billion

Big Tech spent $725 billion on AI. Then they ran out of money in their pockets.

by

This is non-stop. Big tech companies have already spent an irreverent amount of money in 2025 to not lose footing in the AI ​​race, but this year things are getting better. Together Amazon, Microsoft, Google and Meta have announced a capex of $725 billion, which represents an astonishing 77% growth over last year’s (also astonishing) figure of $410 billion. The numbers they are dizzyingbut they are having a worrying consequence. A lot of money saved. For years, Big Tech has been able to boast extraordinary accounting books in which revenues and profits have practically not stopped growing. They’ve built up exceptional cash flow, but now they’re taking advantage of all that money to fund an AI race that doesn’t seem to end at the moment. Cash flow plummets. The amount of investments is of such magnitude that all of these hyperscalers have encountered a problem: their cash flow—the available liquidity— has collapsedthey indicate in the Financial Times, and now it is at levels that we have not seen since 2014. Before, the average was to have 45,000 million dollars since the pandemic, but now that figure is expected to fall to 4,000 million in the third quarter of 2025. Source: Financial Times. Let’s see who spends more. Amazon leads this unique race for spend more than others. The company led by Andy Jassy foresees an investment of 200,000 million dollars in 2026, which will lead it to burn about 10,000 million of its cash flow this year. Meta will continue that same trend in the second half of the year, while Microsoft could enter negative territory in at least one quarter. Even Google, which remains positive, will post its lowest level of cash flow in a decade. Debt, new fuel for AI. To finance this deployment, both Alphabet and Meta have had to resort to massive debt issues and suspend their share buyback programs for the first time in almost a decade. Alphabet issued $48 billion in bonds recently (in February a partdoes some days other), while Meta sumo 55,000 million debt in just six months. Bet now to win later. This strategy marks a paradigm shift: it is no longer investing only with the income one has in cash, but Big Tech is mortgaging its future. The objective is what we have mentioned time and time again: not to lose step in a race where, as Zuckerberg said, staying behind is not an option. Disguising the beads. These companies fear Wall Street’s reaction to these movements, so they are moving billions of dollars in infrastructure but they are doing so outside of their conventional balance sheets. In the FT they explain how Big Tech are using special investment vehicles that allow them to attract external capital and hide debt. They are also more opaque about who will be impacted if the AI ​​does not meet expectations. The memory crisis is also having an impact: in such a way that Microsoft already has added 25 billion dollars to its investment needs this year just to be able to assume the increase in component prices. The danger of going with the flow. CEOs justify these moves by comparing them to what happened with cloud investment two decades ago, but analysts warn: investing when the competition invests is not always a strategic choice, but rather a forced response to staying out of the race. In Xataka | The chip crisis is leaving no stone unturned: motherboards seemed untouchable, but their time has come

Emirates and Oman are building a $3 billion megatrain. The problem is that it crosses a drone battlefield

by

In the midst of a geopolitical climate where tension cut with a knifean infrastructure megaproject emerges in the Middle East that challenges the context of conflict. It is about the construction of the first cross-border rail network of the region. Promoted by Etihad Rail, Oman Rail and Mubadala, this plan proposes a corridor that will integrate the national network of the United Arab Emirates with the strategic port of Sohar, in the Sultanate of Oman. However, the immense work advances under a dense shadow. While the pillars of this train are being raised, the region is going through what in practice It’s the Third Gulf War. The impact of a commercial revolution. To understand the magnitude of “Hafeet Rail”, just look at their economic projections. This mixed corridor—designed for both passengers and cargo—promises to radically transform the flow of trade in the Gulf and lower logistics costs. The network has a monumental investment which is around 3,000 million dollars, equivalent to about 2,500 million euros. Additionally, the infrastructure will link five major ports and more than fifteen integrated cargo facilities directly. The benefits, however, will not be exclusive to maritime trade. For the average citizen, this line will mean an unprecedented change: the trip between Abu Dhabi and Sohar, which currently takes more than three hours on winding roads, will be reduced to just 100 minutes. In addition, it will offer a reliable alternative that will eliminate the usual and costly delays at border crossings. The challenge of operating in a disputed region. The main route of the project will cover a length of 238 kilometers. On these new generation tracks, passenger trains will be able to reach speeds of up to 200 kilometers per hour, while heavy cargo convoys will circulate at a maximum of 120 km/h to optimize international shipping times. Far from being a mirage in the desert, construction is now a tangible and has reached 40% overall progress. On the rugged terrain, backhoes have completed more than 27 million cubic meters of earthworks, and there are currently 80 key structures in different stages of construction. The big question: can it work? Military analysts warn that the recent proliferation of cheap drone attacks has shown that facilities previously considered untouchable are today extremely vulnerable. The fact that the United Arab Emirates host allied infrastructure and bases It makes them latent targets within this tense regional board, adding enormous operational risk to any large connectivity project. Technological avant-garde. On a technical level, the project does not skimp on innovation. According to the technical documentationthe railway fleet will be equipped with the European Train Control System (ETCS Level 2), considered the most advanced and safest in the world in its category. This system, which will be implemented by a joint venture between Siemens and HAC, will allow absolute digital tracking and control of trains using GPS technology. Regarding the execution of the challenging civil works, these were awarded to an Omani-Emirati consortium led by Trojan Construction Group (NPC) and Galfar Engineering and Contracting. A milestone that the consortia particularly celebrate is the extreme workplace safety achieved: to date, 10 million hours of work have been recorded in the field without reporting serious accidents. Closing a historical gap. Beyond the colossal engineering figures, the project carries a deep cultural weight. The unified network has recently adopted the identity of “Hafeet Rail”, a direct tribute to the Jebel Hafeet, the imposing mountain and limestone formation that extends between the borders of both countries and that has historically served as a geographical bridge. Despite business optimism, the success of the operation will not depend solely on laying tracks. Monumental bureaucratic challenges await ahead, such as regulatory coordination between the two sovereign nations and the fluid articulation of port and customs services. In the end, time will tell if the shared vision of progress prevails. For now, Oman and the United Arab Emirates are committed to full economic integration and the creation of a new artery for global trade; All of this, paradoxically, at a time when its immediate environment is navigating a hybrid war defined by uncertainty, intermittent blockades and air threats. It is, in short, a bullet train making its way through a minefield; the maximum expression of risk and ambition in the heart of the Middle East. Image | Photo by Grant Durr on Unsplash Xataka | The US believes that the war in the Persian Gulf is over. Iran believes that it will decide that when it considers

a diamond from 2 billion years ago

by

It was at the beginning of the 20th century when, in a south africa minea foreman named Frederick Wells thought he saw a simple flash in a rock wall and decided to check it with his knife. What he got out of there turned out to be the biggest diamond never found, a piece so large that for years was doubted whether it was just a fragment of something even greater. The iconic scene left a curious idea that is repeated in the history of mining: sometimes, the most extraordinary finds appear just when no one is looking for them. Luck at the last minute. It happened at the beginning of April, when in one of the most remote regions of the planet, a few kilometers from the Arctic Circle, a mine which was already facing its last days of activity has left an unexpected discovery that rewrites its ending. This is not just a new geological discovery, but one that combines extreme rarity, almost unimaginable antiquity and a context that makes it something much more symbolic How usual. In a place on the planet where every extraction seemed to be part of the past, the earth has offered one of its oldest secrets at the last possible moment. An extraordinary diamond in every way. It is not trivial, because the stone found, with more than 158 caratsis among the largest yellow diamonds ever discovered in Canada, a country where this type of gem is already exceptional. In more than two decades of activity, only a few few comparable pieceswhich places the discovery in a practically unique and almost unusual category. The rarity is even greater when you consider that this type of diamond represents less than one percent of the mine’s total production. Two billion years. Yes, because the true value of this diamond lies not only in its size or color, but in its fascinating origin. The researchers said that, formed approximately two billion years deep within the Earth, it is the result of extremely slow geological processes that have remained intact until today. Its yellow color, a product of presence of nitrogen in its crystalline structure, it adds another layer of uniqueness to an already exceptional piece. On the brink of closure. As we said at the beginning, what makes this discovery especially significant is the moment in which it occurs. The Diavik mineoperational since 2003, just closed after more than twenty years of activity and more than 150 million carats extracted. In other words, this diamond appears as one of the last great discoveries before the end, functioning almost as a symbolic closure for an operation that has marking the industry in northern Canada. Extreme engineering in one of the harshest environments. The context in which the discovery occurs is key to understanding its importance. The mine operates in subarctic conditionswith extreme temperatures and in an isolated environment that has forced the development of advanced technical solutions, from containment dams in frozen waters to hybrid energy systems with renewables. This level of complexity turns each extraction into a logistical and human challenge that goes far beyond simple mining. Beyond the stone. During its lifespan, the mine has not only produced diamonds, but has transformed the economy of the region, generating thousands of jobs and important industrial activity. Furthermore, it has established collaborations with communities local indigenous people for the management of the territory and its future restoration, a key aspect now that the exploitation has come to an end and the environmental recovery process begins. The last gift. If you also want, together, the discovery sums up the essence of the entire operation: technology, nature and time converging in an unexpected moment. When everything pointed to a definitive closure without any major surprises, the mine has delivered one of its most extraordinary pieces at the last minute, as if the land itself refused to disappear without leaving a last trace. Thus, more than a simple discovery, the diamond has become the region in the final symbol of a cycle with the most filmy closing. Image | Rio Tinto In Xataka | The diamond industry has been looking for a way out of its biggest crisis for years. Taylor Swift just served it on a platter In Xataka | The diamond industry promised to be happy with lab-grown jewelry. Until prices crashed

Microsoft just turned an $11 billion startup into a Word feature. It’s more than a legal Copilot

by

Brad Smith is more than the vice chairman of the board and president of Microsoft: Smith is also a lawyer and as he himself tellsat the beginning of his career he asked his company for a computer because he firmly believed that computing could change the way lawyers work. In fact, his Wikipedia biography gives more detail: it was the requirement that the Washington, DC law firm Covington & Burling set to join. Said and done: in 1986 he was the first person in the firm to have one, which ran the legendary Word 1.0 processor. Seen in perspective it sounds like marketing, but a tremendous omen: Microsoft just announced Legal Agent for Wordan AI agent designed for legal work. What’s new from Microsoft is not a legal Copilot. Legal Agent is an agent designed to understand and operate within a legal document as a lawyer would: it analyzes risks, compares clauses against the organization’s internal rules, has tracking for the changes it generates, differentiates previous reviews of new proposals and detects potentially problematic provisions. Everything happens within the .docx itself, without leaving Word. What distinguishes it technically is its architecture. The agent does not ask the LLM to generate each edit directly, but instead combines that semantic understanding layer with a deterministic layer that applies the changes in a controlled way. This allows you to insert clauses, delete paragraphs, or add comments while preserving the original formatting of the document, including tables, lists, and change history. The result is a more reliable and predictable system than a chatbot, with fewer hallucinations and with the consistency that legal work demands. Brad Smith’s tweet includes a video that lasts almost a minute and a half where it can be seen in action: Tap to go to the post Why is it important. The key is not so much the technology, which already existed, but rather the distribution: Word is the program par excellence for drafting, reviewing and negotiating contracts around the world. Integrating there means being in the right place at the right time, without friction: it eliminates the need for another service, creating an account and logging in, the learning curve, the workflow between two different apps, data migrations and security. All in one, all easy. The definitive boost is the price. While subscribing to specialty products like Harvey they hover the 1,000 – 1,200 dollars per lawyer per base month, according to market estimates collected by Sacrathe Legal Agent arrives integrated into theCopilot Enterprise subscription of 30 dollars a month that surely many spiteful people already pay per se. The difference in magnitude and the product placement anticipate a voracious entry into this market niche. Context. A troubled river, fishermen’s profit: Microsoft did not start from scratch for this project. At the beginning of the year contract to more than 18 engineers from Robin AI, the legal AI startup that collapsed after failing to close its $50 million round. Probably if Robin AI had not fallen, Microsoft would not have been able to create such a product so quickly. We were talking about other specialized products but the name on the horizon was one: Harveythe sector’s benchmark. Founded by Winston Weinberg and former Google DeepMind Gabe Pereyraoperates with more than 100,000 law professionals in more than 1,300 organizations and is valued at 11 billion dollars. Your latest financing round It was 200 millionclosed in March 2026 and co-led by GIC and Sequoia. It is true that its proposal goes beyond the review of contracts: it has more than 25,000 personalized agents operating on its platform with deep integrations into the document management systems used by large law firms, such as iManage and NetDocuments. Bottom line: It’s not a $30 a month feature. Yes, but. In any case, for now the product is still in early access, only in Word for Windows, with configuration restrictions and some complaints from those who have already tried it. Furthermore, it remains to be seen whether lawyers will trust into a mainstream tool for highly complex cases where a minimal error can be costly. The battle of price and distribution is won, confidence and technical depth is another story. Saying that Microsoft is going to kill Harvey it’s an exaggeration: The Legal Agent is more focused on volume work, that more mundane work of routine reviews, standard contracts, NDAs… that takes legal professionals hours every day. Harvey is strong in more complex and/or high-risk tasks: a multinational with a serious litigation advised by an elite law firm is hardly going to entrust the matter to an agent included in an Office subscription. What the Robin AI story does make clear is that having a good product and customers does not guarantee survival: the group of organizations willing to pay is smaller than the investment rounds anticipated. In Xataka | The relationship between Microsoft and OpenAI is no longer exclusive. It took someone 48 hours to fish in a troubled river: Amazon In Xataka | The results of the technology companies are very clear: the business of AI is not AI, it is renting its infrastructure Cover | Brad Smith on Twitter

Vinted is already worth 8 billion and has achieved it without AI. Just selling your neighbors’ used clothes

by

Vinted, the Lithuanian second-hand platform, has closed a secondary sale of shares of 880 million euros led by EQT which increases its valuation to 8,000 million. It has not raised new capital. It has let investors and employees out, and brought in new shareholders (BlackRock, Schroders, Teachers’ Venture Growth) who can hold out both privately and on the stock market. This IPO does not even have a date yet, but the company says that already operates internally as if it were listed. Why is it important. The technology ecosystem in 2026 has been obsessed with AI for some time, so Vinted is a nice anomaly: it has built a profitable business, with more than €1 billion in revenue and €62 million in net profit in 2025, without mentioning AI anywhere. Its value proposition is different, and its story is not that of a startup that has found a shortcut but rather that of a market that has taken fifteen years to mature and that is now changing consumer habits on a continental scale. The context. Vinted was born in 2008 in Vilnius as a way for neighbors to exchange clothes. It has taken almost two decades to become what it is today: a second-hand trading infrastructure with its own logistics, integrated payments and presence throughout Europe. In 2024, TPG capital entered at a valuation of 5,000 million. In just over a year, that figure has risen 60%. In figures: 10.8 billion euros in gross merchandise value (GMV) in 2025, 47% more than the previous year. 1.1 billion euros in revenue (2025). 62 million euros of net profit (2025). 8,000 million euros of valuation after the operation, compared to 5,000 million in 2024. Yes, but. The profitability is there, but it is modest for that valuation: 62 million profit on 1,100 million income is a margin of 5.6%. Only 1.2 points above that of Mercadonafor comparison. Far below that of any technology. To justify 8 billion, Vinted needs to demonstrate that it can scale that margin and not just volume. The online second-hand market is quite competitive: eBay paid $1.2 billion two months ago to buy back Depop from Etsy and strengthen its position in second-hand fashion. Wallapop It has a generalist profile that also takes away its share in countries like Spain. And in the United States, the large market that Vinted has not yet conquered, the company recognizes which is in the testing phase, not expansion. Between the lines. The entry of EQT as an anchor investor in this round has more meaning than it seems. EQT is the Swedish fund that also controls Idealista and Magnific (before Freepik). Its commitment to Vinted reinforces the thesis that large European private equity funds are building positions in second-generation digital platforms: businesses with real network effects, their own infrastructure and proven traction in Europe, before they are listed. When the time comes to go public, they will be caught in it. The big question. Can Vinted replicate in the United States what it has done in Europe? The company has started allowing buyers from London and New York to trade with each otherbut the American market has its own dynamics, its own consolidated platforms and a different second-hand culture. The answer to that question will determine whether Vinted is an $8 billion company or has the potential to become an $80 billion company. In Xataka | There are too many clothes in the world and there is a company earning billions of euros thanks to it: Vinted Featured image | Xataka with Mockuuups Studio

Google will invest up to $40 billion in Anthropic because the new normal for AI is investing in your enemy

by

May the rhythm not stop. Amazon announced an investment of 25,000 million in Anthropic a week ago, and four days later Google went even further. The Mountain View Company spoke on Friday of an investment of up to $40 billion in that same company. We insist: this is non-stop. The money doesn’t stop flowing. In less than a week, two of the largest “cloud providers” in the world have committed to investing up to $65 billion in a company that, attention, is a direct competitor in the AI ​​segment. None have done it out of generosity, and here there is a lot of covering one’s back and, of course, circular financing. This is the Google agreement. Google will invest $10 billion now considering that Anthropic’s valuation is between $350 billion and $380 billion. From there, it can invest another $30 billion linked to company performance milestones that have not been detailed. What Google gains. In exchange for that investment, Google Cloud will provide an additional 5 GW of computing capacity from 2027, expanding the agreement that Anthropic had already announced with Google and Broadcom to contract 3.5 GW of computing in the form of access to their TPUs. Google already invested 300 million dollars in Anthropic in 2023, but months later he put it on the table another 2,000 million more and in 2025 another 1,000. Anthropic is already worth a fortune. It is estimated that before this agreement its participation in Anthropic was around 14%, and with this new agreement that participation will evidently increase. Anthropic’s valuation has grown dramatically in recent months, and according to Bloomberg There are offers for a new investment round that would place its value at 800,000 million dollars, already at the level of the 850,000 million valuation that OpenAI is around. Its growth is overwhelming, and it is clear that today She is the pretty girl of the industry. No one could wait. The speed with which these announcements have occurred is motivated in part by the competitive fear between Amazon and Google. Anthropic uses Trainium chips from Amazon and TPUs from Google: it needs both and they both know it. Every dollar those companies put into Anthropic is a business case for Claude’s clients to use AWS or Google Cloud, so it makes sense that both want to solidify that “preferential relationship” with the company that is conquering the enterprise market. The circular financing model as a standard. This week’s agreements consolidate what many already consider as the new normal sector: hyperscalers invest in AI startups, and AI startups spend that money on the infrastructure of those hyperscalers. For example: Google Cloud grew 36% in revenue last year to $58.7 billion and Anthropic was most likely one of its heavy clients. The money Google invests in Anthropic comes back in the form of invoices, and the same goes for Amazon and Trainium. But the investment has another reason. These investment agreements not only seek to strengthen ties with the most promising AI startup of the moment, but also have a significant stake in its shareholders. That’s even more striking, because both OpenAI and Anthropic They hope to go public before the end of the year and if so, Google and Amazon will have “bought cheap” their stake in a startup that is expected to skyrocket exceptionally once it becomes a public company. Once again, this is a bet for the future. But there is also the other big reason: the majority of investors (be they funds or companies) do not want to be left behind in this race and are betting because everyone else is doing it too. It doesn’t matter that AI companies are losing money non-stop: the promise is that there will come a time (2029 or 2030) in which the trend will change. It is not certain that this will happen, of course, but OpenAI or Anthropic play with that card and use it to their advantage. We have the last example in Mythos, an Anthropic model that it’s so good (or so they say and some others) who prefer not to make it public. It’s once again selling expectations… and it works. In Xataka | DeepSeek has just released a model that competes with Opus 4.6. It costs seven times less and runs on Chinese chips

The Pentagon wants to invest $54 billion in drones. It is more than the entire military budget of countries like Ukraine

by

The defense budget that the Pentagon has presented for fiscal year 2027 amounts to $1.5 trillion. It is the largest year-on-year increase in military spending since World War II, but in that colossal figure there is another that deserves special attention. This is the $53.6 billion allocated exclusively to drones and autonomous warfare technologies. That amount alone exceeds the Ukraine’s full defense budget either of countries like South Korea or Italy. Spain is even further away. autonomous defense. The money for this specific program will be managed by the Defense Autonomous Warfare Group (DAWG), an agency created at the end of 2025. In the 2026 budget it received 226 million dollars, but in 2027 that figure would be multiplied almost by 240. The United States has realized the relevance that drones have gained in war conflicts and wants to be prepared for this new era of defense. Obsolete investment. The Pentagon itself recognized something striking: the vast majority of the money requested will be used to buy technology that already exists, not to develop future solutions. One of the top officials of the Joint Chiefs of Staff, Lieutenant General Steven Whitney, admitted that technological evolution on the battlefield currently happens in weeks, not years. It’s like admitting that what you buy now may become obsolete almost immediately. Ukraine showed that change has changed. The urgency of this budget does not come from nowhere. The war in Ukraine has rewritten the rules of modern combat In such a way that there are many countries that are processing how to assume these changes. Iranian Shahed droneswhich cost about $20,000 per unit, have proven capable of saturating air defense systems that cost hundreds of times more. Relatively affordable quadcopter drones have destroyed multi-million euro tanks and armored vehicles. Defense budgets in 2025. The US already spent 921 billion dollars last year, this year it wants to spend 50% more. Everything goes very fast. The speed of tactical adaptation on the Ukrainian front has been so high that innovations and tactics that work in January may be obsolete by March. Not because someone has invented something better, but because the adversary has found a way to counter those strategies. The Pentagon has reached an unusual conclusion: the traditional model of weapons acquisition that operated in cycles of years or even decades is structurally incompatible with the speed at which current war conflicts are developing. The irony of the Shahed. Among the most striking details of the budget is the confirmation that the American army has adapted the technology of the Iranian Shahed dronewhich is the same one that has been attacking cities and energy infrastructures in Ukraine for years. The US has done reverse engineering of your adversary’s design to incorporate it into your own arsenal. This clearly illustrates the current war reality: the origin of the technology does not matter, but its effectiveness. Risks. This tension between “we have to spend more” and the speed at which it is necessary to adapt to this reality poses an enormous risk. Buy en masse what works today guarantees that solutions will be available tomorrow. The problem is that these solutions may be technically inferior to those that the adversary has developed in the meantime. The same thing happens if you decide not to buy anything until you have the perfect technology, because that means arriving late (or not arriving at all). It is a dilemma similar to that of technology companies and their investment in infrastructure: they have to buy solutions now that they know that they will end up being obsolete in the short or medium term. Final approval is missing. The US Congress will have to approve the budget, which introduces an important political variable. Beyond that, there is a fundamental question in those 54,000 million in this budget. If drone technology evolves in weeks, there is no money that will be able to buy that adaptability to the modern battlefield. And that even with this immense budget superiority cannot be guaranteed makes clear the sign of the times. In Xataka | The percentage of GDP that each country allocates to Defense, shown in this graph with an unavoidable protagonist

Five years ago, Venice spent more than 5 billion on a system of barriers against the sea. Now look for a plan B

by

There was a time when Venice looked at the Adriatic with ambition. The sea not only shaped the city, permeating its DNA, it also propelled it until it became a naval power who fought for dominance of the Mediterranean. Today things are different. The Serennissima (turned into tourist power) observes with increasing concern the coming and going of the tides, the same ones that in 2019 submerged it under 187 cm of water, flooding 80% of the city. The reason is very simple. Everything indicates that the multimillion-dollar system that Venice was equipped with a few years ago to protect itself from the threat of high water It won’t take long for it to become obsolete. And it is not very clear what the alternative is. One figure: 18. The threat of flooding is not new in Venice. In fact, one of the worst in memory was suffered six decades ago, in November 1966when an intense storm caused the water to reach 194 cm, flooding much of the city. However, experts have been detecting worrying signs for some time. It is not just that Venice sink or the sea level rising (which too). There are increasingly clear signs that suggest that floods will become more frequent in the future. Recently, a group of researchers dedicated themselves to analyzing the “extreme” episodes suffered by the city, those in which 60% of its surface was flooded. Throughout the last century and a half, it counted 28 incidents of those characteristics. The surprising thing is that the vast majority of them (18) were concentrated during the last 23 years. One measurement: 0.42 m. Today more than half of Venice is alone between 80 and 120 cm above the average sea level and projections show that this scenario will soon worsen: in the best of cases, if we manage to drastically reduce our polluting emissions, the sea will rise 0.42m by 2100. In the worst case, it will be 1.8 m, which would greatly complicate the outlook for the Serennissima. In fact, now the high tide already leaves St. Mark’s Square only 30 cm above the water level. One name: Mose. Aware of how much is at stake in Venice, the Italian Government has long been looking for a way to protect itself from floods. The result was Mose (experimental elettromechanical module)a system made up of four barriers and 78 independent mobile gates that allow authorities to protect the Venetian lagoon from what is known as high watertides that flood the city. The objective: to temporarily isolate the Adriatic lagoon and thus protect Venice from the most dangerous tides. To achieve this, the barriers were strategically installed in the inlets of Lido, Malamocco and Chioggia. Each gate also measures 20m wide and between 18.6 and 29.6 m long. An investment: 5,000 million. It is said that the project mobilized an investment of more than 5.5 billion of euros (its execution was marred by corruption). Its work began in 2003 and after several delays it carried out a first test in October 2020, in an event led by the then Prime Minister Giuseppe Conte. A year earlier, Venice had suffered a of the worst floods that are remembered, during which the water reached 187 cm, flooding part of the entrance to the Basilica of Saint Mark. An indicator: frequency. The problem is that the authorities are turning to Mose much more often than expected. EuroWeekly assures that in less than a month, between January 28 and February 19, the system was activated 30 times. Other media report that since their inauguration at the end of 2020, the barriers have saved Venice from flooding in 154 occasions. The problem is that the use of Mose does not come free to the region, neither in economic terms nor on a social and environmental level. Setting up the enormous Mose floodgates has a direct cost, but it also has another indirect cost: by isolating the lagoon, the system alters, for example, the activity of the port sector and interrupts maritime traffic with the port of Marghera. Guardian points out that pressing Mose’s button has an economic impact of more than 200,000 euros for Venice. For this year’s Carnival alone the total bill would be around five million euros. An extra concern: the lagoon. Not everything is measured in operational cost, maritime traffic and economic impact. Altering the tides in the area also has an impact on its ecosystem and that is something that worries experts like Andrea Rinaldo, from the scientific committee of the Lagoon Authority. Especially if two fundamental data are taken into account: first, the frequency of use in recent years; second, the forecasts for sea level rise. “With one more meter, the Mose barriers would have to be closed an average of 200 times a year, which means that they would practically always be blocked,” explains Roinaldo. “When this happens, the lagoon loses its function as a transitional environment. It would become a pond.” A victim: the lagoon itself. As explains GuardianBy blocking the flow of water, the barriers encourage the growth of algae. The problem is that when these die and decompose they directly affect the quality of the water and the rest of the flora and fauna. Does that mean Mose was a mistake? Rinaldo thinks not. The changes are simply happening much faster than engineers expected, forcing authorities and technicians to think about the future in the medium and long term. At the end of the day, if Mose taught anything, it is that projects of his importance are not approved and executed overnight. One question: What to do? The great unknown. Those responsible for Mose are looking for ways to reduce its impact, but it is not an easy decision. Among other things because the Venetians themselves have become accustomed to the barriers and gates coming into operation at the slightest risk, points out Giovanni Zaroti, one of the system technicians. Rinaldo mentions the possibility of launching an international call … Read more

Meta spent 2 billion on a Chinese AI startup. China is clear that it was a conspiracy

by

With China and the United States dancing the dance of artificial intelligenceboth countries and companies want to get the best cards for their decks. Meta is investing millions in the development of AI and, even so, it seems to be lagging behind. To turn the tables, he closed 2025 with a $2 billion purchase: that of a Chinese startup called Manus. The operation was so notorious that the Chinese government itself raised an eyebrow and undertook an investigation to see what was happening there. And they are already clear. It was a conspiracy. The Manus case. Although it has rained a lot and these last few months in China AI companies have come out from under the stones, during the first half of 2025 the proper name was that of deepseek. It was the great competition from the Western OpenAI or Google Gemini, but in March something that looked like an AI agent began to appear: Manus. That’s how they sold italthough it was really a deep investigation mode that helps you perform actions, but does not do them for you. It didn’t matter: the expectation was there and, although there were doubts about his behavior and limits, Manus began to move a lot of money (more than 100 million in estimated income) and attract attention from the big players. One of them was Meta, who took over the company. The purchase. A good question is how China let something like this slip away for a technological and strategic rival to buy. And it’s a good question, but the answer is that, at some point, Manus stopped being a Chinese startup. In the middle of last year, Manus moved to Singapore, allowing the company to bypass export and import controls imposed on China. To the not having your own LLMthey depended on others like Claude which they could more easily access from outside China. This already set off alarm bells in the Government, but with the purchase of Meta the bells echoed. China put to work to various organizations to see what was really happening, the largest of them being the Chinese National Security Commission, which is commanded by President Xi Jinping himself. The reports prepared by this body are directly supervised by the leaders of the Communist Party, so it is a voice that must be taken into account. Conspiracy. And the result of the investigation is clear. As they comment in Financial Timesthe conclusion is that Meta’s acquisition of Manus is a conspiratorial attempt to try to undermine China’s technological capabilities. These are big words that do not remain in a vacuum, since the founders of Manus – Xiao Hong and Ji Yichao – were summoned by the NDRC last March to address issues such as possible violations of foreign investment rules in China. He did not stay for a meeting and, as the FT points out, both have been prohibited from leaving the country during the review process. In fact, there are sources that suggest that Manus would be considering backing out of the agreement, but even so, it is not clear that the Chinese authorities will be satisfied. For his part, Meta points out that they did everything according to the law and it seems that he has already started to integrate Manus systems into their tools, so taking that step back would be very complex. And now… what. That the National Security Commission has classified the case as “conspiracy” is something serious, since it was the trigger for a broader review that involves more agencies in the country that are currently reviewing everything. And the underlying problem is the speed with which everything happened. Manus took off and, just four months later, they moved everything to Singapore to break away from China just before the purchase of an American company. The investigation is shaking the Chinese technology sector because it is not the first time something like this has happened. Although on a smaller scale, it is an operation called ‘Singapore washing’ in which startups founded by Chinese move to the city-state to bypass China’s control and have a more direct line with the United States. The problem is that, at a time when the commercial and strategic war has intensified, calling the Manus case a “conspiracy” sets a precedent. One in which it is stated that China does not want to let artificial intelligence talent and technology escape because this advance has become one of the country’s strategic legs for the next five years. We will see what happens when the case is resolved, but it is clear that Beijing’s objective, like Washington’s, is to prevent its assets from escaping, and Manus can be the example for national technology companies do not follow a similar model in the future. In Xataka | We don’t know if “crisis” means “opportunity” in China, but there is one business where it does: RAM memory

We thought we were 8 billion people on the entire planet. Until some researchers started crunching the numbers

by

In November 2022, the UN celebrated that we were now 8 billion humans on Earth. They are estimates, of course, but beyond the figure, the really interesting thing is that in 2023 we do not reach the replacement rate and that humanity will reach its peak at the end of the century to, inevitably, start to fall. But… to what extent can we trust these accounts? It is something that has been on the table for some time and, according to a study of 2025, we have made a mistake in counting. So much so that we have left several hundred million people behind. Can we trust the numbers? “Calculating the number of people on the planet is an inexact science.” That was demographer Jakub Bijak’s comment to BBC in mid-2024, just when the World Population Prospects study. Something scientific is something exact, but the researcher also commented that the only thing you can be sure of when predicting population figures is the lack of certainty. That, be careful, does not mean that demographers take figures out of thin air. “It is a difficult thing based on our experience, knowledge and every piece of information we have access to,” said Toshiko Kanera, an expert in demographic forecasts. Demographers draw on the data and trends of each country since 1950, but… what if it had not been counted correctly? We are missing millions. In a 2025 study published in Natureresearchers at Aalto University in Finland show how the data sets handled by demographers “profoundly and systematically” underestimate population figures around the world. The serious thing is that we would be talking about hundreds of millions more people living on Earth. Example of the tools that demographers use in their analysis. Each one corresponds to a different bias Rural areas. Josias Láng-Ritter is one of the researchers in charge of the study and points to the accounts carried out in a specific segment: that of rural population. “For the first time, our study provides evidence that a significant proportion of the rural population could be missing from global population data sets,” he notes. As we say, we are not talking about a few million, but billions. “Depending on the data set used, rural populations have been underestimated between 53% and 84% in the period studied. The results are notable, since these data sets have been used in thousands of studies and have widely supported decision making, but their accuracy has not been systematically evaluated,” comments the researcher. The map shows the location of the 307 rural areas analyzed in the study. The populations reported in the graph were found to be underestimated by between 53% and 84% | Aalto University Biases. Attempts to review this data are not new, but previous research has focused on specific countries or urban areas. Researchers from Aalto University wanted to give a more global picture by comparing the five most used population data sets worldwide. They have used maps that divide the planet into high-resolution grids and have taken something very specific as a reference: resettlement figures from more than 300 rural dam projects in 35 countries. Why this bias of the dams? Because when a dam is builtthe population that lives in the area that will be flooded is relocated and accurate resettlement data is usually available. Comparing that population data from 1975 to 2010, the researchers found that the 2010 maps were more accurate, but still left out between 32% and 77% of the rural population. Between 2015 and 2020 the data sets were updated, but demographers continue to believe that underestimation of the rural population continues to exist and is a problem that persists in all regions of the world. Consequences. And we are talking about a problem whose resolution is complex. According to the researchers, no matter how much the data is reviewed, it is a structural problem. Governments do not have the resources to collect accurate data in these rural regions, there is a huge discrepancy between the real population and that reported on the population maps used to carry out demographic studies and that influences decision making. Average percentage of rural population underestimated (red and orange) and overestimated (blue) | Aalto University And it’s important. Current estimates place 43% of the 8.2 billion inhabitants of the world in rural areas -about 3,526 million people- and if we take into account that it is a percentage that has been underestimated between 53% and 84%, we are not talking about a small population, precisely. And it is essential to know exactly how many we are for a simple reason: the redistribution of resources. No data. The lack of accurate demographic records can affect political decision-making. Ritter gives the example of social decisions. “In many countries, there may not be enough data available at the national level, so they rely on global population maps to support their decisions: Do we need a paved road or a hospital? How much medicine is needed in a specific area? How many people could be affected by natural disasters like earthquakes or floods?” he says. Doing quick math, in the best scenario – that of 53% deviation in the rural population – we would be talking about 1,869 million people who would not have been counted. In the worst case, in that of the 84% not registered, we would talk about 2,962 million people. In the Nature study, they give the example of Paraguay, which in the 2012 census may have left out a quarter of the population. Reviewing the methods. In the team’s analysis, there are countries that fare better than others. They point to Finland as an example of reliable data, even in rural regions, because they began keeping digital records of the population 30 years ago. However, in countries where this thorough digital registration has taken longer to be implemented due to crises of any kind, the differences between the real population and the estimated one can be significant. “To provide rural communities … Read more

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.