, which will do so in 2027. Both will be able to work in rack-type systems that have liquid cooling. Qualcomm servers may have up to 72 chips based on the Hexagon NPUs of the company’s Snapdragon SoCs. Inference yes, training no. The company has revealed that its chips focus on inference (the execution of AI models) and not training. Their rack-based systems will have lower operating costs than cloud system providers, Qualcomm says. Each rack consumes 160 kW, a figure comparable to the consumption of some racks based on NVIDIA GPUs. There are no details about the price of these chips, the cards or the racks that will integrate them, nor about how many NPUs can be offered in each rack. What we do know is that Qualcomm’s accelerator cards will support up to 768 GB of memory, more than what NVIDIA or AMD offer in their current models. according to CNBC. Chips for third parties. The other important point is that Qualcomm will sell its AI chips and other components separately, allowing large AI companies to “customize” their own racks based on Qualcomm chips. It is an identical philosophy to the one they have adopted in the world of their mobile SoCs. Investors viewed the news with exceptional optimism, and Qualcomm shares rose 11% in Monday’s session. NVIDIA dominates with an iron fist. In the AI ​​chip segment, the king is NVIDIA. The company is the absolute protagonist of this market and according to CNBC it maintains a 90% market share, which has allowed it to skyrocket its valuation to 4.5 trillion dollars. That dominance could now be threatened by the avalanche of chips that are arriving from various manufacturers. All against NVIDIA. AMD has its excellent Instinct, Google has your TPUsAmazon their TrainiumMicrosoft their Maia and Huawei has your Ascend. All of them make really striking proposals for NVIDIA chips, and little by little these solutions are being integrated into more and more data centers. But the real problem is not in the hardware, but in the software. The great challenge is to defeat CUDA. The de facto standard in the AI ​​industry that developers use It’s CUDAa platform that allows you to take full advantage of the capabilities of NVIDIA chips in the field of artificial intelligence. This hardware+software combination is much more mature than that of its competitors, who have the hardware part resolved (or are on the right track) but do not have a platform comparable to CUDA. AMD has ROCmwhich is especially interesting because it is Open Source, but at the moment its features still do not reach those of CUDA. Reinvent the wheel? CUDA has been on the market for almost two decades, which means that the majority of academic research and pioneering models—such as ImageNet—were written for CUDA. It is not a language, it is a vast collection of libraries, optimized frameworks (like cuDNN), debugging tools and a huge community. Developing a competitor is basically like reinventing the wheel, and migrations are expensive and companies and startups will not have an easy time assuming it. China is also in the fight. And of course, if there is another great protagonist in this race, it is China. The Asian giant, previously dependent on NVIDIA, is seeking to get rid of this manufacturer, and along with the development of advanced AI chips they are also trying to have its own AI software to surpass CUDA. In Xataka | AI is the best thing happening to nuclear fusion. The construction of ITER is already accelerating

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.