Google

Google, Amazon and Microsoft have been burning absurd amounts of money in Ia for years. Finally they begin to see green sprouts

by

The AI boom made Big Tech will increase their capital spending to limits that had never been seen. The fear that The bubble will explode Rondaba, investors They started to get nervous and the profitability of AI remains in doubt. The last results are a green outbreak, the first in a long time, although with many buts. The cloud reaches capex. They tell it in The Information. Capital or Capex expenses of the Big Tech in recent years have climbed unstoppable, much faster than their income, but in the results of the last quarter the gap is finally closing, but not because chatbots and other products are being profitable, but thanks to revenue from cloud services. The crazy one is committed to AI is beginning to show a slight green outbreak, even if it is not directly because of AI products. Income from cloud services are approaching capital spending. Source: The Information (click on the image to access X) The four riders. There are two clear winners of the departure, one that already brought the duties done and one that goes free. Let’s see who is who: Microsoft: The clear winner with a Income increase of 25% In the last quarter, mainly thanks to the growth of Microsoft Azure. Google: Record a 20% increase In your income thanks to Google Cloud and advertising. Amazon: falls 7%but it is the only one that was in positive numbers. Amazon Web Services is the largest provider of cloud services and was already profitable, although its growth is beginning to slow down. Goal: Your income grows 22%but they basically come from advertising. Goal does not sell cloud services, so it does not generate income directly. Indirectly, yes: AI has allowed them Improve the efficiency of your advertising business. Burning money. The increase in capex by AI has reached madness figures that had never been seen in other technological booms. By the end of 2024 we talked about investing a real barbarity In data centers: Microsoft 30,000 million, Goal 35,000 million, Google 25,000 million… The dizziness figures, and have not stopped increasing. Amazon said at the beginning of the year that He wanted to spend 100,000 million in data centers for AI and goal is building several data centers whose combined cost could rise to 200,000 million dollars. Skepticism. This excessive spending frenzy soon unleashed a wave of skepticism. AND If AI is another bubble And is it about to explode? Milmillionaire investments are not translating in income. Even Satya Nadella himself, one of the protagonists of this revolution, was skeptical because At the moment no one is making gold with AI. It is not that they are not making gold, it is that nobody is earning money. In their newsletter, Ed Zitron had accounts And the difference between what is expected to spend in 2025 and the return that is giving them the AI is not that it is a reason to doubt, it is directly no sense: Capex planned in 2025 BENEFITS IN IA IN 2025 Microsoft 80,000 million 13,000 million Google 75,000 million 7.7 billion Amazon 105,000 million 5,000 million goal 72,000 million 3,000 million Green outbreak Yes, in singular. The latest results are hopeful, but we are very far from being able to say that AI is a profitable business, especially As far as generative AI is concerned. Good results are thanks to cloud services, chatbots or audio or video generators are not profitable. Subscriptions to these tools are a way to monetize, but The income they generate is child compared to spending. Despite doubts, unbridled expense has not stopped and this green outbreak can be more than enough for investors to continue throwing banknotes to the AI well. Image | Microsoft In Xataka | The AI industry has become a kind of ‘game of thrones’. And that reveals a worrying truth for your future

The AI is putting the US power grid in trouble. And Google has already taken a measure that shows the magnitude of the problem

by

Google data centers work 24/7, processing searches, videos and now also AI models. But not everything can grow at the same pace. In several areas of the United States, electricity begin to notice the pressure: Energy demand is accelerated and In some places already exceed capacity forecasts. Given that scenario, Google moves: It will reduce the consumption of your data centers when there are peaks, prioritizing the essential and postponing what you can expect. The novelty is the focus: Machine Learning charges. Artificial intelligence progresses. The electricity grid notice. The expansion of AI is going so fast that companies receive more connection requests than they can meet in certain areas. The consequence is no longer only technique: there is an energy restriction that conditions the deployment. It’s not about turning off machines, but moving loads. The “demand response“It consists of adapting consumption to what the network can supply at all times. In practice: displace or reduce non -urgent loads – like the processing of programmable videos or tasks – outside critical hours. It is a tool used in intensive industries and cryptocurrency mining, now applied to data centers with AI. The system has clear limits. This type of flexibility is not applicable in all centers or in all situations. Google recognizes it clearly: there are services that you just can’t expect. Platforms such as Search, Maps or the cloud for critical sectors – such as health or emergencies – require continuous availability, without margin for load settings. There are no “non -urgent” tasks that can be postponed. Therefore, although the response to demand is a valuable tool, its implementation will remain partial and selective. It requires planning, previous agreements and an infrastructure designed to absorb that type of reorganization. Not all centers can do it. But where it is possible, it becomes a real way to relieve pressure on the network without compromising the essential. There is already experience, and now. It is not theory. Google tested this flexibility With the public electric of Omaha and reduced demand associated with Machine Learning in three network events last year. The next step are formal agreements with Indiana Michigan Power (Fort Wayne) and with Tennessee Valley Authority: in Indiana it will be integrated from the beginning of the new center, and in Tennessee it will be applied coordinated with the operator. From experiment to strategy. What began as a pilot becomes operational policy: Managing demand flexible helps stabilize the network and accelerates the connection of large loads without waiting for new lines or centrals. It is not a magical solution, but it wins time while the infrastructure is reinforced. Images | Xataka with Gemini 2.5 Flash | Andrey Metelev In Xataka | Google has signed the largest hydroelectric agreement in history. You no longer know where to get more energy to feed your AI

Google has signed the largest hydroelectric agreement in history. You no longer know where to get more energy to feed your AI

by

Google closed on July 15 A historical agreement With Brookfield Asset Management worth $ 3,000 million to access hydroelectric energy for 20 years. This is the largest corporate cleaning of the world in this modality, which will allow technology to access up to 3 gigawatts of hydroelectric power in the United States. The magnitude of the problem. Artificial intelligence has triggered the energy consumption of large technology to levels never seen. According to The International Energy Agencyfor 2030 the United States will consume more energy by processing data than by manufacturing aluminum, steel, cement and chemicals together. There are studies that indicate that data centers could consume 945 theravats of electricity time globally that same year. Although of course, the same agency also concluded that the energy consumption of AI It is overestimated. What includes the agreement. The initial contract covers two hydroelectric facilities in Pennsylvania: Holtwood and Safe Harbor, which Brookfield acquired between 2014 and 2015. These plants will initially provide 670 megawatts of power, but Google will have the option of expanding the agreement to the 3 gigawatts through future updates of the facilities. At the same time, the company will invest 25,000 million dollars in data centers in Pennsylvania and neighboring states over the next two years. Why now hydroelectric energy. While previously technological ones focused on wind and lots, now They look for “firm” renewable energy sourcesthat is, they provide constant supply regardless of the weather or time. Hydroelectric energy fits perfectly in this category. Besides, Trump’s recent legislation He has maintained fiscal credits for hydroelectric projects until 2036, while those of wind and plot will lose these incentives unless they are launched before the end of 2027. The energy career. Google is not alone in this desperate search for clean energy. Meta signed an agreement for Buy the production of a nuclear plant In Illinois for two decades. Microsoft, Amazon and other technological giants also They are closing similar contracts. The pressure is such that these direct agreements have become a way that technology funds to new generation capacity, thus relieving pressures on electricity prices at home. What comes after. Google Plan Expand this model beyond Pennsylvania towards the Middle Atlantic and the US western environment. The company has already signed other pioneer agreements during the last year, including Geothermal energy No carbon and advanced nuclear. He is also working with the country’s largest electricity operator to use AI and accelerate the process of connecting new energy sources to the network. The arrival of AI in our lives has overcome any energy forecastand now it is the great technological ones that have to guarantee a constant supply of their AI at any price if they want the business to be profitable. Cover image | Greg Bulla and Nuclear forum In Xataka | The AI has folded the price of an ultrarrao metal. The problem is that we need it to store renewable energy

Now you can get the Google Pixel 9 at a good price, with greater capacity, thanks to this Amazon offer

by

Within the high range of mobile, the Google Pixel 9 It is one of those that has triumphed since the market was launched. If you have taken your eye but, until now, you had not considered buying it, this Amazon offer is sure to interest you. Specifically, you have this terminal available, in your 256 GB version Already a fairly attractive price. Specifically, it is available for 649 euroscompared to the almost 1,000 euros that usually cost. * Some price may have changed from the last review A top mobile now at a very attractive price Although the Google Pixel 9 It was launched in 2024, today it remains a reference within the high game. In addition, it is a safe bet, since it has guaranteed operating system updates For about six years. It has numerous features that stand out and one of them is its screen. This is of type Super acts OLED of 6.3 inches with resolution of 2,424 x 1,080 pixels. Its refresh rate is 120 Hz and is compatible with HDR10+. It works with pure Android operating system and the brain of this terminal is the G4 Google, accompanied by 12 GB of RAM and a Internal storage capacity of 256 GB. As for its photographic system, it is formed by a double rear camera of 50+48 MP. Finally, it can be noted that it has IP68 certification and admits fast charge to 45 W. Some accessories to protect this mobile that may interest you Jetech founded for Google Pixel 9 * Some price may have changed from the last review Jotech screen protector for Google Pixel 9 with camera lens protector * Some price may have changed from the last review Some of the links of this article are affiliated and can report a benefit to Xataka. In case of non -availability, offers may vary. Images | Pepu rica (Xataka Android) and Google In Xataka | The best mobiles, we have tried them and here are their analysis In Xataka | The best price quality price. Your analysis and videos are here

Europe wants users to verify their age to navigate the Internet. Google does not need it: you can estimate it with AI

by

In February Google advertisement They were going to start starting an experimental system based on AI and automatic learning (Machine Learning) for Estimate the age of users of its services. That technology, initially thought to avoid access to adult content on YouTube, will now serve much more, although for now it will be restricted to the United States. Automated restrictions. If Google detects that a user is under 18 in the US, it will apply the same restrictions that it applies when users are proactively identified as minors – for example, when creating Gmail accounts. Thus, it will enable reminders to go to sleep on YouTube, it will limit content recommendations and disable both personalized advertising and access to Adult applications in Play Store. What happens if AI is wrong. If Google’s system incorrectly estimates that a user is under 18, said user will be able to send a photo of his identity card (in the US, that of driving usually), of a credit card or even a selfie. First, that Google ends up having a photo of those documents It is quite disturbing. Second, the selfie option is especially surprising, because a priori might not be enough to verify that someone is not a minor. Remembering How-Ord.net. Microsoft a decade ago launched an AI experiment according to which a user could upload a photo of one or more people, and the service He tried to guess the age of those people. The tool quickly went viral, but it was also controversial because of its possible impact on privacy. A study Of 2022 he revealed that various systems of this type were quite precise when guessing age in young people, but not so much in older people. Companies specialized in this area as Yoti explain that The margin of error is already reducedand its system has an average 1.3 -year error in young people between 13 and 17 years old. How to estimate the google age. Google’s age estimation model will use the existing data of its users, including the websites they visit, what type of videos do they see on YouTube or how long their accounts have had. What about privacy. According to Google, this approach “does not imply the collection of additional data”, but of all that information that you already have thanks to the profiles that users use to access their services. They also ensure that they will not share that estimate with apps or websites. Goal already implemented A similar system on Instagram months ago. The age verification fever. The movements in this regard are increasingly striking. Before the excuse for companies and governments to spy on us and knew more about us was the terrorist threat. Now it is that minors They can access adult content. There is a unique fever for developing age verification systems everywhere, and we are seeing it especially in Europe … and the United Kingdom. The United Kingdom teaches us how the future can be. One of the first countries to move in this area is the United Kingdom, which these days these days has activated its regulation so that minors cannot access adult content. The age verification systems must be present in all types of platforms, something that has been especially controversial there and that has generated a surprising (or perhaps not so much) Collateral effect: British users They have begun to subscribe in mass to VPN services to avoid having to pass those age verification controls. The EU wants its own “pajporte”. The controversy introduction of the Digital portfolio beta In Spain —Popularly known as “Pajortport” – is one of the great steps that EU member countries are facing the future European digital identity. Among the options that will enable this mobile application will be to verify the age of the users, and in fact the Spanish initiative is serving as inspiration For a European system. Said system, yes, is not exempt from controversy. It remains to be seen if the Google system is really effective, and if it is, how its deployment progresses and what impact that has on the intentions of various countries to control the age of users. In Xataka | Allowing an app to “all photos” of our gallery looks like the most harmless action in the world. It is not at all

Working in Google, goal or Apple was the great dream of the newly graduated. Now they are afraid of them

by

According to A study The National Society of High School Scholars (NSHSS), in 2015 Google was the company in which millennials wanted to work as soon as they finish studying. Ten years later the situation It has changed remarkably. A few years ago, getting a job in any of the Big Tech was a pride. Not so much today. What happened. A new study of the NSHSS of reveals how a notable number of graduates no longer have large technology as the first options to seek employment. Working on them is no longer a dream, and their last movements make them rather a threat. Source: NSHSS. Big Tech don’t interest so much. In that study of 2015 Google was first among the companies desired by the newly graduated, while Apple was the 4th, Microsoft the 7th and Amazon La 11ª. In 2024 Google has passed to the 7th position, Amazon to the 8th, Apple at the 9th and Microsoft to the 16th. The interest in signing for these companies has been significantly reduced, and in the first six positions four hospitals, the FBI and NASA are now listed. Not to mention Spotify. The company Directed by Daniel EKwhich in 2022 was the third preferred by the newly graduated, is now in position 26, but there are other great displaced. Spacex has gone from position 9 to 22, Tesla from 15 to 33 and Airbnb from 27 to 49, for example. The threat of AI. In the study they point to clear reasons. One of them is that of fear of growing automation: artificial intelligence raises a remarkable impact for those positions that were previously covered by graduate young people. The same industry that previously promised innovation and a career with constant learning is now wrapped in frequent cases of sudden mass layoffs. Companies releasing ballast. The advance of AI both in Big Tech and other companies has allowed to validate the idea of trying to do more. Microsoft is in records of benefits and also dismissalfor example, but other important technological ones Like Amazon They have frozen their hiring while Google has opened the door so that employees who are not happy with the company abandon it freely. Volatility in these large companies is huge. The objectives change. Salary and prestige matter between millennials and gene Z, but according to a Network Trends survey cited In The Times of India, 76 & students put as great priority work stability above those metrics or even location. In fact, there seems to be a turn with a clear social purpose, such as those related to medical institutions. Doubts about whether or not to go to university. Another study Made by Deloitte It reveals how another of the foci of these generations is to continue learning, but not necessarily in academic environments. 31% of the Z Z and 32% of the millennials who participated in the survey indicated that they had decided not to go to universityespecially for the high cost of these institutions and for skepticism about whether that investment ends up being profitable. Changes in labor scene. The technological sector is especially irregular. Mass layoffs contrast with that new phenomenon of the AI Super. What a goal is hiring for hundreds of millions of dollarsFor example. The talent migration in the AI segment – with the money in between – contrasts with the questions that many recent graduates are asked now: what is the ethical impact of my work? Will it continue to exist within five years? There is a redefinition of ambition and the purpose of work itself. Not only should it be profitable, but also have a clear purpose behind. Image | Alex doubt In Xataka | We believed that AI was going to take our jobs. At the moment it has begun to whisper your boss who should say goodbye

A small town in Holland has the solution to mass tourism: to fool Google Maps

by

Mass tourism has unable consequences for residents. Recently we talked about the Avalanche of tourists who go to the lavender fields in Brihuega. In the Parkbuurt neighborhood, located in a small coastal town in Holland, they also have problems with tourism. Specifically because, in the most influx days, it is impossible for them to park. The neighbors have got to work to solve it with a very ingenious method. Fed up neighbors. Parking in Parkbuurt was not a problem until tourists arrived. Especially on weekends, its streets are filled with cars and neighbors, fed up with not being able to park normally, as they say in NH News. What they did was start reporting street blockages on Google Maps so that the app sending visitors to other areas. Google Maps to Rescue. It is the navigation app most used in the world And its function to report incidents on the road makes it perfect for this type of actions. A single report would not take effect, but with neighbors organizing to send reports at the same time the thing changes. The streets of the neighborhood appear as cut in the app and that makes visitors go to other areas. Answer. The trick has worked for them, but it has had some unwanted effects. Gert-Jan Bluijs, councilor of the municipality, is not funny and ensures that this measure has generated more chaos in other adjacent neighborhoods. The neighbors defend themselves from criticism ensuring that it is a peaceful measure that only apply on weekends, during the week they were not applying it. They see it as a way of exerting pressure to do something from the City Council. At the moment, what the City Council has done is to place a sign at the entrance of the town asking visitors to deactivate Google Maps. More cases. It is not the first time that similar tactics are used to combat tourist saturation. A few months ago something similar also happened in Holland, near the area of the Keukenhof Tulipanes gardens. This same year, we knew that Barcelona had eliminated a Google Maps bus route so that tourists would not collapse it. The result was good in that bus line, but others collapsed. The other face. We have seen that navigation apps can be a tool to divert traffic, but they can also be the cause of the problem. Years ago we told you how Waze’s suggestions ruined the tranquility of some neighborhoods. The app, now GOOGLE PROPERTYsuggests faster routes and diverts traffic through streets that used to be little busy. In Xataka | Iceland is so tired of tourism that it has decided to stop it drastically: fringing its visitors Image | Rudi Arlt in Pixabay

The Google Pixel 9 for less than 500 euros, new Nintendo Switch 2, Movistar Plus+ at half price and more. Hunting bargains

by

We return with a new hunting bargains in a week that has been both full of offers and new reservations in some quite expected products. Therefore, in this article we will review the five best chollos we have found throughout the week. Google Pixel 9 by 483.69 euros When using a coupon and when registering in the store. SMART TV Pack + Philips sound bar by 1,099.99 eurosa combo of a TV Oled and a sound bar with Dolby Atmos. Movistar Plus+ by 4.99 euros per montha new offer to access the huge catalog of the streaming platform. Lego Game Boy by 59.99 eurosa Lego set that adapts the iconic Nintendo portable console and can already be reserved. Nintendo Switch 2 + ‘Pokémon Legends: Za’: Za ‘: by 509 eurosthe new console pack can already be reserved. Google Pixel 9 Since its launch, we have seen many offers in the Google Pixel 9but few have been as good as the one that has now Mediamarkt. By 483.69 euros We talk about a fairly cheap mobile for everything that offers: a Excellent multimedia section by its screen and speakerssome cameras that continue to be at a high level, the software is very fluid and works especially well and will receive Android updates for many years. Of course, although it is on offer with a direct discount, to be able to buy it at this price you have to follow two steps: Register in Mediamarkt To receive a second discount. Use the coupon MMPIXEL15 To receive a third discount. * Some price may have changed from the last review SMART TV Pack + Philips sound bar Pccomponents usually launch the occasional pack that gathers several devices. This time, the combo we can find in the store is Smart TV Philips 55OLED819 and sound bar Philips Tab6309/10. All this for 1,099.99 euros. The TV comes with a 55 -inch OLED panel, has ambilight technology and is compatible with AlexaDolby Atmos and Dolby Vision. Instead, the sound bar offers excellent 320W power to 2.1 channels Thanks to including a wireless subwoofer, it is compatible with technology Dolby Atmos and allows connecting other external devices through Bluetooth or through its HDMI EARC port. Smart TV pack and philips sound bar * Some price may have changed from the last review Movistar Plus+ Movistar Plus+ He has launched its summer offer to subscribe to the streaming platform paying half. They can benefit old or new subscribers (Not current subscribers), lasts three months and the offer will be available until August 1. The usual price is 9.99 euros, but now you can get three months for 4.99 euros per month. Monthly subscription to Movistar Plus+ (for three months) * Some price may have changed from the last review Lego Game Boy The collaboration between Lego and Nintendo has re -paid off with the next launch of the set based on the iconic game boy. At the moment, it can be reserved in the official Lego store by 59.99 euros and will be launched on October 1. The design is very cool, but the most interesting and particular thing is that it includes two video games in Lego format (‘Super Mario Land’ and ‘The Legend of Zelda: Link’s Awakening’) that can be introduced into the console. * Some price may have changed from the last review Nintendo Switch 2 + ‘Pokémon Legends: Za’: Za ‘: If you plan to buy the Nintendo Switch 2 In the coming months, attentive because the second pack that includes both the console and the new video game has already been announced ‘Pokémon legends: za’. It is in reserve period in some stores like Mediamarkt for a price of 509 euros and will be launched on October 16. Nintendo Switch 2 + ‘Pokémon legends: za’ * Some price may have changed from the last review Some of the links of this article are affiliated and can report a benefit to Xataka. In case of non -availability, offers may vary. Images | Google, Philips, Movistar Plus+, Lego, Nintendo In Xataka | The best price quality price (2025). Your analysis and videos are here In Xataka | Better sound bars in quality price (2025). Which to buy and seven recommended models from 140 euros

Google has shown the future of our videos and photos in Google Photos. A future full of false videos

by

Google just announce what seemed inevitable after The power of its models of artificial intelligence (Ia) generative In photos and video: Creation of photos and videos in Google Photos from the images of our reel. A novelty that speaks for itself of the future that awaits our content. What’s new. As an AI giant, Google gradually carries all its new advances to the set of services of its catalog. Google Photos has always been a place to experiment with AI, as is the case with the Magic editor or mode Reimagine. Now goes beyond “small touch -ups”, and proposes to create short videos from the photos stored on the reel. At the moment there is not much to indicate to the assistant, who will accept as options “subtle movements” or the classic “I will be lucky” to create a video of up to six seconds. Another novelty of Google Photos is “Remix”, a function that transforms the style of the images we select, being able to choose between styles such as anime, comic, sketches. Google’s anime example remembers Ghibli at Chatgpt. It is a movement that remembers Ghibli moment of Openai, that “Servers melted. Unlike Chatgpt’s global availability, in launch the function will only be available in the United States. Where are we going. With the movement, Google turns a strong trend into personnel and mass: that of the generated videos that seem realistic And social networks and YouTube are flooded. And beyond fun, it presents serious consequences. In Tiktok, for example, there are more and more young people using their mobile to watch mental health videos, with a problem: The most viral are false. The platforms are becoming “Lays of AI“, and far from fighting these trends, they are amplifying them. Meta, for example, has integrated tools such as Advantage+ To create different versions of an ad with little investment. When they try to warn us that the content is created by AI, They are not sure what it is. In text, Openai had its own detector, but ended up closed after admitting that It was not necessary. Our false photos are already a millionaire business. Pieter Levels He is a well -controversial developer and entrepreneur who climbed very soon to the wave of the generative AI with two tools that generate huge amounts to be the only developer, and all using Models ‘Open Source’. One of them, with which he says to enter $ 139,000 Photoai.com. Its function is simple: after training a model with our selfies, we can create an AI model based on us, and then take pictures in the place or possess we want. It goes further that the proposal starts from Google photos, and is already able to make realistic videos of many types, including modeling sessions. Under the slogan “say goodbye to your photographer” points to a future (not very distant) where AI can take care of all the photos we want in paradisiacal enclaves. Fake places where to posture On Instagram, but without the need to move at home. The Spanish Freepik It also has Your alternative. And there are people doing chilling things: inventing their own memories From photographs. Apple plays to improve our memories, without creating them. With iOS 26Apple has presented a function that “gives life to our photos, but in a different sense. The company showed from its inception to Vision Pro as a platform to remember family moments with fidelity. Now it allows photos captured with any iPhone in normal mode become a spatial scene with depth (visible on our mobile, without the need for a virtual reality kit). Image | Xataka and Google In Xataka | How to create images from photos with artificial intelligence

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.