search

exit Google search

by

Europe has proposed let depend on the US in technology and now is taking real action to get the desired digital sovereignty. According to Politicalone of these measures will come into force tomorrow and consists of something very specific: saying goodbye to the Google search engine. what’s happening. According to an internal email to which Politico has had access, as of Thursday, June 4, “Qwant will replace Google as the default search engine on the European Parliament’s computers.” The reason for this change is to continue “Parliament’s commitment to digital sovereignty and the protection of users’ personal data.” Why is it important. The European Union has realized (too late) that The US is not the reliable partner I thoughtand that depending on their technology is a problem. Changing the search engine on parliament’s computers is a symbolic measure, in the sense that it is one of the easiest to execute and will not have a real impact on technological independence, but it is important to send a message: it is possible. Qwant. It is a search engine that focuses on privacy protection, in fact its slogan is “the browser that knows nothing about you.” Qwant does not store search history or resell browsing data, has an extension for Chrome, Edge, Firefox and Safari and comes with built-in AI to summarize web pages. And most importantly, it is European, specifically French. Yes, but. When an MEP searches from the address bar in the browser of one of the parliament’s computers, Qwant will be the default search engine. However, they will be able to both access Google to search and change the settings to make it the default search engine. That is, there is no prohibition or block to use Google or any other search engine. More measurements. A few days ago we learned that the European Union is preparing a “technological sovereignty package” with four key measures to reduce its dependence on the United States. These measures focus on four critical areas: Cloud: audits and stress tests for suppliers. Chips: the European Commission will review the Chips Act to be able to cancel contracts and centralize emergency purchases. Software: creation of a strategy to promote European free software. Data centers: 200 billion for the construction of data centers and 20 billion for digitalization and AI in energy. It’s going to be complicated. The EU’s dependence on US technology is enormous. According to Synergy data70% of the European cloud market is owned by just three American companies (AWS, Microsoft Azure and Google Cloud) and only 15% is controlled by European companies. In the field of chips, although there is ASML, the external dependency is enormous as evidenced with the case of Nexperia and China. Not only is Europe late, its ability to compete in these critical sectors is negligible compared to that of the United States or China. Image | Guillaume Perigois and Alex Dudar on Unsplash In Xataka | The CEO of Mistral sends a message to Europe: the end of being the technological vassal of the United States

We have small and giant black holes, but the intermediate ones do not appear. Now some scientists have designed a method to search for them and they already have two candidates

by

Today astrophysicists have a lot of information about black holes. They have even been photographed. However, there are only two types of black holes for which a multitude of evidence has been found: supermassive black holes, which are colossal in size, and stellar black holes, which are formed by the collapse of a star when it runs out of fuel. Supermassive ones usually have masses between 100,000 and 10,000 million solar masses. The stellar ones are much smaller, with approximately a mass equivalent to that of 3 to 100 suns. So what happens in the intermediate range? Don’t black holes of intermediate mass exist, between 100 and 100,000 solar masses? This is a question astronomers have been asking for a long time. Theoretically, they could exist, but no evidence has been detected. Now, a team of scientists from Yangtze University in China, has devised a method which could be useful to find them once and for all. Gravitational microlensing of fast radio bursts. These scientists have used a method that is based on searching for fast radio bursts that have experienced a gravitational microlensing deformation. These deformations are caused by massive objects that stand between the path of the blast and the Earth. By studying the effects of these disturbances, its mass can be calculated. For this reason, these scientists have analyzed a catalog of these bursts, looking for those that may have been distorted by gravitational microlensing and have been left with two candidates whose mass would correspond to an intermediate black hole. The best? They also fit with primordial black holes, so they could even serve to better understand dark matter. Clarifying concepts. Now let it be understood. Fast radio bursts are short bursts of radio waves, which come from far away, beyond the Milky Way. There is no consensus on its origin, but many have been detected, it even seems that there are a large number in a single day. For their part, gravitational microlensing are formed when a very massive object comes between a light source and the Earth. It is so massive that, due to the action of gravity, it doubles space-time and, with it, the path of light that reaches Earth. As a result, multiple and/or magnified images may form. The point is that fast radio bursts themselves can be altered by gravitational microlensing when a very massive object crosses their path. Gravitational microlensing Canadian Hydrogen Intensity Mapping Experiment (CHIME). This is a Canadian radio telescope initially designed to map the presence of hydrogen in large fractions of the observable universe. Thanks to its large collecting area and field of vision and its bandwidth, over time it was seen that It was also very useful to detect fast radio bursts. They are very fast, but when observing so much sky at once they do not go unnoticed. For all this, the authors of the study that has just been published have analyzed the CHIME catalogwith special attention to bursts that at some point have suffered a disturbance by gravitational microlensing. Two candidates. Of all the distorted bursts they found, there were two whose size matched possible black holes of intermediate size. One had between 539 and 609 solar masses and the other between 1,544 and 2,571 solar masses. Curiously, there were no galaxies or galactic clusters around it. When black holes form through physically well-known collapse processes, they are usually in the centers of galaxies. However, when they are isolated, as is this case, what is expected is that they are primordial black holes. That is, black holes that formed in the early stages of the Big Bang, before there were even stars that could collapse. something unexpected. These scientists expected to find intermediate black holes, but they may also have found evidence of the origin of dark matter. One of the hypotheses about this mysterious matter that makes up most of the Universe is that it is partly composed of primordial black holes. The problem is that the existence of primordial black holes has not been proven. With this new study, two birds could have been killed with one stone: demonstrating that intermediate mass black holes and also primordial black holes exist, in turn helping to unravel the dark matter mystery. Without a doubt, it is a cosmic carom that is worth continuing to investigate. Image | THAT | POT In Xataka | Stephen Hawking made a prediction about black holes in 1971. A new signal has proven him overwhelmingly right

Google search engine is killing blue links. So more and more users are finding refuge in DuckDuckGo

by

Search things on Google it’s not the same anymore. The technology giant is turning its traditional search engine into something very different in which AI is taking control whether we like it or not. The good news is that if we don’t like it, we have alternativesand precisely in DuckDuckGo They are taking advantage of this opportunity very well. Times change. For nearly three decades, Google has been the gateway to that traditional list of blue links that the search engine presented as results. In recent years, the search engine is becoming an AI engine that tries to anticipate the user’s intentions and autocomplete queries with long answers. AI Overviews. The arrival of AI Overviews (AI Summaries) has made the search engine answer questions directly at the top of the page, but that move is being heavily criticized. The reason is simple: Google is killing those blue links it defended for more than 25 years. The new AI Mode goes in that same direction and points to a future in which we will not ask the Google search engine things, but simply talk to it. DuckDuckGo takes advantage of its opportunity. DuckDuckGo CEO Gabriel Weinberg recently stated that “Google is imposing AI with no way to disable it.” This has not pleased a certain sector of users who are beginning to use alternatives that give them back control. That’s good news for DuckDuckGo, which has seen more and more users start using its services: Downloads of the DuckDuckGo app in the US grew by an average of 18.1% weekly, and peaks of 30.5% were reached on May 25 The adoption rate on Apple devices was even higher, with an average rise of 33% and a peak of 69.9%. Visits to specific page free of AI, noai.duckduckgo.comincreased on average by 22.7% from week to week. Searches like before, and private. The advantages of DuckDuckGo are striking for those looking for alternatives to the Google search engine. To begin with, DuckDuckGo has that aforementioned page that avoids any trace of AI. If you want to use AI canbut in that case you will have access to somewhat more modest models (Claude 4.5 Haiku, GPT-5 mini, Llama 4 Scout, you can pay to access better models) but with which your conversations will remain private. In addition, DuckDuckGo deletes IPs and prohibits the use of those conversations to train models. It is an extension towards AI that iron defense of privacy that since its inception has differentiated this search engine… although there have been some controversy about it. Google doesn’t even flinch. Despite this growth in downloads and visits to its search services, DuckDuckGo remains an alternative with an anecdotal market share in the search segment. Worldwide it has a 0.71% share compared to Google’s 90.02% and Bing’s 5.14%. Anti-AI shelter. Google’s dominance here is absolute and it seems almost impossible for that to change, but even so DuckDuckGo is clearly positioning itself as the refuge for those who see AI as an unnecessary—or unwanted—complication when performing searches. Kamyl Bazbaz, head of company policy, explained it with a simple statement: “people simply want to be able to choose.” Business model at play. The shift towards AI seems to be beneficial for Google, whose search revenue grew by 19% in the first quarter of 2026. It is a striking fact considering that theoretically AI results make it difficult for advertising to enter the search engine, but Google seems to be solving it with native advertising from AI Overviews or Keyword auctions in AI Mode. This is added to the recommendations of affiliate products that also seem to be working and promoting a business model that is being renewed in parallel to what the search engine is doing. In Xataka | Google already knows how to make its AI Overviews more human: use Reddit content ruthlessly

A new search method detects dozens of possible planets with two suns that had gone unnoticed by us

by

Nowadays, more than 6,000 planets are known which, like ours, revolve around a star. On the other hand, only 18 planets have been found that orbit a binary system, with two stars. For this reason, they are considered a rarity. They are so rare that the best known of them all is Tatooine and, far from existing, it belongs to Star Wars fiction. However, a team of scientists from the University of New South Wales has decided to change the method we use to search for planets. In doing so, they have suddenly found 27 candidates for planets with two suns. An underused method. The method that have used These scientists know as apsidal precession. It was already used in the past to characterize binary star systems. However, until now it had not been used to check if there is another object, such as a planet, within that same system. Broadly speaking, it consists of locating possible changes in the eclipse calendar of the two stars. If these changes cannot be explained with general relativity or stellar physics, there must be something disturbing their movements. This is how, thanks to the TESS telescope, 27 candidates for circumbinary planets (with two suns) have been located, although it will still be necessary to confirm which ones are really planets. Stars playing hide and seek. Eclipses occur when, from the position of the telescope observing them, one star interrupts the light of the other. In a known binary system, these eclipses are predictable. Therefore, if we see something that doesn’t add up, there could be a planet in the way. TESS typically relies on the transit method to detect exoplanets. It detects periodic disturbances in a star’s brightness, which could indicate that a planet is crossing between it and the telescope. However, if the planet has an irregular orbit that is not in the telescope’s line of sight, it may go unnoticed. However, with this new method that is not a problem, because you do not have to see the planet or the changes in the brightness of the star. It focuses rather on the gravitational effects that affect its two stars. It doesn’t matter that its orbit is not visible to us. It’s just the beginning. This team has detected 27 candidates for circumbinary planets in a total of 1,590 two-star systems analyzed. That means about 2% of these systems could host planets. If this is true, thousands more planets could soon be detected. For a long time, exoplanet detection would have been highly biased. Artist’s representation of a system with two stars A great variety. The smallest possible circumbinary planet that has been detected has a mass similar to that of Neptune, while the largest is 10 times more massive than Jupiter. The closest one is 650 light years away from us, while the furthest one is 18,000 light years away. There are candidates in the northern and southern skies. In short, there is a great variety. That also supports the hypothesis that there are a wide variety of planets out there that, until now, have been invisible to us. And now what? Now it will be necessary to check which of those 27 candidates are really planets. Some stars, such as brown dwarfs or white dwarfs, could also alter the eclipse calendar of the binary system. Even black holes could do it. Therefore, it must be ruled out that it is any of these phenomena. To resolve this question, another instrument will be used, the Anglo-Australian telescope from Coonabarabran. With it these other very massive objects could be detected. If no plausible explanation is found, it would be concluded that it is a planet. What we can learn. Having techniques to detect another totally different type of exoplanet gives us much more information on planetary formation. There could be planets similar to Earth, whose only difference is the existence of two suns. Some of them could even host life or have hosted it in the past. The range of possibilities would open up greatly. About half of the stars in the universe are found in binary or multi-star systems. And all of that is still unexplored. Image | Star Wars | NASA/JPL-Caltech/T. Pyle In Xataka | There is only one chance in 11,000 years to reach the planet Sedna. Some Italians want to use this nuclear engine

The hantavirus outbreak has sparked a search to find its origin. Ushuaia fears that this will cost it its tourism

by

Until a few days ago, the vast majority of Spaniards had never heard of hantavirus. Probably not Ushuaiathe capital of the province of Tierra del Fuego. The crisis unleashed by the MV Hondius cruise ship, however, has united both names in one of the most worrying episodes so far in the turbulent 2026, at least in health terms. And so worries in Ushuaia. A lot. Although there are still unknowns about how, where and when the viral outbreak broke out, in the southern region they fear that what happened affects one of the pillars of their economy: tourism. They even talk about a “smear campaign”. And the hantavirus arrived. Although there are still many months ahead until December, 2026 seemed basically doomed to be the year of the iran waroil through the clouds, Maduro’s arrest or (in another order of things) the North American World Cup. Now that list is expanded with one more item: the hantavirus. Since on May 2 A viral outbreak was confirmed on board the MV Hondius, the world is waiting for what happens to the ship, its passengers and the chain of possible people infected by a virus that until almost two weeks ago was almost completely unknown in Europe. In a place in Patagonia… In Ushuaia, capital of the province of Tierra del Fuego (Argentina), one of the southernmost cities of the planet, the evolution of the outbreak is also being followed with interest. Although there what really generates debate is not the last hour about the outbreak or what may happen from now on, but what has happened in recent weeks. The reason is very simple: the fateful voyage of the MV Hondius left its port April 1. When the ship set sail, there were about 140 passengers on board, including the septuagenarian Dutchman who only a few days later (on the 6th) began to show symptoms of infection. More than a month has passed since then, four long weeks during which events have happened at breakneck speed. That first patient died on April 11, days later his wife did and since then at least one more dead and half a dozen infected. As for the MV Hondius, after the evacuation of the last hours on board the ship there are only a few dozen of people who will continue heading to the Netherlands. What has not changed is the question that authorities have been asking for days: Where is the origin of the outbreak? Where the hell did the Dutch couple get infected? A remote landfill. Taking into account the incubation period of hantavirus, which ranges between one and six weeks, authorities are working with the hypothesis that the epicenter of the outbreak is not on the ship. That is, the most plausible theory (at least in appearance) is that the virus they took him to the MV Hondius one or more travelers who were already carrying it before navigation began. That made all eyes turn first to the couple of Dutch retirees and second to Ushuaia, the place where they embarked. In recent days the conversation has revolved around a very specific point in the town: a garbage dump located about seven kilometers from the center of Ushuaia, a place where, assures The Countryepidemiologists are looking for traces of infected rodents. Hantavirus infection, let us remember, is contracted mainly by coming into contact with the urine, feces or saliva of certain rodents. The most common thing is that contagion occurs by inhaling remains of this waste in poorly ventilated spaces, but it could also be contracted in a large landfill. The unknowns begin. That a Dutch couple (he 70 years old, she 69) stop by a mega urban garbage dump before embarking on a luxury cruise through the South Atlantic may sound strange, but it makes a lot of sense. The MV Hondius trip was not just any trip. It was planned as an expedition cruise to contemplate Atlantic fauna. And the Ushuaia landfill is not just any garbage dump either. Lovers of birds and natural photography usually go there to enjoy the species that fly over it, including scavenger birds such as the white matamico. In recent days it has been published that some of the travelers who boarded the MV Hondius visited the dump. The Country even interviewed to a guide who was in the area with some of the tourists from the cruise, although the Dutch couple was not among them. Did they get infected there? Is that the epicenter and genesis of the outbreak? Hard to know. First because the marriage (unfortunately) has died. Second, because before boarding the cruise the couple had made a long road trip that took them to different parts of South America. In fact, it is believed that they were four months visiting several countries on the continent, including Chile and Uruguay. “We have no evidence”. This fact (that the first deaths were in other parts of South America) has been strongly emphasized by the authorities of Ushuaia, who do not quite understand that the couple was infected in their territory. The reason? The main one, insists Juan Facundo Petrina, general director of Epidemiology and Environmental Health of the province, is that the hantavirus is not a problem in the area. “In Tierra del Fuego we have no record of cases in our history,” clarifies to the BBC. “Specifically, since 1996, when the National Surveillance System included it among the notifiable diseases, we have not had a single case in Tierra del Fuego.” More than 1,000 km north. As if that were not enough, Petrina details a few more facts. To begin with, the hantavirus endemic area is more than 1,500 km to the north. Also remember that there are no records confirming that the mouse subspecies that transmits the disease lives in the area. “And if rodents begin to move, since they do not respect borders, it is important to remember that we are an island,” duck. Another key that … Read more

the search against time to locate the “loose ends” of the hantavirus cruise

by

Although how much the experts insist In which the risk for the populations is very low, the fear of the hantavirus continues to travel through each of the countries through which the passengers of the MV Hondius have circulated. Actually, it’s normal. Fear is a very human emotion, which helps us be alert in situations as new as this one. There is considerable uncertainty in this situation, especially in relation to passengers who got off the ship before the hantavirus outbreak was confirmed. However, little by little this uncertainty is fading as information about their countries of origin arrives. Saint Helena, April 24, 2026. On April 24, 13 days after the death of the first infected passenger with hantavirus, the ship made a stop on the British island of Saint Helena. He took the opportunity to lower the body and prepare it for repatriation. His wife traveled with him, who became the second victim, after traveling to South Africa. In addition to the two deceased, it is known that on the island another 28 people got off. It was not yet known that the first victim had a contagious virusso no passenger control was carried out. Then what? Since it became known that there was a hantavirus outbreak on the ship, attempts have been made to locate those 28 people and their close contacts. It is known that among them there were passengers of at least 12 different nationalities. There were mainly British and Americans, but also people from other countries such as Singapore, Canada or Germany, among others. The countries of origin or residence of each of them have also been taking charge. Thus, some have already achieved the goal of finding them and putting them in quarantine. From the United States to Singapore. In the United States there are five people in quarantinespread across Texas, California, Arizona and Georgia. In Singapore too have been quarantined and to the two passengers who got off in Santa Elena. France had no citizens among that group of passengers, but has quarantined eight citizens who were in contact with one of those people on a flight from Saint Helena to Johannesburg. In short, many of these people are already under observation, although it is true that there are others to be identified and monitored. Many of the passengers who got off in Santa Elena have already been identified. It is not very contagious. The positive part of all this is that hantavirus is not as contagious as other pathogens with pandemic potential such as the COVID-19 coronavirus. In general, contagion between people, which only occurs with Andes variant (that of the boat), requires very close contacts. It is possible that he was a super-spreader on board the cruise ship. That is, a person from whose body a virus is transmitted exceptionally well. We saw some cases with COVID-19, for example. However, it is normal that the rest of those infected are not super-spreaders. Furthermore, it cannot be ruled out that there were several simultaneous infections when coming into contact with mice on an excursion. It would not be a patient who is super contagious, but there could have been an initial event that caused a good part of the cases. In both scenarios, the contagion capacity would decrease over time. dead end. Another positive point about hantavirus is that has what is known as a dead end. It cannot be contagious continuously. It is known that, at most, it usually reaches three links. That is, one person infects another, that second infected person passes the disease to a third and that’s it. The chain cuts itself. For this reason, even if some of the people who got off in Santa Elena could continue to be contagious, they would not infect many people beyond their closest contacts. And what about those who are still on the boat? Spain is prepared to repatriate all passengers, with or without symptoms, or treat whoever is necessary. In that case, the transfer will be carried out with the necessary safety conditions so that there are no more infections. Therefore, from now on, the important thing is that these people undergo the necessary quarantines. Each country will decide how long they will be, but always taking into account that the incubation time in the most extreme cases can exceed 40 days. Easy to detect. The genome of this virus is very well known. For this reason, the diagnosis is simple through PCR. This test will be carried out on passengers on several occasions during the quarantine, to check their evolution. They cannot be discharged without repeating it once again. That said, although it is logical that uncertainty worries us, it is important that we keep in mind the limited pandemic potential of this virus. There is a lot of work to be done to ensure that the situation does not get worse, but there are already many people working on it. Image | CDC/Magnific | Fdesroches In Xataka | We believed that hantavirus did not jump between humans. Until someone went to a birthday party in Argentina

We are increasingly looking for human answers on Reddit. That is the reason why the Google search engine is now a Reddit in disguise

by

Google has updated its search platform for the umpteenth time, but it has done so with an especially significant change. The user experience in its AI search engines (both AI Overviews and AI Mode) attempts to become more “human”. And to do this, in these searches Google will add more context to the links, such as extracts from internet forums and blogs. And if there is a beneficiary (or harmed) of that movement, it is Reddit. Google was already a gateway to Reddit. There is a behavior that Google has been seeing in its data for years and that for a long time it preferred not to publicly acknowledge: when someone wants a real answer to a real question, add “Reddit” to the end of the search. Not because Reddit is necessarily a reliable source, but because Reddit brings together real people who have experienced this issue, tried to solve it, and written about it without anyone paying them to do so. Google, with all its infrastructure and all its algorithms, had not managed to replicate that. So instead of trying it’s going to incorporate those answers directly. What exactly has changed. The search engine update will make in AI Overviews Fragments from forums, social networks and other “first-person sources” appear. When someone searches for something for which there is no single objective answer, Google’s AI will include perspectives and opinions found in all kinds of (supposedly) human sources online. Doing so will add the name of the creator of that content (or their avatar) and the origin from which said perspective comes. Google also promises to add more context about the origin of its AI-generated answers, similar to how ChatGPT or Claude include links supporting their answers. Tired of so much SEO. The reason is obvious: Google’s organic results for practical and subjective questions—”what vacuum cleaner should I buy”, “how do I cure my dog’s ear”, “what is the best neighborhood to live in in Valencia”— They are dominated by SEO and those techniques optimized to appear on Google. It is important to position, not answer the question well. That is precisely where Reddit, like other forums or personal blogs, has something that this content usually does not have: the real experience of someone who was in the same situation. Google sums it up in its own statement bluntly: “For many searches, people are increasingly looking to other people for advice.” The contradiction that Google has not resolved. There is a potential problem in this new way of conceiving these searches with AI. AI Overviews were designed to answer questions directly and thus save the user the work of clicking, reading and researching. Now they will include diverse and even contradictory perspectives from forums and social networks. So, will AI Overviews answer the question, or will it make us go back to the sources to find the answer? If it is the latter, it will not be very different from what I already did the traditional Google search engine. There is an interesting imbalance here between “we give you the answer” and “we give you context so you can find the answer.” In a sense, Google’s decision complicates searches. AI models are becoming less prone to failure. The famous cases of add glue to pizza are much less common now, and new models often boast a significant reduction in “hallucination” rates that they have. GPT-5.5 Instant, released this week, “produced 52.5% fewer hallucinations than GPT-5.3 Instant,” OpenAI indicated in its official announcement. The problem is that these hallucinations are increasingly difficult to detect because these chatbots hide these mistakes very well. That the system now includes unverified or validated content from networks like Reddit can be problematic: community votes do not always measure how truthful or useful a certain thread is. Using Reddit has its drawbacks. This platform has value precisely because it is not optimized for Google algorithms: It is chaotic and contradictory.. Sometimes there are brilliant responses from people, but other times there are completely wrong comments. When a user adds “Reddit” to their search and reads the results, they are automatically weighing which comments are useful and which are not. But that step disappears if Google extracts fragments of those discussions to include in an AI Overview. Eliminate that human filtering step and presents those answers with an authority that perhaps they should not have. Google will have much more difficulty than a human in distinguishing the comment of someone who has been working in plumbing for twenty years from that of someone who tinkers as a hobby. The shadow contract. This is not just an editorial or technological decision. In 2024 Google signed a deal worth $60 million a year with Reddit to access their data and train their models. You are not incorporating content from this social network as a public service: what you are doing is monetizing a commercial contract. Your message that you are highlighting those “original voices” is really saying that you have paid for that privileged access to Reddit content and now you are going to take advantage of that access and make it profitable. That revenue is interesting for Reddit, no doubt, but there is a problem: clicks. The Stack Overflow Precedent. There is no need to speculate much about what may happen because it has already happened. Stack Overflow is the largest technical Q&A community on the internet, but has lost most of its traffic in two years because AI companies They started collecting all those answers. to train your models and then serve them to your users directly. That caused users to stop visiting Stack Overflow and experts to stop answering questions. The quality of the new content on this network was clearly affected, and it became clear that if the AI ​​already gave you the answer without having to enter Stack Overflow, why enter? The danger for Reddit is exactly the same. Google didn’t have many alternatives. ChatGPT, Claude and Perplexity They have been capturing … Read more

Meta has signed an agreement to search for it in space

by

Back in 1941, Isaac Asimov already played with an idea that for decades sounded more like literature than infrastructure: capture solar energy in space and send her back to Earth. It was not a minor occurrence. Basically, it posed a question that today no longer belongs only to science fiction: what do we do when the energy available down here is not enough to sustain what we want to build. More than eighty years later, that question has found a new protagonist: artificial intelligence. What we have seen in recent years is a race to build AI infrastructure at enormous speed. More models, more servers, more data centers and, as a direct consequence, more need for stable electricity. Meta places the problem there: current clean sources help, but have obvious limitations when looking for continuous supply. Solar doesn’t produce at night, the wind doesn’t always blow, and the grid needs storage to turn that intermittent energy into a more reliable basis for its operations. The energy that AI is pushing beyond Earth The Meta movement arrives in the form of two agreements who attack the problem from different sides. The first is with Overview Energy, a startup with which Meta has reserved until 1 GW capacity of orbital solar power to support the company’s data center operations. The second is with Noon Energy, with whom Meta has reserved up to 1 GW/100 GWh of very long duration storage capacity. The idea is not to replace one technology with another, but to combine generation and storage to get closer to a more continuous supply. Overview Energy’s proposal is based on a premise that is simple to tell, although difficult to execute. Its satellites would be in geostationary orbit above the Earth’s equator, where sunlight is constant. From there they would capture energy and send it to existing solar installations on Earth as low-intensity near-infrared light. According to Meta, these plants would convert the beam into electricity and inject it into the grid just as they do today with direct sunlight, also during the hours in which they now remain inactive. Capture of a video about the project shared by Meta It’s a good idea to put things in perspective. The company itself places this technology in an early phase: Overview plans a orbital demonstration in 2028when your system should try to send power wirelessly from space to a solar plant on Earth for the first time. If successful, commercial delivery to the US grid could begin, at the earliest, in 2030. In between, the most difficult part remains: proving that the system works, that it scales, and that it can do so in an economic sense. Noon Energy Energy Storage System The second alliance looks at a less striking, but equally important problem: what happens when clean energy has already been generated and needs to be conserved for longer. Noon Energy works with reversible solid oxide fuel cells and carbon-based storage to offer more than 100 hours of storage, well above what Meta says lithium-ion batteries can offer today. These two alliances fit into a much broader energy strategy. Meta assures that it has already contracted more than 30 GW of clean and renewable energyand places these agreements alongside its next-generation geothermal projects with Sage Geosystems and XGS Energy, in addition to 7.7 GW of nuclear energy linked to Vistra, TerraPower, Oklo and Constellation Energy. What remains is a fairly clear snapshot of the moment: AI is not only pushing technology companies to buy more chips, it is also forcing them to look for electricity in increasingly unconventional places. Images | Xataka with Grok In Xataka | Kimi Code is eight times cheaper than Claude Code and does 75% of your work. The question is whether it is enough

Mercadona has gotten rid of its search engine and replaced it with its own. They did it in a month with Claude Code and saved 90%

by

Mercadona’s online store processes 4.4 million searches a week. Until recently, that volume was managed Algoliaa well-established search service used by companies like Sephora or LVMH. They had been with him for eight years. Now They have replaced it with their own search enginebuilt largely by José Ramón Pérez Agüera, CTO of Mercadona Tech. He has done it largely by himself, from his home, over a long weekend. This is how he told it in a successful LinkedIn post which now extends us in a video call with Xataka. “I’m going to be very honest and I know that this is going to look tacky, but it’s the truth,” says Pérez Agüera. “70% of the work (implementing the search engine, improving search quality and laying the foundation) took three days. One weekend plus an extended Monday.” The result: an 85% improvement in the quality of the ranking, the complete elimination of searches without results (previously 4% of the total) and a reduction in the monthly cost of between 9,000 and 15,000 dollars with Algolia to less than 900. That is, a saving of between 90% and 94% depending on the month. A decision that had been on hold for years The idea of ​​abandoning Algolia is not new at Mercadona Tech, it had been ruminating for a long time. The reasons are not surprising either: the search engine directly moves between 30 and 35% of the products that end up in the cart, which makes it a critical piece of business. And Algolia, like most SaaS services, has a pricing model that scales with use: as the company grows, the cost grows, with no way to stabilize it. “In the end you end up in a vendor lock-in of very critical software that is then difficult to get rid of,” explains Pérez Agüera. But Every time the team considered building something of their own, the work estimate was pushed back.. “The most optimistic vision we had, and with a much more basic version than the one we are going to release now, was five months. And it already seemed fast to me.” Then came the era of AI agents in software development. Pérez Agüera used Claude Code as the main tool and began to experiment on his own, without a formal project or assigned team. More out of curiosity than anything else. For playing. What AI did and what it didn’t The technical process combines hybrid search (by keywords and semantics) with a machine learning system that optimizes the ranking of results. AI made it possible to iterate on dozens of experiments in hours, analyze 479 MB of catalog and analytics data in days, and explore different ranking configurations by chatting with the agent instead of manually implementing them one by one. “I easily did 40 or 50 experiments in a weekend. That would have traditionally taken me weeks,” he explains. But the speed has a precise limit: the 29 technical decisions that AI did not make. Documentation generated during the experimentation process with Claude Code: the 14 parameters that Mercadona’s search engine evaluates to order results (from the popularity of a product to how well it fits semantically with what the user is looking for), its relative weight in the final ranking (popularity and semantic similarity account for two thirds of the decision) and the configuration of the machine learning model used to train it, based on click and purchase data from the last four weeks. Each of those parameters was discussed and validated with the AI ​​agent, but the final selection was made by the human team. Image provided by Mercadona Tech. The most representative was the choice of the indexing engine. Most systems, and probably any AI agent consulted, would have recommended Elasticsearch, the most widespread solution. Pérez Agüera chose Tantivy, a much smaller library written in Rust that integrates as an embedded component, without the need for a separate Java virtual machine. An impossible decision without knowledge of the Mercadona ecosystem. “The AI ​​always recommends the most generic option,” he says. “I made that decision because I have the context and the knowledge to make it.” The transfer to the team When the core of the search engine was ready, the project passed to the engineering team. What they found was not bad code, but it was ccode that did not follow Mercadona Tech’s internal standards. The architecture was hexagonal, as is the company’s style, but it used a different approach than usual. The tests existed (Pérez Agüera applied TDD during development) but some did not make sense or were missing cases. The agent had written thousands of lines of code in a few hours and reviewing them all was unfeasible. “The team’s Tech Lead took two or three days to adapt the project to our good practices,” he summarizes. “Not because the code was wrong, but because it didn’t meet our standards as a company.” In total, adding the initial phase and the launch into production, which includes load testing, infrastructure adjustment and integration into the Mercadona Online architecture; The project has taken approximately a month of work. And “two and a half people” have been in charge of it: Pérez Agüera, the Tech Lead of the Shop team and a part-time Staff Engineer for infrastructure. The original five-month estimate required five or six people. “FWe have easily done a x5 to the speed of the projectand what we have now is much more advanced than what we would have had in five months,” he says. What changes for the teams For Pérez Agüera, the search engine is one more experiment within a larger transformation that Mercadona Tech continues to process internally. The question on the table is not whether to use AI in development, but how to redesign the entire development process based on it. His diagnosis of the profiles is forceful: “AI is going to mean that fewer developers are needed and more engineers are needed. Coding loses value per se; the … Read more

Someone has taken more than 12,000 Spanish laws and converted them to source code. It is a real gem to search for legislation

by

If you have ever prepared for competitive exams and are looking for the legislation that you have to prepare for or need to consult a law for any management, you will have already realized that the Official State Gazette is a pain. (also applies to regional versions) to find out what is current and what has changed: transpositions, various PDFs, annexes and cross references that make you go crazy. You are not alone: ​​sooner or later it has happened to everyone. Until now you only had two alternatives: consult with someone who did know about the subject to clear your doubts or resort to artificial intelligence to then carefully check that nothing is left out. To the computer engineer Enrique Lopez It must have happened to him too and he took action on the matter. The project. Is called Legalize and it is in a few words a digital repository of state and regional legislation available on GitHub, as if it were a computer project. Thus, it has translated more than 12,000 regulations in force in the state (both state and regional), each one into a Markdown file with plain text on which you can search for what interests you. In addition, each of the laws are grouped in folders based on their jurisdiction. In short: one law, one file, one folder, one jurisdiction. The organization follows the standard ELI (European Legislation Identifier). As the project’s GitHub explains, all content comes from the BOE Consolidated Legislation APIthe text of the legislation is public domain. What Legalize-es provides is structure, version control and metadata. What has changed about this law. But the laws have their drafts, consolidated texts and subsequent reforms, so sometimes being clear about what is in force and what is not is an odyssey. So you added each reform as a commit, with the actual publication date. This way, even if you have no idea about laws, you can see what exactly has changed in the regulations: in red is what is deleted and in green is what is added. We see it better with an example, that of Royal Decree-Law 8/2010: Royal Decree-Law 8/2010 Why is it important. Beyond the practicality of access of this format, the true relevance is that anyone can know what has changed in a law without tricks or cardboard. It is true that the BOE is public, but it is far from friendly. On the other hand, when a law is reformed, it is easy to lose sight of previous regulations. With this format it is easy to know what has changed and when. Context. In a state like Spain where the normative production report of the CEOE for 2024 (the last one released) lists 719 regulations, being up to date with regulations that affect matters as important as taxes or retirement is an arduous task. The digitization of current legal regulations is a pending issue that this project addresses as a civic hack: using technology to simplify and clarify what the administration hinders. How it works. The core of legalize-es is the automation of legislative data through a pipeline, that is, with a “robot” that periodically monitors the BOE’s Consolidated Legislation API. The system extracts the text from the official PDF and cleans it of strange formats, leaving it in plain text. Once processed, the law is integrated into a Git version control system where each reform does not overwrite the previous one, but is saved as a new layer to allow access to the history of changes, which allows traceability. In Xataka | The “ChatGPT for lawyers” exists, it was born in Spain and has just reached a milestone: becoming a unicorn Cover | Flickr

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.