Month: March 2026

Why on earth donkeys arose has always been one of the great mysteries of natural history. Until now

by

When we think about the animals that have been accompanying humans since time immemorial, helping us in agricultural and daily tasks, surely the first candidates are horses, dogs, and even cats. Probably the donkeys (Equus africanus asinus) are a little further down the list. Until now we believed that the domestication of the donkey was an event that was repeated in different places and times in prehistory. However, the largest genetic study of these animals carried out to date revealed a different story: that of a single domestication of the donkey, which occurred about 7,000 years ago in the area around the Horn of Africa and what is now Kenya. The closest relatives of the domestic donkey, wild donkeys (Equus africanus) still live today in this African region. The common donkey is sometimes seen as a subspecies of these African donkeys or as an independent species closely related to it (in which case its “scientific” name would be Equus asinus). According to the team, led by French researchers, the donkey was domesticated in this context, and then began to spread throughout the rest of Eurasia, already as a domestic animal about 4,500 years ago, that is, about 2 and a half millennia after being domesticated. The genetic study has not only pointed out the unique origin of this species, but It has also “advanced” the date of domestication by about four centuries. The domestication of the donkey would have made sense in its spatiotemporal context. About 7,000 years ago the Sahara environment witnessed an aridification process that led the desert to expand. The donkeys they had an advantage Compared to other equids, they are more resistant to lack of water, which could have made them ideal for use as an aid in transportation or agricultural work. For their analysis, the international team of researchers analyzed samples of 207 modern donkeys from 31 countries, as well as remains of skeletons of 31 other donkeys who lived in the last 4,500 years. They also used genetic information from other equids to expand the study. The work of the researchers was published in the journal Science. The variety and the mules The study also offers us some curious stories about this animal. For example, genetic analysis of Roman-era remains found in France tells the story of a generation of giant donkeys (up to 25 centimeters larger than the average modern donkey). The Romans They didn’t raise these donkeys colossal for their direct use, but because mules (crosses between male donkeys and horse mares) were of great use to them. The Romans took advantage of an animal that combined part of the robustness of donkeys with the ability to travel long distances more typical of horses. After the fall of the Roman Empire, mules once again gave way to donkeys since economies had become more local, so it was not necessary to use them to transport large loads along the popular Roman road network. The donkey is perhaps the most maligned of domestic animals. Despite having played a key role in human development over the past four millennia, the donkey is often seen as synonymous with stupidity or clumsiness. Such is the point that the donkey has become a threatened species in places like Spain or Mexico. For better or worse, the donkey continues to form part of our cultural heritagefrom the donkey with which Sancho Panza accompanied Don Quixote to that of Friar Perico. Now, thanks to science, we know a little more about the history of what could be the least popular cousin of the equine family. In Xataka | The Iberian lynx is reconquering Spain and that is good news. The challenge now is to understand why In Xataka | Science had always believed that only humans understand geometry. Until we noticed the crows again Image | Ansgar Scheffold

which cars can circulate and which rest on March 21

by

This Saturday the Hoy No Circula Saturday scheme comes into force again, the program with which the Ministry of the Environment of Mexico City (SEDEMA) restricts the circulation of certain vehicles to contain pollution in the Valley of Mexico. Once again, those planning to use the car should carefully check the finish of its license plate and the verification hologram before hitting the road. The restrictions are not limited to the 16 municipalities of CDMX, but also extend to several suburban municipalities in the State of Mexico. The program also operates in: Atizapan of Zaragoza Coacalco de Berriozábal Cuautitlan Cuautitlán Izcalli Chalco Chicoloapan Chimalhuacan Ecatepec de Morelos Huixquilucan Ixtapaluca Peace Naucalpan de Juárez Nezahualcoyotl Nicolas Romero Tecámac Tlalnepantla de Baz Tultitlan Chalco Valley Also, remember that if your journey includes passing through any of these locations, the Saturday No Circulation Day also applies. What cars and license plates does Hoy No Circula Saturday affect? The logic of the program is to take a portion of the cars off the road to reduce emissions, but specific rules apply on Saturdays that are added to the scheme from Monday to Friday. Not all vehicle owners rest on the same weekend: the hologram, the finish of the license plate and whether Saturday corresponds to an even or odd week are the elements that determine who must leave the car parked and who can drive. It is also key to keep in mind that the Not on Saturday Circulation Today It does not operate 24 hours a day. The hours of operation are from 5:00 a.m. to 10:00 p.m., so that outside of that period—during the night and early morning—the program does not limit vehicle traffic, unless an environmental contingency or other extraordinary measure is declared that imposes additional restrictions. In the case of March 21, 2026, the calendar indicates that it is the third Saturday of the month, which is why it is considered an “odd week.” Under this configuration, vehicles with hologram 1 and plates whose ending is an odd number are those that must remain out of circulation during program hours. If your car meets that combination, you will have to keep it stored until after 10:00 p.m. On the contrary, cars with hologram 0 and 00 retain permission to circulate without restrictions within the framework of Today No Circula Saturday, while those with hologram 2 cannot circulate under any circumstances on Saturdays. In addition to the previous cases, remember that there is a group of exempt vehicles that can move without being affected by these limitations. These are: Electric cars, powered by natural gas or with hybrid technology Units placed in the name of people with disabilities All those intended for urban public transport services (including funeral services) Those used for school or passenger transportation Those assigned to public security and/or civil protection tasks Those who choose to ignore the provisions of Hoy No Circula risk a considerable financial penalty: the fine for violating the program ranges from 20 to 30 times the Measurement and Update Unit (UMA), which is equivalent to approximately 1,924.40 pesos at the minimum and up to 2,886.60 pesos at the maximum. Added to the financial burden is the possible immobilization of the vehicle and the time that will have to be spent resolving the problem before the authorities. Which cars and license plates are affected by Hoy No Circula Saturday? In conclusion, if you are going to travel by car this Saturday through CDMX or through the suburban municipalities of the State of Mexico included in the program, it is advisable to verify before starting what hologram your vehicle has, what the ending of your license plate is and if the calendar indicates an even or odd week. Today No Circula Saturday is designed to remove the most polluting cars from the streets, but at the same time it requires better organization of trips and consideration of mobility alternatives when the vehicle has to rest. Photo | Chandler Cruttenden In Xataka | The countries that pollute the most in the world, gathered in a detailed graph

We thought that 3D printing a gun was already disturbing. Now someone has gone one step further with a homemade guided “missile”

by

Talking about 3D printing is no longer just talking about prototypes or industrial environments. In recent years, this technology has been established as a tool available to enthusiasts and creators who can design and manufacture complex objects from home with relative ease. That accessibility has expanded the possibilities of use, but has also opened debates about its limits, especially when it intersects with weapons development. The precedents of 3D printed guns They have been on the table for some time, and now a new project once again pushes that debate into even more delicate terrain. The disturbing jump. What has now put the focus on this issue is a video of just five minutes in which the amateur Alisher Khojayev shows a prototype that is reminiscent, at least in its approach, of portable anti-aircraft missile systems. The project includes a launcher, a projectile and several electronic systems designed to assist in guidance. What does it teach. In practical terms, what Khojayev shows is a set divided into three parts that the creator presents as a coordinated system. The launcher acts as the base of the system, the projectile concentrates a good part of the 3D printed components, and an additional node with a camera can be incorporated to reinforce tracking. How the system is laid out. The architecture proposed by the project is based on linking several devices through a wireless network that coordinates the flow of data. The first step is to connect the launcher with a control computer via WiFi, which analyzes the information received and calculates the trajectory. In a second phase, the projectile becomes part of that network and receives instructions to adjust its orientation using moving surfaces. The system combines ESP32 microcontrollers with sensors such as GPS, barometer, compass and an inertial measurement unit to estimate variables such as speed and position. The cost data. The project is not only presented as a technical demonstration, but also as a low-cost exercise. According to the creator, the entire system can be assembled for about $96 from commercial components and 3D printed parts. That, of course, doesn’t mean that anyone can make something similar at home, not least because such a development would probably be illegal in many parts of the world. But it does leave a broader reading: 3D printing is reducing barriers and costs in a growing variety of projects. Images | Alisher Khojayev In Xataka | We thought that the war in Iran was about missiles. Until Germany has started counting them: it’s about what will happen in May without them

just completed a critical test before delivery

by

Spain is in a moment of weapons renewal. A few days ago we said that the country has been selling his old submarines for scrapa movement conditioned by something very specific: the new generation of S-80 submarines is beginning to gain traction. Or impulse, rather. And the ignition of engines S-82 Narciso Monturiol It is one more step towards delivery to the Navy. And it has cost… Milestone. It was last February 28 when S-82, the second submarine of the S-80 Plus series, started its diesel engines for the first time. The Ministry of Defense has become again echo of the achievement achieved a few weeks ago, and it is no wonder. The ignition is one more step in the commissioning of the submarine after the electrical ignition, the boarding of the batteries and the launching. Overcoming these phases brings the ship closer to its delivery. The S-82. The Narciso Monturiol measures about 80 meters in length, has about seven meters in diameter and is capable of moving about 3,000 tons while submerged. It has capacity for more than 30 crew members and can last submerged for three weeks. Regarding weapons, account with a system developed by Navantia and Lockheed Martin that allows launching ground attack missilesbeing the only of the European Union’s conventional submarines that have that capability. It can also arm DM2A4 heavy torpedoes with a range of up to 50 kilometers. Issues. And if the ignition of the engines is announced as a milestone, it is because the S-80 program has not had it easy. These are the first submarines of Spanish design, and since 2004 they have been suffering delays. The first in the class, the S-81 Isaac Peralshould have been delivered in the early 2010s, but design errors and a budget that went from 1.8 billion euros to 3,900, ended up being delivered in 2023. The biggest problem was weight, something crucial in a submarine. It had between 70 and 100 tons of overweightsomething that was discovered just when it was due to be delivered and that compromised the entire project because it is something that would cause buoyancy problems. There were also other problems in the propulsion system and disputes between the Spanish Navantia and the French Naval Group. Next steps. And, again, completing these tests is the step prior to delivery at some point this year. It also shows that there is someone doing the calculations somewhat regularly because the S-82 was scheduled to be delivered at the end of 2024, generating a delay in the plans. This delay is what caused the very veteran S-71 to have to be revised so that Spain could keep two submarines operational at the same time, giving itself the situation of having a new generation one alongside the aforementioned S-71. The next steps for the S-82 are to continue testing systems before delivery, and then the S-83 and S-84 will arrive, which were dated for 2026 and 2028 respectively and will arrive with a novelty: the BEST AIP system to improve immersion capability and long duration missions. Both the S-81 and S-82 will also be upgraded with the BEST AIP at some technical stop between 2029 and 3031, respectively. After that ignition of the S-82 engines, they start the most critical tests, such as those of acoustic signature and immersion, key steps before it can be launched into the sea to definitively retire the S-71. Photos | Ministry of Defense In Xataka | In the 19th century, Spain designed the first “torpedo launcher” to defend itself against the United States: the Sanjurjo Badía submarine

They are not more taxes, they are updating them in the new digital economy

by

On March 12, it was published in the Official State Gazette the resolution which approves the general guidelines of the Annual Tax and Customs Control Plan of 2026. The text does not announce any new taxes, any additional fees and any direct obligations for the average citizen. What it does do is much more relevant for the fiscal control of Spaniards: the Treasury is updating the digital economy that citizens were already using. The Tax Agency has expanded the amount of information it receives on how money moves in the digital economy: mobile payments, neobank accounts and Fintech platforms or sales on second-hand platforms. The difference is not what you will pay in taxes, but how much the treasury now knows about your income and the new ways to receive it. ​New financial information. The most important novelty of Plan 2026 is not regulatory, but informative, and has raised a lot of controversy even before coming into force. Starting this year, the Tax Agency will have information every month on the ownership of bank accounts and also on the income obtained by businessmen and professionals who use any collection management system through POS cards and payments associated with mobile phone numbers. This directly includes payments through Bizum. That is, the Treasury now equates Bizum to credit cards. What has changed is which platforms are now required to report to the Treasury. Previously, this duty fell mainly on traditional banking entities. But the economy it is becoming more digital and neobanking, Fintech and digital payment systems platforms have increasingly become most common among Spaniards. Therefore, the Treasury has now extended that obligation also to companies in the new digital economy. All of them must periodically report data on certain professional profiles or important financial movements, which gives the Treasury a much more complete and updated view of money flows that were previously off its radar. Neobanks under the radar. The Plan dedicates a specific section to controlling the activity of digital financial entities, or neobanks. The BOE text recognizes that these entities, which offer services through technological platforms in many cases without physical presence in Spain, have transformed the banking landscape and pose challenges for tax control. The concrete response of the plan is that the checks will be focused on those taxpayers in which improper use of neobank accounts is detected to hide income or assets abroad. The objective declared by the Treasury is to improve the traceability of operations in a digital and cross-border environment, without implying any direct restriction or penalty for the user for the simple fact of having an account in one of these neobanks such as Revolut. or N26. Wallapop and the DAC7: selling online leaves a tax trail. Electronic commerce and sales platforms are another priority objective of the 2026 Annual Tax and Customs Control Plan. The BOE itself states that the operations carried out through these platforms sale of items between individualslike Wallapop or Vinted, have grown by double digits between 2020 and 2025, and that its trading volumes have doubled in that period. This growth is precisely what justifies the Treasury pay more attention to what happens in them. The tool that makes this control possible is the European DAC7 directivewhich forces platforms like Wallapop to communicate to the Treasury the income that their sellers obtain from selling products, providing services or renting goods through these apps. With that information, the treasury can compare the income that a person declares with what the platform has reported, and open an investigation if the numbers do not add up. ​Individual or professional: the distinction that changes everything. The new Treasury measures are aimed at controlling fraud by obtaining more information from the platforms that move money. That does not mean that the platform must send information about all its users. You should only do it from those to use them professionally or who, although they are not professionals, have a very intensive activity (hidden economy). If you use Wallapop to sell your bicycle or Vinted to get rid of clothes that you no longer wear, this measure does not affect you. However, if you have a clothing store and sell through this platform without declaring income, the Treasury will knock on your door. The same happens with payment platforms. If you use Bizum to pay your share of dinner to a friend (or have it paid for you), Treasury will not receive information about these movements between individuals. On the other hand, for self-employed professionals and companies the scenario has changed. The Tax Agency will receive the data on what they charge by card or Bizum. This allows you to compare that income with what you declare in the quarterly VAT more efficiently and quickly than before. In Xataka | Income Simulator 2025: how to use it to know if the declaration you make in 2026 will pay or return Image | Revolut

OpenClaw is the AI ​​agent that is blowing the AI ​​industry’s mind. We have tested it: Crossover 1×42

by

ChatGPT and Claude are great, but they only do things when you ask them to. OpenClaw It’s something else. It is an AI agent that takes advantage of the power of ChatGPT or Claude (or other models) He becomes your personal employee and does everything you ask of himbut in an autonomous and proactive way. This is something that the industry has been promising for years, and although some steps had already been taken in that direction with AIs that, for example, can reserve a table for you in a restaurant, OpenClaw goes a little further because you basically “give them the keys to the office”. So, when you install it on a machine (or a VPS, or a Raspberry Pi, or a Docker container, or wherever you want) you give this AI agent superpowers, because it will be able to do everything it wants on that machine. You will be able to use all the apps you have, browser included, and use all those tools to do things for you. It is, we insist, like having an employee who works for you 24 hours a day and who, if you don’t want to, will never rest. The concept is super powerful, but of course it has some buts. The most important one is security risks, and in this episode we talk about how to protect yourself so that that virtual employee doesn’t end up messing you up and causing chaos. We also have to talk about costs, because this AI agent is a true “token glutton” and you will have to be practical when choosing which models you want to use it with. We talk about all that and many more things in this episode Crossover 1×42, which serves as an introduction to a fascinating topic. Be careful, this is addictive. On YouTube | Crossover In Xataka | OpenClaw changed the rules of the AI ​​race. Technology companies already have their answer: copy it

The model challenges benchmarks in a key area

by

When we think of Xiaomi, it is normal that its mobile phones come to mind or, at most, its foray into electric cars with models like the SU7. However, what we have seen now points to a much more ambitious move: the company also wants to compete in the artificial intelligence race. It has done so with the launch of MiMo-V2-Proa model that, according to the data shared by the company itself, seeks to position itself close to the most advanced systems on the market, but with a very different focus on costs. And that changes the conversation quite a bit. What Xiaomi proposes. The company presents its model as the “brain” of systems capable of executing complete tasks, not just responding to specific requests, which in the sector is known as agent-oriented models. According to official information, we are looking at an architecture that exceeds one trillion total parameters, although it only activates 42 billion in each execution, and that can work with contexts of up to one million tokens. On paper, this allows you to maintain long, complex processes without fragmenting them, something designed for large tasks and more demanding workflows. Performance against the greats. If we look at the data, Xiaomi does not present its model as the best on the market, but as one that can compete in certain scenarios. In the GDPval-AA benchmark, oriented to real agent-type tasks, it reaches an Elo of 1426, surpassing Chinese models such as GLM-5 (1412) and Kimi K2.5 (1309), although it falls short of proposals such as Claude Sonnet 4.6, which marks 1633. The external reading is provided by Artificial Analysis, which assigns it a score of 49 on its intelligence index, which places it in the group of most competitive models on the market. The key is in that closeness in some benchmarks, not in general leadership. The key to the price. This is where Xiaomi’s proposal changes the board. According to data collected by Artificial Analysis, running your IQ with this model costs approximately $348, compared to $2,304 for GPT-5.2 or the 2,486 of Claude Opus 4.6. It is not exactly the same comparison as the price per API use, but on both levels Xiaomi appears clearly below several Western rivals. In its own API, the company sets prices of $1 per million tokens for entry and $3 for exit in the range up to 256K, a lower rate than models such as Claude Sonnet 4.6 and Claude Opus 4.6 at the same level of use. Beyond chat. What Xiaomi is proposing with this model is not only to improve the quality of the responses, but to change the type of work it can do. The company insists on moving from conversation to action, with a system capable of using tools, interacting with environments and completing chained tasks. In this context, it presents it as a model optimized for agentic scenarios and links it to frameworks such as OpenClawin addition to mentioning collaborations with OpenCode, KiloCode, Blackbox and Cline. On paper, this reinforces the idea of ​​an AI designed to execute workflows and not just answer questions. behind the scenes. Xiaomi enters the race with a model that, according to available data, is close to the major benchmarks in some scenarios, although without generally surpassing them. Where there does seem to be a clear bet is on the price, and that is where it tries to differentiate itself. The question is whether this balance between cost and performance is maintained outside of benchmarks and in real environments. We will have to wait to know if what the data shows is also projected in the real world. Images | Xiaomi In Xataka | China has immediately understood the future of the technology industry: “one-person companies” powered by AI

The rain has transformed the driest desert on the planet into a sea of ​​flowers. It’s a sight to behold and a problem for experts

by

The Atacama Desert bloomed again in spring. After the August rains, more than 200 species from the Chilean region were activated and provoked the first major flowering since 2017. The Internet was filled with impressive photos, but (beyond the hype) there is a central problem: increasingly clear signs of a destabilized climate system. What has happened? In August 2025, a storm left accumulated between 40 and 60 mm in the Chilean Atacama Region. Specifically in the south: in Huasco, Freirina, Vallenar and the Llanos de Challe National Park. As a consequence, flowering started in the third week of September and reached its peak between the end of September and mid-October. He show was amazing: a mantle of red and yellow añañucas, of sighs, of huilles, of guanaco legs and lion’s claws. And why are we talking about this now? It’s a good question. Historically desert blooms occurred between 5 and 7 years. Typically linked to El Niño phenomena. In the last 40 years, Chile has recorded about 15 superblooms. The striking thing about this case (as happened in 2022 and 2025) is that it is linked to La Niña conditions. And, indeed, one may be a coincidence, but three so close together mark a trend. And the problem is that more blooms are not always good news. And so? As explained Maria Fernanda Pérezan ecologist at the PUC of Chile, out-of-season blooms generate a gap between flowering and pollinators. What’s the point of having pollen if we don’t have bees to do their job? Indeed: absolutely nothing. What’s more, if climate change causes this type of blooms on a regular basis, this deregulation could cause very serious problems. After all, just think that a guanaco paw seed can spend fifteen years on the desert floor until its time comes; If it germinates and there is no one to pollinate it, there will not be another seed. Climate change is going to cause us more problems than we are able to imagine. Because the serious thing is not the sea level, the melting of the glaciers or the rise in temperatures (that too). The most important thing is these little things that change everything. Things so small that we haven’t thought about them. Image | In Xataka | The Atacama Desert is one of the driest places on the planet. And right there a bunch of “crazies” are trying to get water out of the fog.

a Spanish company is the key piece

by

Europe has embarked on the adventure of technological sovereignty. It is pointing to several fields at the same time, being the space sovereignty one of them. Pursuing this objective, the European Defense Agency -EDA- has just awarded a research contract to an aerospace consortium with the aim of creating a military satellite Optimized for very low Earth orbit. And the Spanish Sener will be the one to lead that space A-Team. In short. The EDA contract is for 15.65 million euros and the objective is as mentioned: to create the first European military satellite concept especially used for VLEO space. Spain, France, Luxembourg, Portugal and Slovenia are the countries that are financing the project baptized as VLEO-DEF, and the Spanish Sener will have the task of leading 16 other companies belonging to those five countries. This is not the first time we have talked about Sener Aeroespacial. It is the subsidiary of the SENER group and is one of the Spanish companies who participates in the ambitious rearmament plan of the European Union. It has more than 4,000 employees and its experience covers space, guidance, control and unmanned systems. Very low Earth orbit. Before seeing what the satellite will do, let’s see what very low Earth orbit is. Call too VLEOis the orbital strip that is between 150 and 400 km altitude. It is the lower end of low orbit and, although it may not seem like it, it is actually very close to the Earth’s surface. This brings key benefits such as the ability to capture images with much greater detail, a better signal-to-noise ratio in optical and radio frequency sensors and, above all, very low latency. After all, it is closer than other satellites and the signal must travel a shorter distance. However, it is not a comfortable strip. The atmosphere at that height generates very intense friction and there is an aggressive chemical environment. This implies that the satellites are not “floating”, but rather require almost continuous propulsion. And, in addition, the materials must be very resistant to resist corrosion and, basically, not disintegrate after a short time. VLEO‑DEF. And the idea, precisely, is that. The consortium must find a way to develop a military satellite specifically designed to operate at around 250-350 kilometers altitude in a sustainable manner. The duration of the project will be 36 months and the 17 companies will have to find the key to the technologies that allow the future construction of satellites to operate in VLEO. Because, although this field is very interesting for scientific and observation research, in the military spectrum, flying at that distance from the Earth seems very interesting to achieve what we have mentioned: a much clearer and more detailed observation of the territory. And it is important because we constantly see that they “keep an eye” on what neighboring countries are doing, which has allowed us to know some Chinese operations or the North Korean military ship disaster. Sovereignty. If the program comes to fruition, such an observation satellite can provide key data in intelligence, surveillance and reconnaissance missions by being capable of offering much faster communication between the satellite and military commands. With VLEO-DEF, the ultimate goal is to pave the way for future VLEO satellite constellations for border security, protection and intelligence, all within the aforementioned sovereignty. The Ukrainian War and the gas cut by Russia, the case of Greenland with the United States and blackmail of the American president have awakened in the EU that idea that they should start to fend for themselves in fields where they previously delegated to the allies. That is why rearmament began, but also the search for energy alternatives, rare earth, defense programs with European AI and cconstruction of data centers and semiconductor factories. And in all these programs, Spain is emerging as a key partner with space programs, chip development, renewable matrix and with projects for data centers. In Xataka | “Elon Musk can monopolize everything,” warns Arianespace, which has been launching all of Europe’s satellites for 40 years

OpenAI’s big problem all these years has been a chronic lack of definition. Now he wants to solve it with a super app

by

OpenAI spent much of 2025 announcing new features, not new models (that also), but new products. We saw him with his Sora 2 video generator or with ChatGPT Atlas browser. Now, the company recognizes that they were diversifying too much and their plan is… to launch another app. The super app. They have an exclusive Wall Street Journal that OpenAI is preparing a desktop tool that will unify the ChatGPT app, its Codex code platform and the Atlas browser. This super app will offer agentic capabilities, not only oriented to code, but also to productivity. This is aiming directly at the business field, a field in which its rival, Anthropic is quite ahead of him. Too many products. The company’s goal with this move is to simplify the experience and reduce fragmentation between products. Speaking to the Wall Street Journal, a company spokesperson assures that it will allow them to unify the different teams, which will be able to focus their efforts on one product instead of several. In an internal note, OpenAI explicitly acknowledges that they were spreading their efforts across too many apps and needed to simplify them. The change will be led by Fidji Simo, the head of apps at OpenAI, who recently brought the employees together to give them a message: “We cannot waste this moment because we are distracted by parallel projects.” And diversifying consumes many resources, both economic and computing capacity, and OpenAI is not to be wasted none of them. Without direction. OpenAI has the most used chatbot in the world, but what they don’t have is a clear product strategy. They have wanted to be too many things at once without a clear strategyand in addition, half-abandoned products have been left along the way. The Atlas browser is the best example of this. I had all the potential to be a serious alternative to Chrome which had not yet integrated Gemini. The reality is that, five months after its launch, ChatGPT Atlas is still exclusive for Mac and also has lost functions. Something similar happened with Sora 2: they got the viral moment they were looking for, but today the app remains exclusive for users in the US and Canada. Competition where it hurts most. While OpenAI launched its video memes or its browser, the competition moved forward with a much less flashy, but better thought-out plan. According to a Menlo Ventures reportin 2023 OpenAI had a 50% share in the enterprise segment, while Anthropic had only 12. In 2025 the tables turned: Anthropic had 32% and ChatGPT 25%. If we focus only on programmers, 42% prefer Claude and only 21% ChatGPT. ChatGPT still has many more users, but the vast majority are for personal use. Financially, business users are much more valuable because they have no qualms about paying for subscriptions that often exceed $200 per month. Image crisis. In case Anthropic was not eating enough toast, the image crisis caused by the agreement with the Pentagon. ChatGPT began to lose users at a worrying ratewhile Claude was placed in the top of most downloaded applications. What they were missing. Image | Amparo Babiloni, Xataka In Xataka | There was a time when ChatGPT was a magical and free tool. That time is about to end

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.