reason

With O3 and O4-mini, chatgpt reason better than ever. But more important than the model is that we can ask you certain questions

by

I read Alberto Romero a few days ago in his great The Algorithmic Bridge counting Three stories who shared something disturbing: they were Three people saved by doing something as simple as asking chatgpt About his symptoms. They did not use Prompts sophisticated or advanced techniques. They only described pain, discomfort, a concern. And they lived to tell. The interesting thing in the conductive thread of these three stories is not that the AI ​​was right (although he did). Is that Natalia came up with asking Chatgpt for that tension in his jaw. It is that Cooper thought about sharing his dog’s analytics. And that Flavio, with an increasing pain, will value consulting a Chatbot In addition to a doctor. The protagonist of these three stories It is not chatgpt, but that previous momentthat moment that someone thinks “What if …?” We live obsessed with what the machines are taking us away. Employment, singularity, creativity. But along the way we are losing something almost equally valuable by our own foot: The ability to imagine unplanned uses, unlikely connectionsways of use that break the instruction manual. The gap is not only between those who dominate technology and those who are not between those who can reimagine it and those who use it following established patterns. Is Like when we buy a new oven. There are those who feel patiently to read the instruction manual memorizing each official function, and those who immediately begin to tighten buttons, turn dials and create combinations that perhaps or the manufacturer conceived. Both will end up cooking, but it is the seconds who will discover that putting the grill A medium power while turning the fan, an impossible finish is achieved to achieve the official recipes. True innovation has always emerged from those who consciously ignore theoretical limitations. The good thing is that, in the same way as Talent is very democratic, This capacity does not understand credentials. A programmer senior You can get stuck by writing code with Chatgpt while a teenager discovers that asking him to simulate difficult conversations before having them. And it turns out that it works as an emotional mirror that never appeared in any manual. The stories that Alberto collects have something more in common: they occur in the medical field, where knowledge has historically been power. Where the diagnosis flows from top to bottom. Where the informative hierarchy is so rigid that it has its own uniform. However, there are Nature, Cooper and Flavio, investing the order, leading to the expert a diagnosis that they could not have formulated but that the AI ​​could detect. It is not a matter of blind faith in technology (Everyone confirmed with professionals). It is something much more subversive: a silent rearrangement of who has access to what information and at what time. While we keep wondering if the machines are going to steal our imagination, the reality is much simpler: The decisive is not what AI takes us away, but what we think about doing with it. There is the true competitive advantage of the 21st century: in that ability to see possibilities where others only see tools with predefined instructions. Natallia, Cooper and Flavio did not save their lives for having access to superior technology, but for daring to think differently. In Xataka | O4-mini is much more than another model of AI. It is the Tesla Model 3 of OpenAI Outstanding image | Xataka

Goal has an excellent reason to launch a gigantic flame variant 4: Specialization capacity

by

Goal advertisement Last Saturday the launch of flame 4, his new family of Open Source models of IA. The company takes chest with three multimodal variants and one especially striking for being absolutely huge in size. But it is for a good reason. Hello, call 4. It has been almost a year since the goal announced Call 3and his new family of models arrives with three different variants: Call 4 scout: the most “small”, which competes with Gemma 3, Gemini 2.0 Flash-lite and Mistral 3.1. Call 4 Maverick: Compete with GPT-4O, Gemini 2.0 Flash and Deepseek v3 Call 4 Behemoth: An absolute monster that as Meta surpasses GPT-4.5, Gemini 2.0 and Claude 3.7 in various benchmarks. The latter is not demomently available publicly. Amazing context window. These models offer a context window of 10 million tokens, something simply spectacular. That means we can enter a gigantic amount of data such as entry (PROMPT), for example huge code repositories on which to work directly. Mixture-Of-Experts. These models use mixture -of-ou-explet architecture that already took advantage for example Depseek. As we explain then, this allows to divide the model into “experts” that activates according to the type of request. That improves efficiency and has proven to be a fantastic technique for models to behave optimally with much less resources consumption. Scout has 16 experts and Maverick has 128. This type of architecture also favors the inference phase, or what is the same: the models will respond not only efficiently, but fast and fluid. Call 4 Behemoth, the “Model Professor”. It is not yet available, but this variant is absolutely huge and has two billion parameters (2T in English), when calling 3, which was huge (405b) was a five times smaller model. Deepseek R1 has 671,000 million parameters, three times less that calls 4 Behemoth. The key to this model is that it serves as “teacher” for smaller and, above all, specialized variants. The flame 4 comparative table with respect to some of its rivals. Specialization. This variant is also a perfect candidate to be “distilled” and starting from it to obtain much smaller but equally capable models that “learn” from that “teacher teacher” that is call 4, but adapting to more concrete areas and scenarios and in which they can highlight. And less censorship. OpenAI’s image generator He already took a 180º turn and applies much less inspired censorship By Grok 3. Goal does the same with flame 4, which according to the company has become “responds with a strong political inclination at a rate comparable to Grok in a controversial set of political or social issues.” Thus, we have a somewhat less “politically correct” model. For now debatable results. Although the model seems to score very well in benchmarks, experts like Simon Willinson They have tried it And they ensure that their first impressions are not especially remarkable. Gemini 2.5 Pro seems to behave much better in one of the tests he performed when summarizing and analyzing a text. However, with flame 3 something similar happened, and both calls 3.1 and calls 3.2 significantly improved their behavior. They can already be tested. Call 4 is now available In WhatsAppInstagram, Facebook or the website of Goal AI. And once again, experts offers the possibility of download italthough you will basically need a cluster with a lot, a lot of memory to be able to run them at home. They are also available in Huggingface. And soon “will reason”. Mark Zuckerberg Indian In its Instagram account that in addition to these models, next month we will see a model called 4 reasoning that will be the company’s first reasoning model. It is an especially interesting variant, especially when competing with Deepseek R1 (and its successor, which It will appear soon). Images | Goal In Xataka | Great technology have determined something this year: that we end up talking to an AI

Europe will invest a lot of money in countries as far away as Uzbekistan or Kyrgyzstan. The reason: rare earths

by

Surely it went unnoticed by the vast majority of the planet Between tariffs and war conflicts. Kazakhstan announced last week the discovery of his Greater rare earth sitewith an initial estimate of one million tons of key elements such as Cerio, Lantano, Neodimio and Ititrio, all fundamental for the global energy transition … or to begin a new arms era. And now the news that did reach more people: the EU will invest a fortune in Five Central Asian countries. The official reason? Strive ties. The truth? The track is one of the five countries: Kazakhstan. The news. In full escalation of commercial tensions with the United States, the European Union surprisingly announced an investment of 12,000 million euros in Central Asia during its first summit with the five countries of the region (Kazakhstan, Uzbekistan, Kyrgyzista, Tayikistan and Turkmenistan). The president of the European Commission, Ursula von der Leyen, stressed that these funds will go to key sectors as transport, clean energy, connectivity and sustainable development of strategic natural resources. Tariffs and foreign trade. In a context marked by the new 20 % tariffs imposed by Washington to European imports, von der Leyen stressed that the EU seeks to offer A reliable alternative Faced with powers such as Russia and China, betting on egalitarian associations and investment in local capacities. In addition, the common commitment to the Territorial sovereignty and peace in Ukraine, condemning the Russian aggression and reinforcing the message that respect for international law will be a cornerstone of this new strategic association. The EU, which already represents 22.6 % of foreign trade And more than 40 % of foreign direct investment in Central Asia, seeks with this summit to consolidate its regional influence and open new trade routes that avoid Russian territory, such as the Transcaspiano corridorKey to reduce the Eastern Energy and Geopolitical Dependence. A key region. Behind good words are not only sustainable development and regional cooperation, but a critical geoeconomic priority: ensure the supply of essential minerals For the European green transition, the strengthening of its industrial base and the development of its defense capacities, all in a context of growing global tension and structural dependence of China and Russia. Strategic minerals. The urgency of this strategic turn was evidenced after the recognition of a disturbing vulnerability: in 2023, 94 % of imports European rare earth came from China, Malaysia and Russia. In addition, China controls the 60 % of world production of critical minerals and 85 % of its processing, while strengthening its own green industry. This concentration of power, added to political proximity between Beijing and Moscow, has led to worrying episodes, such as Chinese restriction to Antimony exporta key mineral in military technologies such as precision optics and night viewers. Abundance, but with limitations. In this panorama, Central Asia emerges as a realistic and attractive alternative. Kazakhstan currently produces 19 of the 34 minerals critics defined by the EU and could expand this figure 21 in the short term After the announcement of last week. Uzbekistan, meanwhile, is the fifth major supplier Uranium World and has important reserves of gold, silver, titanium and molybdenum. The region also has lithium, silicon and tungsten, fundamental for batteries, solar panels and electronic defense systems. However, much of these resources are trapped in a poorly developed mining sector, lacking modern infrastructure and technological capabilities for sustainable extraction. There, a priori, money would be destined. The European strategy. They counted on DW That, in the face of the geoeconomic competence of China and Russia, Brussels seeks to differentiate offering cooperation models based on industrial associations and mixed companies with local actors, favoring direct foreign investment, regional business growth and progressive industrialization. This approach is especially attractive to Central Asian leaders, who see in it a way to diversify their economies, reduce dependence from Moscow and gain greater strategic autonomy. The cornerstone of this approach would be the Gateway Global Initiativethe ambitious European project of 300,000 million euros conceived as an alternative to the New Silk Route China. The transcaspian corridor and a promise. A crucial component of the European Plan is the development of the Transpian International Transport Route (Titr), that logistics corridor that would unite China and Europe through Central Asia and the Caspian Sea, reducing traffic times to 15 days and avoiding the step by the Suez Canal or Russian territory. The problem? The dimension of investment. The implementation of this corridor requires an estimated amount of 18.5 billion euros In infrastructure, of which more than half have already been mobilized by the EU through an investment forum with support from its member states, the private sector, and institutions such as the European Investment Bank and the BERD. To get an idea, the Expert Samuel Vestterbye That this route could multiply container traffic from the current 100,000 to 800,000, with a transformative economic impact for both regions. The Russian “friend.” No doubt, the European turn also has a clear geopolitical dimension: stop use which makes Russia of Central Asian countries for avoid sanctions Western imposed after the invasion of Ukraine. The European diplomat Kaja Kallas was explicit in that sense when warning in Asjabad that Russian companies should not use the region as commercial escape route. In this context, the EU needs to balance a incentive and pressures policy: Offer real economic development through infrastructure and commerce, while demanding cooperation in compliance with the western sanctioning regime. Something like the “carrot and stick” approach that analysts see as an opportunity to consolidate strategic relationships that transcend the economic. Challenges and Emergency. Despite the advances, the European strategy has notable challenges before him. Experts Like Marie Dumoulinof the European Council on Foreign Relations, warn that the concrete projects of the Global Gateway take to materialize and lack visibility in the region, which could weaken the EU’s ability to compete with Chinese or Russian offers. In other words, that what is said is credible Brussels must Accelerate implementation of infrastructure works, show tangible results and … Read more

The Spanish car will be unscathed from US tariffs for a very simple reason: we manufacture cheap models

by

25% of tariffs on all cars that enter through the borders of the United States. That is the last major measure announced by Donald Trump that, if nothing changes, will enter into force on April 2. A new economic decision that has evident winners and losers. How does Spain affect you? 25% tariffs. It had been rumored for a long time and Donald Trump has ended up confirming it: the United States will tax with 25% tariff to all cars that enter through its borders. If the car has been manufactured outside the United States, the importer will have to disburse 25% more through the car. The announcement has a small print that was later clarified by the White House, they collect in The country. For cars manufactured within the economic space of Canada-Mexico-United States (which have a special treaty) only the tariff will be applied to the pieces that are not produced in the United States, which should partially mitigate the norm. The latter is important since it is estimated that tariffs will raise the purchase cost of a car in the United States between $ 4,000 and $ 10,000 Since half of the cars bought in the country come from abroad and, from those assembled there, 60% of the components are also manufactured outside their borders, they point out in Motorpasion. And a threat. “If the European Union collaborates with Canada to economically harm the United States, they will be imposed on a large -scale tariffs, much greater than expected, to protect the best friend that both countries have had,” Donald Trump threatened in words collected by The country. The words of the president of the United States are a notice for a European Union that had threatened to reactivate tariffs that were already in force during Donald Trump’s first mandate and even go one step further, as reported The New York Times. Germany is especially vulnerable to US policies. It is the country that exports the most cars to the United States. According to Bloombergin 2024 they exported 446,566 cars worth 24,800 million dollars. In addition, its main manufacturers produce a good part of the cars that sell in the United States in countries such as Mexico or Canada. So, 80% of cars that Volkswagen sells in the country come from outside the United States. Mercedes does the same with 63% of its cars and BMW with 52%. What about Spain? According to the Data from the Department of Commerce collected by The country. Spain only exported 8,316 light vehicles that added a value of 178.5 million dollars. Both figures are, obviously, minimal if we take into account those that we have been renovating previously. According to Observatory of Economic Complexity (OEC)the main shipments from Spain to the United States represented Volkswagen and Ford. Despite this, those 178.5 million dollars contrast with the more than 8,000 million dollars that Spain exports in cars to France or Germany, the main markets for the Spanish industry. Small and European. If the Spanish industry will not suffer with the tariffs raised by Donald Trump it is because the cars we manufacture here are focused on the European market. Most of them are Compact and utilitarian vehicles of relatively low prices that are manufactured in Spain because the labor is cheaper and, therefore, allows to better adjust the benefits margins. The data pointed out that Ford was one of the most affected companies in Spain because it produced in Almussafes (Valencia) the Ford Transit Connect. This vehicle was exported to the United States because it was used as a taxi in New York. However, It does not occur since 2024. To this we must add that the country has been taking positions in recent years to reposition its industry in the electric car. And production plans for this type of cars also indicates that They will be small carsdestined for a European public. Consumers. So far we have talked about the industry but, obviously, a European response will touch sectors outside the car. While in the United States half of the cars that are bought are manufactured outside the country, only 12% of cars imported to Europe come from the United States, According to Acea. If Europe decides to respond by raising more commercial barriers to the cars that come there, the main damaged will be the Tesla Model S and, above all, the American sports models such as the Ford Mutang or the Chevrolet Corvette. Models that, like the Ford Broncothey arrive at droppers and are especially expensive to compensate for the European Emissions Standards. Therefore, if the European Union decides to answer we can expect an increase in the prices of other products. For example, it was already valued to impose Harley-Davidson motorcycles. Photo | Stellantis In Xataka | Europe wants to complicate things to China but Spain has other plans: go for free to attract MG and Byd factories

Openai is about to be worth as much as Coca-Cola even though it is still a money losing machine. The reason: SoftBank

by

Openai is about to close a gigantic financing round led by SoftBank that will double its assessment as reported Bloomberg. This new round will coincide with a non-tan-positive scenario for the company: that of increasing doubts about the rhythm of real progress of generative AI. The panoramic. The startup is finalizing a financing round of 40,000 million dollars led by SoftBankwhich will raise its value to 300,000 million. This assessment would place Openai in the select club of the most valuable companies in the world, drawing with a whole Coca-Cola, and getting closer to others such as Netflix or Tesla. The context. The investment is structured in two phases: An initial of 10,000 million (7.5 billion softbank and 2.5 billion of other investors). A second of 30,000 million scheduled for this year. Among the investors that accompany SoftBank are the Magnetar Capital Fund, which could contribute up to 1,000 million, in addition to Coatue Management, Founders Fund and Altimeter Capital Management. This round almost doubles the assessment achieved in October 2024, when OpenAI captured 6,600 million led by Thrive Capital at an assessment of 157,000 million. Between the lines. This spectacular increase in value could be interpreted as an attempt to maintain the perception of exponential growth at a time when the generative AI seems to have reached some plateau in its practical development. And now what. OpenAI must demonstrate that it can transform its technological leadership into a sustainable business model that justifies this astronomical assessment. And Altman will have to overcome the sandwich made by his investors (demanding returns) and the infrastructure (which requires continuous capital). After the initial euphoria with the generative AI, with the arrival of Chatgpt in the late 2022 as a clear turning point, we are going to a somewhat more selective phase in which only market leaders will continue to capture large amounts of money. In Xataka | The new OpenAi image generator takes a 180º turn in a section that we did not see coming: censorship Outstanding image | Mariia Shalabaieva in Unspash

For some reason, there are people trying crash against a wall. Makes more sense than it seems

by

A Tesla Model and, a wall that simulates a road, a youtuber of 65.8 million followers, its reputation. The reputation of Tesla, its security systems and semi -autonomous driving. Another YouTube. A cybertruck tesla. Another wall with another painted road. Indeed, there are people by throwing their tesla against a wall: what is happening? The coyote game and the roadrunner. “I am in my tesla at 64 km/h with the activated highway launched against a false wall in the style of the coyote and the roadrunner.” This is how it starts Mark Rober’s videoknown NASA engineer with a YouTube channel in which 65.8 million subscribers are counted. Rober travels in his car thrown against a false wall in which he has simulated that the road on which circulates continues and perfectly simulates what it really is: a decoration. The objective is to check if your Tesla Model and equipped exclusively with cameras offers better or worse performance than a car equipped with a Lidar system As far as autonomous driving systems are concerned. The background. Rober uses his video to explain the differences between a car that bases its semi -autonomous driving in the cameras exclusively and those that use the Lidar system. In 2021, Tesla already advanced that would base its entire system of driving aid in the cameras. In 2022 it was confirmed that the company would stop riding radars and sensors. The company ensures that its software is able to analyze everything that happens exclusively using cameras with reliability that matches that of radars and sensors. Along the way, evidently, the cost of assembling these components that irremediably increase the final cost of the vehicle is saved. To check if this is true, Rober tests A Tesla Moder and that only uses cameras In its Autopilot system and a lexus RX equipped with a lidar sensor of the luminar company. The cars are launched with activated semi -autonomous driving functions (Autopilot in the case of Tesla and adaptive cruise control in the lexus) to verify what happens in different scenarios. Among them, children who appear in the trajectory of the car or very intense rains. The wall. During his tests, the vehicle equipped with Lidar December and saved a child from being hit if he was already on the road, if he appeared behind an obstacle, in cases of fog and intense rain or if he was dazzled by an intense light. And yes, also if there was a wall on the road painted with the simulation of a road. The Tesla Model and could not say the same and the cameras did not exceed the tests of the fog, the rain … and the famous wall. In the latter case, the car was cheated and thought that the road continued, not stopping at all. And has exploded the controversy. The video Mark Rober has exploded the controversy on social networks for various reasons. The first and most obvious: does this type make sense? When can a car find a wall that completely covers a road simulating that it continues later? Beyond that this has to value each other, what is true is that Rober activates the highway a few seconds before reaching the wall and in the video it is observed how the system is disabled when it is going to collide. Rober has been accused of turning off the system but has defended himself on social networks publishing a complete video in which it is seen how the system is disabled before the crash. A criticism that has been repeated To Tesla vehicles, accusing the company of washing their hands before the authorities. But not only that, Rober has also been criticized for using a vehicle with autopilot and not with Full Self Driving (FSD)the most advanced autonomous driving system in Tesla while Lexus was equipped with Lidar, one of the most expensive technologies at the moment. Cameras vs lidar. The experiment has something interesting although not a resolution. The truth is that one might expect that a cameras based system is less effective than one accompanied by a lidar. The single combination of mounting a camera and a radar has been demonstrated The most effective option So far both day and night. The cameras analyze with images what they have in front but a lidar sends pulses of light constantly around them. That light bounces on the object and returns to the lidar. With the time that has passed since the shipping of the pulse of light until it receives it again you can create a 3D map that, in the case of the wall, serves to verify that there is an obstacle on the road even if it can deceive the camera. Autopilot and FSD. The other great criticism that has been made to Rober is that he uses a car equipped with Autopilot but not with the FSDthe most advanced driving aid system of Tesla. The FSD is offered as an additional option through a subscription and allows the driver to disregard driving (even in urban environments) as long as it remains attentive to what happens in front of his eyes. For example, the Autopilot demands from the driver the confirmation of lane change but the FSD navigates between the streets by itself, without the need for driver’s validations. In fact, in the United States it allows The car travels 65 meters without a driver Inside. Thus, we can ask a tesla parked to come to pick us up from a distance. And there are people by throwing their cars against a wall. Rober’s experiment has given wings to other YouTuber who have come out to defend the company. In response, Kyle Paul, who has 237,000 subscribers on his YouTube channel and specializes in Tesla videos, He has published his own test. The goal has been the same: will your Tesla Cybertruck before a wall that simulates the continuation of the road? And his answer is that if the car has … Read more

I’ve been trying electric cars for years. I bought one of combustion for a single reason

by

We live complicated times to Choose car. Or, perhaps, we believe them more complicated than they really are. Yes, it is true that we can have doubts about What a course will take The automobile industry, what technology will be imposed, what not or, directly, What restrictions will raise Against my car in the coming years. But let’s look at this from a positive point of view: we have never had so much variety and so many technologies that will better adapt to our lifestyle and the trips we made on a day -to -day basis. This is what I try to explain to those who, knowing that I write about the automobile industry and try some of the most recent models, ask me what car they should buy. What should be the best option For them. To give them an answer I give a series of tips. The most important, although it seems nonsense, always leave it for the end. Choosing car Diesel (yes, diesel), gasoline, Microhíbrido, electric hybrid, plug -in hybrid, electric or gasoline bifuel and liquefied or compressed natural gas. It is also possible to buy a Electric of extended autonomyalthough the possibilities are very small. Or a car moved by hydrogen but, in this case, Recharge options are almost non -existent. Therefore, if we are in this situation and we want to buy a car, we have to think about several aspects that I summarize in the following steps: What type of car do we need (by size)? Here we must contemplate what we are looking for exactly. How much trunk do we need? Are we going to use the rear seats usually or do we travel as a couple? Are we a family that will grow, are we thinking about forming it or do we plan to have children? Do children already travel on their own and do not need such a big car? Personally, I recommend doing a Very rational reading of our situation. If we usually do not carry much luggage we should not prioritize the trunk. If we usually do not use the rear seats, their habitability should have a secondary weight in our decision. It is simply not to over -dimension our needs. Yes, if we are traveling with little luggage in our weekend getaways and, for example, one day a year we want to move a bicycle because we will have a problem or we will have to better square the luggage. That day, perhaps, we regret a little of the choice but this will be passenger because the rest of the year we will omit a quality for which we have probable we have paid a surcharge. If, on the contrary, we are fond of mountain bike and every weekend we load and download our bicycle, I will not only recommend that you buy a car that facilitates this task. I say more: Take the bicycle to the concessionaire And check for yourself if the bicycle enters the trunk and how simple or complicated it will be its load and download. What use are we going to give our car? I like to define the size of the car in the first place because, for almost all kinds of sizes there is the technology we are looking for. Largely because tAmaño and type of use go hand in. If we are looking for a car to move exclusively in the urban environment, to go and return from work and leave children at school, we are not interested in a car too large, because in this type of journey it will be much more uncomfortable to use. On the contrary, if we move regularly on the highway and highway and perform hundreds or thousands of kilometers every week, we will prioritize comfort and, most likely, the car is larger. In this case, I would prioritize the following aspects, from greater to less importance: Are we going to use this car as the only one in the house? What kind of displacement do we do in our day to day? Do we usually go on a trip? How many long getaways do we do a year? How many kilometers do we do a year? What technology do we choose? Once we have the adequate car size for our needs and we are clear about the type of use that we are going to do, it remains to choose which technology that interests us most. The money is decisive here and, if necessary, access to a plug to load an electric car or a plug -in hybrid. My recommendation is as follows: Diesel: Who performs many kilometers on the highway and highway and site the city little. I would not buy it if I did less than 20,000 kilometers a year. Gasoline: We want a car “for everything” at home. We move in the usual figures for a Middle driver And the use we give to the car is mixed between city and getaways on highways and highways. Microhíbridos: If the option exists, the microhíbrido is a good option to obtain the ECO LABEL AND ITS BENEFITS If we are looking for a gasoline car because the cost barely differs and we will not notice any change when driving. Hybrids: We also have the car as a “vehicle for everything” but most of the kilometers are made in city. Although we can leave the weekend, we make a good dozen kilometers in the urban environment or we need enter a zbe where access to more polluting cars is restricted. Plug -in hybrids: same case as the previous one but we have the possibility of loading a plug at home and weekend getaways are usually to a second close home. This is important, once the battery charge is exhausted, with the combustion engine working exclusively, the car can have too high consumption that spoil daily savings. Electric: We have access to a plug. You can choose to dedicate it as a second … Read more

China needs cardiologists in an increasingly aged society. His answer is an AI that “reason” as the best

by

820,000 people. That is the number of patients that the Department of Cardiology of the Zhongshan hospital attended in 2024. A very high figure for a country in which doctors are missing and that sees that, every so often, every so often, His hospitals overflow. Increased higher health coverage, especially for an increasingly aging populationand the answer has been found in artificial intelligence. In a cardiologist, specifically, which has trained with a base of hundreds of thousands of people and has all the support of the government. Cardiomind. That is the name of artificial intelligence that has been developed by the Zhongshan hospital, the University of Fudan and the Shanghai Artificial Intelligence Academy for Science. It is a joint effort to create an AI that, according to them, is able to reason as their best cardiologists. GE Junbo is one of those outstanding cardiologists of the Chinese Academy of Sciences and Comment They are feeding the system with data, but they also teach him “to think as an expert first level cardiologist.” Co-pilot. In this way, Cardiomind compares patient history and the results of tests based on global research to generate diagnostic suggestions. And if you were raising one, leaving the possibility of leaving your heart’s health on the awareness of a machine, in the word “suggestion” is the key. According to its creators, the idea is not that doctors are replaced, but that it helps those members of a congested system to work faster and more accurately. “With their help, our doctors can attend more patients, reduce the workload and improve both the quality of the diagnosis and the treatment,” commented The doctor. Pillars. The secret of the pizza is in the dough, in the case of AI, that special sauce is the training. Cardiomind is a multimodal system that can analyze and process data from various sources (electrocardiograms, ultrasound and laboratory results, for example) to provide its diagnosis, and it is what has been trained. The source has been hundreds of thousands of medical records that the Hospital Cardiology Department had stored for decades, learning in the process how doctors think both in the diagnosis and during treatment. In addition to multimodal and very trained, it specializes in one thing: cardiovascular diseases. Total government support. And all this comes from the hand of itself. China is investing strong in development of artificial intelligences and, although in the West only a few as Deepseek They are the ones that attract attention, the country has been betting for the development of these systems to decongest several organisms. And that of medicine is fundamental. In January of this year, and apart from Cardiomind, established In Shanghai the first test and verification center of large models of AI in the field of medicine. Their fields will be assistance, disease prediction, personalized treatment according to the history of each patient and support to develop drugs. Necessary. Such is that impulse that several hospitals already have their own AI platforms, even some that They combine Contemporary and traditional medicine. And the reason why this technological development is being promoted is the one mentioned above. The population is getting older and doctors are missing. It is estimated that the country has 1.9 doctors per 1,000 inhabitants and They want have 5.5 nurses per 1,000 people for this year. In other countries, the average is devastating. In Spain, for example, there is 6.21 doctors per 1,000 inhabitants. In the case of nurses, the figure is 7.12. Challenges. Now, despite that government impulse, there are three elements to take into account. One is that of the data security to guarantee privacy. In the case of Cardiomind, those responsible developed a firewall with encrypted data. Other countries are studying How to apply AI in cardiology and, precisely, data leakage and ethics is a point that is usually present Among the concerns. Ethics also plays a role, since if it goes from suggesting to a more important role, it would be necessary to legally regulate it. At the moment, or they are in charge “only” of paperwork and records or, if they attend decision making, the last word has a human. In addition, there is the technological issue itself. The development of these artificial intelligences needs servers and calculation power. It is something that China is solving more or less controversial forms buying items that They should not be able to buy due to the commercial war with the West, but also using solutions developed at home, such as the chips of a huawei that is the spearhead in development of hardware for AI in the Asian giant. Images | Huawei, In Xataka | It is impossible to escape from Huawei: not even the oldest hospital in Spain in its ambitious digital transformation

Zuckerberg dismissed 5% of the goal template “for low performance”. His former employees say there is another reason

by

In the same way as It is not the same to resign That they fire you (although the result is that you are left without employment), it is not the same as they fire you For a staff cut that for “low performance”. Goal, ha 5% of its workforcearguing that they were low performance employees. However, this public justification has enraged the 3,600 meta -employed since they stigmatizes them as bad workers when they apply for new jobs. The former employees say that their work performance has only been an excuse, but not the real reason for their dismissal. A “cruel” trick to cut staff. “I have decided to raise the level of performance management and fire low performance employees more quickly,” wrote Mark Zuckerberg in the internal memorandum sent to its employees in January. Such and as he published Business InsiderThis was raising the scale of what the company considered “low performance” to put employees in the pillory who, until that moment, considered that they had a good performance. Sally Maitlis, a professor of organizational behavior and leadership at Saïd Business School, assured to Fortune That the “low performance” label could be a ballast for the professional career of the former employees dismissed, doubly penalizing them with the loss of their employment and closing the door to find a new one. “It is a terrible way of being labeled and is clearly not useful for anyone in the labor market,” Maitlis said. They have shouted the four winds. Before such injustice for them, employees have not hesitated to raise their voice from social networks to claim their professionalism and denounce the bad faith that, according to them, has used a goal in their dismissal. “Today they have fired me, but not for being a low performance employee. Let’s be clear: that label is misleading and, for many of us, it is totally wrong,” ” published on his LinkedIn account Steven S, former product of Instagram product. Kaila Curry, who worked as a content manager, said in her LinkedIn profile that: “I often asked for comments and they always told me that I was doing a good job. They never assigned a PIP (performance improvement plan), they never gave me correction indications and never guided me or gave me clear expectations. I just struggled … I am not a person of low performance.” A mortgaged future. The main complaint of the employees focuses on the argument that has publicly wielded a goal to justify the dismissal of the 3,600 employees since, as they argue, it harms their options for free to get a new job. “The most difficult thing is that goal says publicly that they are cutting low yields, so it feels as if we had the scarlet letter on our backs. People need to know that we do not have low performance,” declared one of the employees dismissed to Business Insider. “It is possible that these people are superstar in other places. Therefore, it seems to me that it is an additional punishment because they are people who probably have a very high market value,” He pointed out to Fortune Dan Cable, Professor of Organizational Behavior at the London Business School. Employers are answering. One of the most recurring advice when running for a new job is not criticizing the company that has just fired you. However, against any forecast, the strategy of the finishing targets has had a positive impact. Many recruiters and companies have shown interest in the profiles of the target employees. “Good employers look for people who are good communicators, capable of defending themselves, just as they would in their position within the organization,” He stood out to Business Insider Ashley Herd, Recruitment Expert from LinkedIn and founder of a manager recruitment company. “Those who speak of course can be found, being realistic, with a smaller group of possible employers, but it will be a group that really aligns with their values ​​and supports professional success and long -term happiness,” said the expert. Now stigma is target. Mark Zuckerberg said the layoffs were not due to economic difficulties, but to a restructuring to focus on the IA development And, in fact, vacancies had opened to incorporate into this new profile. Since the hiring process is a relationship of two, the complaints expressed in social networks by their ex -employed can complicate that process. User forums as Reddit or Teamblind have filled the new candidates about the new goal policy with warnings. “It seems that it was more money than performance. Be careful to join this company. Zuck does not care about its employees. Only the company,” wrote A meta -employed. “Meta is now the most cruel technological company that exists,” added another former partner. Other great technological companies Like Microsoft or Amazon have already announced that they will take more into account the performance of their employees when signing their layoffs. In Xataka | A 1.3 million rolex and a brilliant suit: the new Mark Zuckerberg has celebrated his wife’s birthday Image | Goal, Unspash (Annie Spratt)

Satya Nadella is clear that at the moment AI has not revolutionized anything. The reason: money

by

“Teach me the pasta!” That is what Satya Nadella wants. That the AI ​​shows the pasta. Give (a lot) money. It is something that Microsoft and Openai have already referred to both. For these companies, AGI’s definition is precisely linked to money, and have established that An AGI will not be an AGI until you give 100,000 million benefit. That argument now serves Microsoft’s CEO to talk about the fact that AI has not revolutionized anything. From agi nothing. In one recent interview With the Dwarkesh Patel podcast, Satya Nadella, CEO of Microsoft, made some important statements in which her position was clear about the current situation of the artificial intelligence segment. To begin with, he indicated that “that we claim that a milestone has been achieved in AGI It is only to hack the meaningless evidence “, that is, to give greater importance to the benchmarks they really have. And of revolution, either. In fact, the best metric to measure the success of the AI ​​segment is simple: it should increase the gross domestic product of any country. “When we say: ‘Oh, this is like the industrial revolution’, we should have that type of growth that caused the industrial revolution. For me, that means 10%, 7%for the developed world. Adjusted to inflation, growing at 5%, that is the real marker.” AI needs a ‘Killer app‘… Nadella commented how that growth has not yet occurred because most users have not yet understood how to use AI effectively. It is the same that happened with the PC, it took time to find its place and demonstrate its potential. … like what Excel and email achieved. Microsoft’s CEO remembered how before the PC, email and spreadsheets, companies made their business forecasts almost handmade: “Faxes circulated, someone received them and then made a memorandum between offices that then circulated, and people introduced figures, and in the end a forecast came out perhaps right in time for the next quarter.” But then Excel arrived and email and revolutionion that type of task, like many others. “That is what we need to happen with AI when it is introduced into jobs in the field of knowledge.” The AI ​​will supervitaminar. The debate on the impact of AI on work is constant, but for Nadella this technology will help us extraordinarily. Thus it will allow any worker to focus on high value tasks, and not on routine tasks that can be automated. He joked talking about the amount of time dedicated to filtering his email, and how to avoid that will be an exceptional time and productivity gain. The AI ​​not only does not give money, but it loses it. The vast majority of large companies that are betting on AI and developing large language models are losing money, and they are also losing it to beast. Openai is the best example of Burn money as if there were no tomorrowbut it is also the best positioned to win that race. But this is a bet. In fact, probably the greatest in history. It is demonstrated by the colossal investments that Microsoft, Amazon, Meta, or Google are doing to create data centers dedicated to AI. All of them will dedicate dozens of millions of dollars to that taskand they will even when it is not clear that AI will be profitable. Of course, these companies believe it will be, and much. Image | Sony Pictures In Xataka | Satya Nadella asked to lower her salary for not having complied with security. Microsoft compensated with 79.1 million

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.