called

It is called Anthropic and it is going to pay you 200,000 million, according to The Information

by

Anthropic has agreed to pay Google about $200 billion over five years for more computing power, according to has published The Information. The figure would thus place the AI ​​startup as Google Cloud’s largest individual client, representing more than 40% of the backlog of earnings that Alphabet communicated to its investors last week. From commitment to commitment. A revenue backlog reflects contractual commitments already signed by a cloud provider’s customers. That Anthropic occupies more than 40% of Google Cloud says a lot about the extent to which the startup has become a structural piece of Alphabet’s business. There is also another nuance to highlight: that large AI companies like Anthropic or OpenAI still need the hyperscalers to continue growing, so in this sense, both Microsoft and Google can afford not to have the best AI models as long as they receive such an amount of income from offering such computing capacity. What the agreement consists of. According to they count In The Information, the pact, signed in April, includes massive capacity of TPUs (Google’s own AI chips), supplied in collaboration with Broadcom. However, this infrastructure will not be ready after 2027. Anthropic, for its part, not only works with Google hardware, since also uses Trainium chips from Amazon and Nvidia GPUs, playing its cards well to diversify suppliers and not depend on a single company that supplies computing capacity. The now classic circular financing. Alphabet has been investing in Anthropic for years: first it was $300 million in 2023, then another 2 billionafter 1 billion more in 2025. A few days ago we also discovered an investment of up to 40,000 million additional payments by Google, of which 10 billion would be disbursed immediately and the rest would be conditional on objectives met. In exchange, Google Cloud will provide an additional 5 gigawatts of computing capacity. This way, Google invests in Anthropic and Anthropic spends that money in Google. Is called circular financingand it is the key to how the foundations of AI are made of promises. According to account In the middle, the contracts signed between large cloud providers (Amazon Web Services, Microsoft Azure, Google Cloud) and startups like Anthropic and OpenAI already add up to more than two billion dollars in committed backlogs. Hyperscalers invest in AI startups and AI startups spend that money on the infrastructure of those same hyperscalers. Anthropic can’t afford it… and yet they do it. Estimates suggest that Anthropic’s server costs could reach 20 billion dollars only in 2026. The company is not yet profitable, but demand for its model family Claude continues to grow strongly in the business segment, which forces it to secure long-term computing capacity before infrastructure shortages prevent it from doing so. The agreement with Google adds to another recent one with CoreWeave and the forecast of securing almost a gigawatt of additional capacity through Amazon chips before the end of the year. Almost symbiotic relationship. Alphabet is at a time of maximum competitive pressure in AI. Your cloud business grew by 36% last year, and Anthropic is one of its most intensive clients. Losing that relationship, or seeing it migrate to other providers like AWS, would be a significant blow. Furthermore, with an Anthropic valuation that Bloomberg situates around 800,000 million dollars, and with a possible IPO Before the year is out, Google’s accumulated stake in the company could become one of its most valuable financial assets. It is not just infrastructure: it is also a capital bet. Cover image | Wikimedia and Fortune Brainstorm Tech In Xataka | If at some point NVIDIA has to choose between giving its best chips to the US or China, its choice is very clear.

Freepik is now called Magnific. And the name change is the least of it

by

Freepik has rebranded itself as Magnific. The Malaga company, founded in 2010 as a search engine for graphic resources, has decided to adopt the name of the Murcian startup that it acquired in May 2024 and reorganize its entire generative AI offering for creatives under that brand. The move comes accompanied by figures that explain why it is worth taking the step: $200 million in annual recurring revenue, more than one million paying subscribers and 250 business teams using the platform, including those from BBC, DeliveryHero, Huel, R/GA, Damm and Job&Talent. Why is it important. Few European companies can stand up to the wave of American creative platforms (such as Midjourney, Runway or Leonardo) without having raised a single round of venture capital in the United States. Freepik, now Magnific, is one of them. And it is doing it from Malaga, with a different model from the rest: instead of competing for the best image or video model, it aggregates the leading models on the market and integrates them into a single professional production environment. It is a commitment to being the layer that unites, not the one that generates. The context. It is worth remembering where this story comes from. Freepik had been there for years stealthily becoming one of the most relevant players in the global graphics sector: in 2020 EQT bought the business in one of the largest Spanish technological operations, and since then the company has chained acquisitions (Iconfinder, Videvo, EyeEm…) and a turn towards generative AI. The purchase of Magnific in May 2024 It was the turning point. Magnific was then a five-month-old startup founded by Javi López and Emilio Nicolás that had popularized the concept of reimagined upscaling: enlarge images generating new details in the process. The operation was carried out without the two brands merging. Until today. Magnific Spaces interface. Image provided. Between the lines. That the resulting company adopts the name of the acquired company and not that of the parent company says something: Freepik clearly carried a perception of a bank of images of stocksa business perceived as conservative and little linked to AI, to novelty. Magnific, on the other hand, had less of a brand, but was synonymous with cutting-edge AI and a tool admired by the international creative community, even commented by Elon Musk a few weeks after its launch. Adopting the Magnific name is, above all, a positioning move: the company does not want to continue to be associated with vectors and templates, but with AI-assisted audiovisual production. It’s a rebranding to where the future money is, not where your legacy is. In figures. The data that the company has shared outlines an unusual trajectory in European AI: $200 million ARR (annual recurring sales). 1 million paying subscribers. 100 million monthly visits. 175 million images and videos generated per month. 250 business teams in production. 2,000 subscriptions to the Business plan in its first six weeks, with a current rate of 150 new devices per week. Andreessen Horowitz ranks it as the largest European generative AI web company by number of users. In detail. What is offered under the Magnific umbrella covers the complete visual production cycle: 4K image and video generation with audio, upscaling own, collaborative space in real time (Spaces), 3D environments, multilingual lip sync (Speak), speech synthesis, sound effects, and a legacy library of 250 million assets. The business promise is not to have the best model in a category, but rather that a creative team can do all the work without jumping between five different tools. He’s not doing bad at all with that proposal. And now, with the unified brand and the financial muscle to accelerate, it is time to convince the market that this promise also applies to the giants that come after it. In Xataka | Freepik, winner of the special Xataka Award for the best Spanish technology company of 2025: from image bank to Adobe rival Featured image | Magnificent

It’s called “Acubi”, K-Pop stars wear it and Zara is already taking note

by

The long-awaited return of the third season of Euphoria to HBO screens this April seemed to dictate a clear sentence about the future of fashion and beauty: excess is back, but in a darker and more intentional version. The series has abandoned the ethereal, innocent aesthetic of its beginnings to embrace high-contrast makeup, deep ’90s inspiration, and “fierce, unapologetic glam.” As Donni Davy saysthe main makeup artist of the fiction, this new season is “a campaign against clean-girl makeup.” However, while Hollywood pushes hard towards this drama and visual aggressiveness, there is a much quieter rebellion in the streets and on global social networks. An entire generation has decided to turn its back on the aesthetic chaos of the West to embrace Acubi: the quiet cool South Korean style with muted tones and loose silhouettes that is redefining the youth wardrobe. A trend born on the internet. The Acubi style is not an invention of recent weeks, but its current explosion is undeniable. The term comes from the south korean brand Acubi Cluba pioneer in mixing 2000s minimalism (Y2K), “subversive” basics and cyber fairy grunge. It is an aesthetic that avoids metallic pastel tones and takes refuge in a neutral color palette—white, black and gray—, constantly playing with proportions: tight-fitting tops or those with strategic cuts (cut-outs) combined with very loose cut pants. This fashion is gaining traction on platforms such as TikTok and Pinterest since the summer of 2022, standing out for being a creative and mature mix. The accelerator has been K-Pop. If Acubi has conquered the world, it has been thanks to its most powerful ambassadors. Heewon Yuh, youth fashion strategist at WGSN, clarify in cnn that “K-Pop functioned not so much as the origin, but as an accelerator, transforming a local style approach into a globally recognized look.” Female groups with global impact such as Blackpink, NewJeans and Aespa have taken this aesthetic to the stages and fashion weeks. In fact, Ning Ning and Winter, vocalists of Aespa, they showed off their best performance of the Acubi style, dominated by black and asymmetry, during the promotion of his album Armageddon. The data supports this aesthetic tsunami: the hashtag #Acubi generates about 65,000 daily posts on TikTok and 87,000 on Instagram. In addition, interest in Google Trends in Korean fashion reached its peak in the United Kingdom and the United States in February of this same 2026, coinciding with the appearance of idols at London Fashion Week. It is not a simple algorithmic whim. Jaana Jätyri, founder of the forecasting agency Trendstop, explains to cnn that “in periods of economic and social tension, fashion tends to soften.” The Acubi allows young people to be fashionable “without feeling on display.” Along the same lines, Rose Coffey, analyst at The Future Laboratory, maintains that new generations seek “stability and a sense of control” through modular and adaptable garments. However, this search for stability through neutrality does not convince everyone, and has a deeper and more controversial reading. The general obsession with derived aesthetics such as clean girl proposes a “neutral, apolitical and universalized” image. According to Marta De la Rochaan expert from the European University of Madrid, “we have lost the political and identity message that more striking urban tribes previously had.” This is where the analysis of journalist Noemí López Trujillo in Newtral provides a fundamental critical layer. Based on the test Reaction by Susan Faludi, details that the rise of aesthetics that require women to be discreet and ultra-clean can be read as a conservative and anti-feminist reaction. The goal, as journalist Brenda Otero explains, is that women “do not make mistakes, that they all appear equal, that there is no chaos, that they are static and do not change.” Fortunately, although the Acubi shares that silent color palette, its heritage grunge and its asymmetrical cuts save it from falling into the total and apolitical submission of the clean girlgiving it a subversive edge against the desire for women to go unnoticed. From digital niche to universal basic. Far from being a passing fad, the business model behind Acubi predicts a long life. The trend has already made the leap from screens to cash registers. Retailers of fast-fashion such as Shein and British brands such as Minga London already market these garments. Furthermore, mainstream market giants (high street) like Zara and COS have begun to incorporate similar loose silhouettes in their recent collections. Even haute couture has taken note: Gucci’s Cruise 2025 collection and Fendi’s recent catwalks They have presented minimalist designs relaxed people who drink from this same fountain. This success is also a victory for Soft Power South Korean. Professor Dal Yong Jin, from Simon Fraser University, explains in cnn that the growing visibility of Seoul aesthetics is a reflection of the expansion of the Hallyu (the Korean wave). Consuming Korean fashion has become deeply attractive to the international public, strengthening the economy and institutional image of the Asian country globally. The key lies in its extreme practicality. As illustrated in The Straits Times When analyzing urban youth in Singapore, this aesthetic triumphs in all types of contexts – even in hot and humid climates – thanks to the fact that it allows you to show off a well-groomed appearance effortlessly. It is a style that is built on interchangeable and breathable layers: a basic tank top, loose cargo pants and the finishing touch of some metallic accessories or sunglasses are enough to complete the look. Furthermore, although it was initially promoted by women, Acubi has broken gender barriers. Men’s fashion has embraced this trend under an influence Techwear. The “tiny top and big pants” formula translates to men wearing tight shirts or ripped sweaters paired with extra-wide parachute cargo pants, a line that retailers like Lewkin already carry under the “Acubi Men” label. The silent noise of a generation The current fashion landscape draws an interesting dichotomy. While giant Western productions try to impose dramatic or high-contrast aesthetics, global youth has chosen to … Read more

Millionaires are fleeing the Middle East. And their unexpected destination is a small Swiss canton called Zug.

by

In 2011, during the Arab Spring, several European private banks detected an unusual phenomenon: Within days, high-net-worth clients began transferring large sums from the Middle East into accounts in Switzerland without prior notice. It wasn’t the first time something like this happened, but it was one of the fastest. That left a clear lesson in the financial sector: when stability falters, money does not wait to understand what happens, it simply moves. War moves money. we have been counting. The war in the Middle East is not only altering military and energy balances, it is also causing a silent movement but massive capital. What were previously fiscal decisions or lifestyle They have become urgent security decisions, where the priority is no longer optimizing profits, but protecting assets. In this context, an idea begins to prevail: billionaires do not wait for the situation to get worse, they go aheadand that movement is redrawing the global map of wealth in real time. Dubai is no longer an unquestionable refuge. For years, Dubai was the natural destination for international fortunes seeking stability, tax benefits and a secure environment in a complex region. However, the conflict with Iran has introduced a variable that previously seemed controlled: the direct risk. That perception has been enough for activate discrete outputs but constant numbers of businessmen, executives and large assets who are now looking for more predictable alternatives outside the gulf. This is not a collapse, but a change in mentality: when security is no longer absolute, attractiveness quickly erodes. Aerial view of Zug And, suddenly, Zug. In this displacement, the place that is attracting attention is not a great global capital, but a small swiss canton of just 135,000 inhabitants: Zug. Traditionally known for its role in commodities trading and, more recently, in crypto ecosystemhas become the first destination that many of these capitals look to. Reasons? counted the financial times that both wealth managers and bankers agree that demand has grown significantly since the beginning of the conflict, to the point that for many clients the request is direct and automatic: move there. The call effect. This growing flow is having immediate consequences in an already limited market, especially when it comes to housing. Demand has rapidly outstripped supply, generating intense competition for any property available and lines even for modest rentals. Added to this are administrative barriers that make entry difficult, especially for those who do not belong to the European Union, forcing residence to be linked to employment, investment or specific tax agreements. Zug attractsbut it does not absorb without friction. Switzerland reinforces its role in the geopolitics of money. What is happening in Zug is not an isolated phenomenon, but rather part of a broader dynamic in which Switzerland consolidates again as a refuge in times of uncertainty. Its political stability, its legal framework and its financial tradition make it a almost automatic destiny when overall risk increases. In fact, other cantons like Lugano have begun to capture part of this growing demand, expanding the phenomenon and confirming that the movement has only just begun. A map of wealth that changes with each conflict. In short, the result is a progressive movement of money from risk areas to safe enclaves, where each crisis acts as a catalyst. The war in the Middle East is accelerating this process and leaving one conclusion abundantly clear: global fortunes are no longer driven only by opportunity, but for threats. And in that new balance, places so small and discreet like Zug They can become, almost without noise, the great beneficiaries of an increasingly unstable world. Image | Schulerst , IDF Spokesperson’s Unit, LohriPR In Xataka | The most buoyant market right now is selling streaming and satellite images of US movements to Iran. In Xataka | Commercial aviation is based on very old aircraft. The Iran war is going to make it even worse

Someone has created the first complete advanced malware by vibecoding with AI. It’s called Voidlink and it leaves an important question

by

For a long time, develop malware advanced seemed reserved for actors with experience, time and considerable technical capacity, especially in an environment in which operating systems and many platforms have been tightening their defenses. But the table is changing. What we have seen in recent years is that artificial intelligence not only serves to summarize texts or answer questions, it can also very visibly accelerate the software creation when given precise instructions. And that leaves us facing a reality that is difficult to ignore: the same tool that simplifies legitimate tasks can also reduce part of the effort necessary to create malicious code. That change begins to take concrete form with VoidLink. In his analysisCheck Point presents it as one of the strongest evidence so far of advanced malware developed largely with the help of AI. There is, however, an important nuance in the investigation itself: the company assures that it detected it at an early stage, that it was not deployed against victims and that it was not used in active attacks. But that is precisely why the discovery is so revealing, because it allowed access to development materials that rarely come to light. How VoidLink was built and why it changes the dashboard VoidLink was not, at least on paper, a minor piece or a rudimentary experiment. The cybersecurity firm describes it as a malware framework for Linux with a modular architecture, designed to maintain stealthy and prolonged access in cloud environments. In his analysis he mentions components such as eBPF and LKM rootkits, as well as specific modules for cloud enumeration and subsequent activities in container environments. That level of maturity is just what separates it from other previous cases associated with simpler code. One of the most striking twists in the case is who seems to have been behind it. Check Point explains that, due to its internal structure and the pace of evolution observed, VoidLink gave the impression of having come from a large team, with different profiles and a fairly defined work plan. But the evidence collected by the firm points to something very different: a single actor who, according to the investigation, would have had AI support during different phases of development. There is also another relevant element: that actor would not be a rookie, but rather someone with a solid technical base and previous experience in cybersecurity. The most revealing part of the case is how the project would have been built. The firm describes a working method based on what it calls Spec Driven Development that works as follows: You define what you want to build This idea is translated into architecture, tasks, sprints and delivery criteria The implementation is delegated to the model. In the exposed materials, development plans, technical documentation, coding standards, deployment and testing guides appeared, as well as an organization by teams and phases that supports this model. One of the recovered artifacts, dated December 4, 2025, further suggests that VoidLink had already reached a functional phase in less than a week and exceeded the 88,000 lines of code. That is precisely what separates VoidLink from other precedents. Check Point maintains that this is the strongest evidence of malware created almost entirely with the help of AI. “This is the first confirmed case of advanced AI-generated malware, created with the speed, structure and sophistication of an entire engineering organization,” claims the company. The question now is how far malicious actors can go with these types of techniques. Images | Xataka with Nano Banana | Check Point In Xataka | The Booking hack is a little more disturbing: “Tracking phishing” attacks are here to stay

The developers who get the most out of AI are also the ones who sleep the least: it’s called "AI psychosis"

by

Andrej Karpathy, co-founder of OpenAI and who coined the term vibe codinghas been in what he describes as a state of “AI psychosis” since December. He works 16 hours a day directing swarms of code agents. And he admits that he feels “extremely nervous” when he has left tokens without consuming at the end of the month. This has been admitted in an interview with Sarah Guo. It is not an isolated case but rather the pattern that is beginning to repeat itself among the developers who get the most out of this type of agents. Why is it important. The dominant narrative about AI has been that of unlimited productivity and the famous “10x“What is beginning to be documented is its dark side: the most intensive users are also those who show the most worrying signs of behavioral deterioration. And they are not anecdotal profiles. Garry Tan, CEO of an entire Y Combinator, has called his own experience “cyber psychosis“. A CTO picked up by Axios says he needs prescription medication to sleep. If the most productive tools in history generate the same patterns in their most intensive users as games of chance, the debate about the impact of AI at work enters another dimension. In Xataka Having an AI on my phone that works without an Internet connection is more useful than I thought: this way you can start it Between the lines. Karpathy’s nervousness at the tokens Being left unused is the behavioral signature of someone who has internalized scarcity as a threat, exactly the same mechanism that keeps a gambler hooked on a slot machine. Developer Armin Ronacher talked about this in January: “Many of us fell into code addiction with agents. We barely slept, we built incredible things.” The context. Agents like Claude Code either Codex from OpenAI do not work like a chatbot that is asked a question. They operate autonomously for hours, writing, testing, and deploying code while the developer monitors, fixes, and re-delegates. The promise is enormous and so is the cognitive cost: the human brain is not designed to supervise processes that advance at machine speed during 16-hour days. {“videoId”:”x9f93vm”,”autoplay”:false,”title”:”Claude Code Presentation”, “tag”:””, “duration”:”234″} Yes, but. Programmers have always had a reputation for working in marathons of concentration. Sleepless nights before a launch are part of industry folklore. What distinguishes this phenomenon is its compulsive nature and its continuity: it is not the specific pressure of a deadlinebut an activation that does not turn off when the job ends, because with an agent that can keep running, the job never completely ends. In Xataka |I have lived the “miracle” of vibe coding: this is how I programmed an Android TV app without having any idea about programming Featured image | Anthropic (function() { window._JS_MODULES = window._JS_MODULES || {}; var headElement = document.getElementsByTagName(‘head’)(0); if (_JS_MODULES.instagram) { var instagramScript = document.createElement(‘script’); instagramScript.src=”https://platform.instagram.com/en_US/embeds.js”; instagramScript.async = true; instagramScript.defer = true; headElement.appendChild(instagramScript); – The news The developers who get the most out of AI are also the ones who sleep the least: it’s called “AI psychosis” was originally published in Xataka by Javier Lacort .

It is called CY8, and it is the missing piece to transport tons

by

In some of the most extreme environments on the planet, such as plateaus above 4,000 metersair density can reduce the takeoff ability of an aircraft by more than 30%forcing a complete redesign of how any cargo is transported. For a long time, that has limited what can be moved… and, importantly, where. The “air truck” of Beijing. Yes, China has just successfully tested a pilotless “air truck” in the form of a drone. It’s called CY-8 and its first flight in Zhengzhou was not only a technical test, but the confirmation of a concept: an unmanned aircraft capable of combining large load capacity with operational flexibility that until now was reserved for much more complex or infrastructure-dependent platforms. Designed to charge. The CY-8 stands out for a very clear logic: moving weight efficiently. With a maximum weight of 7 tons and a 3.5 ton loading capacitypractically equals its own useful weight, something key in logistics operations. Your closed cellar 18 cubic meterswith front and rear access, allows you to accelerate loading and unloading processes, reducing time on the ground and increasing the pace of operation in environments where every minute counts. Operate where others cannot. The local media reported that the true advantage of the system is not only in how much it transports, but in where can you do it. The CY-8 can take off in less than 500 meters and operate on basic tracks, allowing you to access high mountain areas, isolated regions or island environments. It is optimized for extreme altitudes such as those in Tibet and for territories with limited infrastructurewhich greatly expands its radius of action beyond conventional airports. Autonomy, range and versatility. With more than 3,000 km rangethe drone not only connects remote points, but does so without the need for pilots and with the ability to adapt to multiple missions. Plus: can be set up quickly for logistics transportdisaster relief, emergency communications or reconnaissance tasks, making it a hybrid platform between civil and military, designed to respond to changing scenarios. Pilotless logistics. The development of the CY-8 is part of a growing competition to dominate heavy unmanned aerial transport. China is already working in even bigger modelswhile the United States explores alternatives with vertical takeoff that completely eliminate the need for runways. In this context, the CY-8 represents a more or less intermediate bet: it does not eliminate the infrastructure, but it does reduce it to the essential minimum. More than a drone, a strategic piece. Beyond its numbers, this “monstrous” drone redefines a key idea: in conflicts or crises, it is not enough to arrive, we must be able to sustain. This type of platform makes it possible to maintain supply chains in places where it was previously complex or impossible to do so. Therefore, more than an isolated technological advance, this kind of Chinese “air truck” points to a paradigm shift: logistics, silent and constant, as the true engine of any modern operation. Image | x In Xataka | China just showed the world what comes after the combat drone: 96 drones with a science fiction launch In Xataka | For years we have associated drones with propellers: in China they explore an alternative inspired by nature

The number one enemy of the Spanish mountain is called climate change. And we have data to prove it.

by

In 2024, they burned 47,700 hectares. In 2025, 340,000 were exceeded. And honestly, the reasons are manyalmost too many. Well, Marco Turco, from the University of Murcia, just demonstrated something that we already sensed: at a global level, the days of extreme fire risk They have increased 65% since 1980. That’s 12 more days a year. And, if that were not enough, the Mediterranean region is where lthe signs are clearer. What does all this mean? In general terms, this means that although the causes of the fires remain human (in Spain between 80 and 95% of firesin fact; the intentional ones there are many less), climate change has a lot to do with its spread. Increasingly. Why is it interesting? Because this study is the first to apply formal climate fingerprinting techniques on a global scale to fire risk. That is, that figure of 11.66 more days of extreme risk in 44 years is achieved with the most advanced methodology that we have at our disposal. And if the global data is bad, the Mediterranean data (where the days have doubled in these almost five decades) they are horrifying. But it’s not all bad news. After all, as Turco points outdespite the increase in risk, the burned area has not increased proportionally. And the reason, according to him, is the improvement of the means of extinction. However, “when extreme conditions coincide with ignition, the resulting fires are more virulent and extensive.” Why is it news now? Besides because the article has just been published in Science Advances, because the precedent of 2025 (a rainy spring and a terrible summer) It resonates a lot with what we have in 2026. We don’t even have to remember that we are talking about a handful of months with truly incredible accumulated rainfall and that is generating an amount of material in the field that can easily be end up turning Spain black. Because the core of Turco’s work is that the conditions that allow fire to spread and become a big fire They are stronger than ever. Furthermore, human exposure to these types of fires is increasing: according to recent work in Cataloniabetween 42 and 138% for each area burned since 1992. The great debate of the future. As we have repeated on several occasions, there is no debate about the effect of climate change on increasing the risk of fire. The work is summarized in how much, how and where. Therefore, the central debate is another: what. What we do with the cards that nature is dealing us. And the truth is that there is a lot to cut: whether to bet on extinction or preventionif investing more in the landscape management or begin to integrate the entire territory into urban planning schemes more ambitious and extensive. Etc, etc, etc. The debate is endless and we are always late. Because what is clear thanks to Turco is that the distance that separates the spark from the megafire is increasingly shorter. Image | Mikhail Serdyukov In Xataka | In Ourense there are towns that fear running out of water in the middle of the rainy season. The reason: the hangover from forest fires

It’s called MACROHARD and “no one else can do it”

by

Elon Musk is determined to unify his empire. After the purchase of xAI by SpaceXhas now announced a joint project between Tesla and its artificial intelligence company. It is about the AI agent MACROHARD, and before going into details, let’s go with what is most important for Musk himself: it is a mockery of Microsoft and software that wants to emulate the functioning of entire companies. Also something that no other company can do right now. According to Elon, of course. MACROHARD. If Microsoft is “small soft”, MACROHARD is “huge hard”. Elon Musk with his usual things, but beyond the joke, the project is an umbrella that unites the capabilities of Grok and those of Digital Optimus. Grok is the conductor, xAI’s large language model that calls the shots thanks to its “deep understanding of the world.” Digital Optimus is the executing “arm”, the software that manages the execution of a task in real time. From what Musk has stated, it is a vitamin agent, something in which American industryand Chinais putting all his efforts into it. Also with the General Artificial Intelligencebut that is a different story. like a brain. The tycoon has compared the operation of the agent with the theory of dual processing of the brain. While Grok is the one who sets the reason, the pause and the intention, Digital Optimus is the one who carries out the instinctive action, whatever that means. If we land it better, Grok has the knowledge Being an LLM and Digital Optimus is the one that executes the action. As? Well, being able to “process the last five seconds of what appears on a computer screen and execute actions with the keyboard and mouse in real time. If you want to know more about how it works, you will have to wait because in Musk’s message there is not much more to say about this. Basically, he has described an agent like those in which the competitors are already working from xAI, but Musk claims it can “simulate all of a company’s operations.” “It’s a funny reference to Microsoft” – Elon Musk AI4 chip. The most interesting thing about MACROHARD is that part of “hardware” that can also give rise to the name. Tesla and xAI have been working on their own chips for some time that allow them to better train AI and more optimal inference. After abandoning its chip project a few months ago, this 2026 Tesla rescued it and noted that they were making enormous progress. Maybe not yet to train the model with heavy load, something that NVIDIA chips largely take care of (because NVIDIA is involved in all AI companies, big or small), but yes for that inference task. And that’s where this Tesla and xAI project has a key advantage. The agent is designed to run on Tesla’s AI4 chip along with “relatively moderate use of the NVIDIA hardware that already has xAI and it is much more expensive.” The key is that the AI4 is much cheaper than NVIDIA’s solutions and is also the ‘home-made’ solution, allowing Tesla and xAI to scale the system as far as they want. Provided that TSMCwhich is the company that manufactures them, has bandwidth, we already know who the two main clients are in case of a bottleneck. Revived project. And it is also curious that, just like the own chip projectMACROHARD has been rescued from the underworld he was in. Days before Musk revealed the project, the media Business Insider He noted that the agent had stalled due to factors such as a hiring freeze and the departure of key engineers such as Toby Pohlen, a founding member of xAI who was going to lead the project, but who left the company two weeks later. Perhaps with that part of it being a system powered by its own and cheap chips, what Musk says that “no other company can do it yet” is fulfilled, but it will not be because they are not trying. Outside the long shadow of NVIDIA, OpenAI continues to develop own chips from Broadcom. And Goal, after some lurchesthere is also preparing chips for inference. Photo by Planet Labs Now all that remains is to see less talk and something tangible about MACROHARD’s capabilities and whether, being under the Tesla umbrella, the project comes to fruition. And one last curiosity: MACROHARD is like Elon christening he Colossus 2 data center…with which he already laughed at Microsoft. Images | sherwood, Gage Skidmore In Xataka | Elon Musk’s Grokipedia is not exactly the best place to get objective information. ChatGPT doesn’t care

A startup from Malaga is the most used European AI app in the world according to Andreessen Horowitz. It’s called Freepik

by

The venture capital firm Andreessen Horowitz has prepared its already traditional ranking with the world’s top 100 end-user AI applications. There are many predictable ones in the top positions, but we are surprised because among the top 15 is none other than freepikthe platform created by the Malaga startup of the same name. Freepik in the world top. On the list we have many usual suspects (and some not so usual) in the top positions, but one of the big surprises on the list It is the Freepik platformwhich is ranked number 11 and is the only representative of our country in that ranking. But besides that, it is the first of all Europeans included on the list. This specific list is made with the number of unique monthly visits as a criterion. USA dominates. In that list the dominance of apps from American companies is clear, and only the Chinese one DeepSeek sneaks into the top 10 list. Here ChatGPT dominates the ranking with Gemini and Canva completing the podium, but it is surprising to see the relevance of Grok, ahead of Claude. And Google shines with its own light. Within the list, the presence of Google is also notable, which has four tools on that list: Gemini (in number 2 on the list), Google AI Studio (10), Google Labs (25) and the splendid NotebookLM (30). Most of these apps come from the US. Graphic: Xataka with Gemini. Data: Andreessen Horowitz. China tightens. It may seem that China’s role here is less relevant than it should be, but it must be taken into account that many Chinese startups focus on platforms and applications for the Chinese market. Even so, there are clear protagonists such as Capcut and Doubao (ByteDance), Qwen and Quark (Alibaba), Kimi, Kling, Cutout and of course the aforementioned DeepSeek. Europe has its protagonists. Freepik is the clear standout on this list among the European AI applications, but there are others that stand out and manage to make it onto the list such as Photoroom (France), Turboscribe and Veed (United Kingdom), Remove.bg/Kalleido (Austria) and another standout, ElevenLabs (based in London). An evolution towards hybrid apps. As Andreessen Horowitz points out, three years ago the distinction between “native AI” products and traditional software was clear. Today that barrier has disappeared, since massive tools like CapCutCanva or Notion have integrated generative AI as the core of their experience and revenue engine. They have taken advantage of their inertia, they have adapted and they have won. In mobile apps the ranking changes, and a lot. The most popular AI mobile apps in the world based on their number of active users each month is very different. Here Freepik disappears from the list, for example, and it is China that totally dominates with 22 of the 50 apps (44%). The US has 13 apps on the list (26%), while Europe only has four (8%) and other countries share the other 11 (22%). Here China benefits from its huge user base, who also very frequently use AI applications for all types of functions. ByteDance is especially eye-catching and has five apps on the list (CapCut, Doubao, Cici, Hypic and Gauth). Divergence of approaches. In general, all apps try to build user loyalty through their ecosystems and try to integrate more and more things so that one does not leave them. However, there are important approaches among some such as ChatGPT, very oriented towards being a “super app” for mass consumption, and Claude, from Anthropic, which focuses on professional and technical users. AI wants to be almost invisible. AI is no longer a destination, a website to go to, but is becoming part of the experience, a function integrated into the application. Thus, it now resides directly in the browser, in development environments or in office suites. In Xataka | The war between Anthropic and the Pentagon points to something terrifying: a new “Oppenheimer Moment”

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.