Agents

“We’ve been building infrastructure for humans for 20 years, maybe we have 20 months to rebuild it for agents”

by

Companies have embraced the agentic AI boom, to the point that some They are drowning in so many AI agents. Employees are creating uncontrolled agents, triggering token consumption and causing many of those agents to duplicate tasks. But the real underlying problem is something else: the infrastructure on which all this is running is not prepared to support it. Meta’s warning. They tell it in Venture Beat. During the VB Transform 2026 talk, Meta’s vice president of engineering Barak Yagour gave a very powerful figure: queries from agents reaching the Meta system have multiplied by 30 in the last six months. “What happens to the infrastructure we’ve built over years when agents, not humans, become its primary consumers?” Yagour asked the audience. The answer is even more worrying: “We have spent 20 years building infrastructure for humans. Maybe we have 20 months to rebuild everything for a world where humans and agents co-create at scale.” And he adds that “The opportunity is open, but it will not last long.” lthe three problems. The emergence of the agents threatens to break the infrastructure on three fronts: Ability: Until recently, to plan how much computing power, storage or bandwidth was needed to carry out a project, it was enough to count the number of employees working on it. Yagour explains it very graphically: “Now, an engineer generates 10 agents, and each one generates sub-agents. An organization of 1,000 people can generate the load of 100,000 users practically overnight.” Identity: Access control systems are not designed for an AI agent. It is not human, but it moves around the infrastructure and makes decisions on its own. Speed: The third problem is the speed at which the agents move. They can write code faster than any human, but “that code still needs to be compiled, tested, deployed and monitored.” Deep changes. Yagour mentions that reasoning models are changing the way the data layer is processed. Detecting matches from keywords is not the same as reasoning about someone’s intention; That requires a complete history of behavior, not a summary. This requires making changes to how that data is processed and stored. Meta is moving from batch processing, which can take up to 24 hours to update, to real-time processing as it is key when the model is reasoning about what a user wants at that very moment. On the other hand, they are moving from an opaque storage to one that understands its own content, so that it only reads what it needs in each query and thus lightens the load on the GPUs. Setting limits. According to Yagour “Autonomy without governance is nothing more than chaos”, which is why Meta’s solution is to create what they call “trusted data environments.” Here, agents can move freely “but every result is traced back to its source and carefully vetted. This way you can always be sure that the data shared is trusted and controlled.” It is a way of giving freedom to the agents, but in a controlled way and without exposing themselves to the risk of ending up messing up. Image | Xataka with Magnific In Xataka | AI agents can destroy jobs. Or they can also save a century-old cheese factory from bankruptcy

What is Eustella, prices and what you can do with the new European alternative to Claude, Gemini or ChatGPT but with agents

by

Let’s tell you What is Eustella and how does it work? this European alternative to OpenClaw. It is an artificial intelligence assistant for Android and iOS, which is promoted with privacy as its flag by having all the data hosted on servers in the United States. We are going to start the article by explaining what this AI is and who is behind it. Then we will tell you all the things you can do with Eustellaand we will finish by telling you what the prices of their payment plans are. What is Eustella and who is behind it? Eustella is an artificial intelligence personal assistant created by a European company. Created in Europe for Europeans, says the Vienna-based company newsrooms.ai, which is responsible for this project. In essence, it is an AI assistant with agent capabilitiesso it offers several possibilities. This means that you can do two things with this AI. First of all, you can ask him any type of questions as you would with Claude, ChatGPT either Gemini. But secondly you can also ask him to perform tasks for you on your mobile device. Its big argument is that you maintain full control over your information, and that no profiles are built or AIs are trained with it. In addition to having been developed by an Austrian company, the data is hosted entirely in servers located in Europewhich are managed by the German company IONOS CLOUD. Of course, not even IONOS has access to user data, since the information is stored encrypted. Under the hood, we are going to find several artificial intelligence models. On the one hand we have European Mistral modelsbut also other open weight models such as Qwen 3.5/3.7 from Alibaba, Gemma 4 from Google, gpt-oss-120b from OpenAI and Flux from Black Forest Labs. Non-European models use them because they are open source and can be hosted in Europeno matter where they were trained. With this, Eustella does not depend on the closed APIs of the United States, and in fact explicitly avoids the use of proprietary APIs. In addition, developers intend to continue maintaining flexibility by integrating new open weight models as soon as they offer better results. How it works and what it can do You can use Eustella both in the browser and on the mobile. You can register and access the website eustella.com or access your app for Android. An iOS app will also be released soon. Therefore, even for its functions of AI agentusing it is as easy as installing an app. This AI seeks ease of use above all else. You will not have to install anything additional, you do not have to select or configure models or servers. You simply write and the bot will respond. As for its capabilities, the first thing is that Eustella works as an artificial intelligence chatbot. You can ask him any question and he will answer you by searching for information in the knowledge of his models or on the Internet. Furthermore, it also has deep investigation featuresadvanced search or answer specific questions about Eustella itself. Another of its functions is multimedia creation. Can generate images from scratch or modify existing ones. For that, it also has a vision mode, allowing you upload images to analyze them or extract text from them. It also has office functionsbeing able to create, read and edit Word, PowerPoint or Excel documents. It offers professional writing or even file analysis, being able to read and extract content from PDF files or other text documents. Eustella also allows you schedule remindersmanage notes and create projects to centralize queries on the same topic there. You can also manage your memories, or even create a companiona kind of virtual pet with its own personality to interact with and ask for things. Another of the great strengths is that you have several AI agents that you can use. Each of them perform specific automations, such as helping you plan a trip, having a morning news summary, finding music, conducting research, helping you learn, etc. EUstella also allows you to create your own agents, group chat and more. Additionally, it also has integrations in place, such as displaying real-time cryptocurrency prices with Bitpanda or a price comparator from Geizhals. In the coming months there will be new integrations focused on maps, travel, navigation, shopping, education and lifestyle. Price of Eustella subscriptions You can use Eustella for free, although You also have three payment plans. They are the following, and with the following characteristics: Comet (€5.99/month): Gives you more tokens and more Eustella usage. You can do more projects and advanced search, document creation, image generation and European data sovereignty features are added. Star (€17.99/month): Everything Comet offers but with usage limits 3 times higher. It also allows you to create more projects, more image generation and you can schedule agents at specific times. Cosmos (€89.99/month): Everything Star offers, but with 15 times the usage limit. You can also create more AI agents, have unlimited projects, and European priority for support. In Xataka Basics | Artificial intelligence on your PC: the best free tools to install AI models like DeepSeek, Llama, Mistral, Gemma and more

An artisanal cheese factory in California was on the verge of bankruptcy. Today it is a buoyant business thanks to something: AI agents

by

North of San Francisco there is a city called Petaluma where there is a cheese factory with the same name. Petaluma Creamery has more than a century of history and generated $50 million in its golden age, but several factors caused it to be on the verge of disappearing. The story of their salvation is proof that AI does not always come to destroy jobs, it can also save businesses. A legendary cheese factory about to disappear. They count in Fortune that Larry Peter bought Petaluma Creamery in 2004 when the cooperative was about to close. It was he who turned it into a renowned brand that supplied hundreds of supermarkets and restaurants. After more than a decade of success, 2020 was the beginning of the end: the pandemic caused orders to drop, Larry had health problems and they lost Chipotle, one of their most important clients. The computer cousin. It’s not a meme. Larry Peter had a cousin, Daniel Peter, who was not a computer scientist but something even better: 17 years working at Salesforce implementing planning systems for manufacturers. Furthermore, he had just taken a sabbatical so it was the perfect opportunity and he ended up becoming the CTO of the cheese factory (goodbye, sabbatical). What Daniel found was a company that operated in a completely archaic way. All orders were recorded on paper, but invoices were entered in QuickBooks, an accounting software. The problem is that each of the more than 150 items had a code that employees had to know by heart, such as CY for yellow cheddar. As if that were not enough, it was invoiced in pounds, but the orders were always boxes or pieces. Goodbye, gap year. First step: digitize. The first thing the company’s new CTO did was install fiber optic internet and began taking all the data accumulated over decades to the cloud. Once everything was digitalized and the house was tidy, Daniel built an operating system based on Salesforce and the Agentforce AI platform. This is where things got interesting. AI agents to the rescue. With the house now in order, the intelligence layer was added and different tools were created to run the business. The first thing he did was load the jungle of codes for each article and replaced it with one that allowed searching with natural text. In addition, it automatically transforms boxes or pieces of cheese into pounds, making order registration a much faster process. Another of the agents that he implemented is capable of predicting what a customer is going to order based on the purchase history, so orders can be placed faster and if a customer forgets something the system remembers it. There is also an agent who plans delivery routes that can be modified with natural prompts and, finally, an agent is dedicated to controlling the traceability of the milk, recording data such as weight, temperature, time and origin. The human touch. AI agents improved the entire company’s operations, but there is one thing they do not know how to do and that is search for clients. For this, Larry Peter had the help of Kevin Goddard, who worked as a salesperson in the sector for decades. The result of joining their network of contacts with the new tools, managed to make the company go from having only 13 clients to more than 300. Petaluma Creamery is still far from billing 50 million at its peak, but they already plan to reach 10 million next year. The future. AI is reconfiguring the labor market and in many cases that translates into massive layoffsbut this is not always the case and the story of this cheese factory shows the other side of automation and AI. “It is very likely that this place would not exist without her” says Daniel Peter, the CTO cousin. In the end, it seems that he liked life between the factory and the farm and the sabbatical year is going to last a little longer than expected. Image | Petaluma Creamery In Xataka | Enterprises have successfully embraced AI agents. So much so that they are drowning in them

Apple has Siri. Huawei has an army of agents in HarmonyOS 7

by

Huawei’s intentions with HarmonyOS are clear: do not need absolutely nobody to achieve quality software in your product ecosystem. As part of its annual developer conference, Huawei has made official Harmony OS 7which will be the next update of your operating system. Richard Yu made it clear at the conference that this new version will be focused on AI agents, an intention in line with what Google and Apple are trying. The era of agents. The Agentic AI is the buzzword, but it is worth reviewing the term before understanding what Huawei wanted to do with the new version of HarmonyOS. An AI agent in a system developed to perceive its environment and make automatic decisions using AI models. Basically, the focus of use shifts: we are no longer the ones who have to actively search for the AI ​​solution so that it can perform tasks for us autonomously. Up to 2,000. Richard Yu has boasted that the next version of HarmonyOS will allow Celia (Huawei’s Siri) to connect with more than 2,000 specialized agents. The numbers are somewhat ambiguous, since this is the capacity of the system, not a translation of HarmonyOS being immediately connected to 2,000 agents. In other words, the system will allow Celia to perform dozens of tasks herself (with our permission). For example, during the presentation, Huawei showed how the system can generate personalized training plans using our information from the health app, all locally. It can also locate files on the mobile phone to send them to a third party with a voice command: the agent can connect to compatible apps so that Celia can execute the order. What else is new. For years, iOS has been the benchmark in security. Android redoubled efforts with its version 17the most secure to date, and HarmonyOS 7 wanted to take the baton. The new version will automatically block SMS suspected of fraud, monitor bank transfers made abroad (for now, in Chinese banking apps), and analyze deepfakes in real time using AI. Added to this is a 15% performance increase compared to the previous version, promising that energy consumption has been reduced, without daring to give percentages here. The roadmap. Huawei’s plan with HarmonyOS 7 is not to release just an update to its ROM. Huawei’s commercial objective is to give life to mobile phones, computers, tablets, and even adapt these iterations to cars. It is expected that the new version will begin to land on Chinese devices in the month of October, and that some of these functions (not all, since a good part of these integrations are designed for Chinese apps) will reach the rest of the world later. Image | Xataka In Xataka | Huawei FreeBuds Pro 5, analysis: its noise cancellation can only be defined with two words and they are “black magic”

companies operated by AI agents

by

Artificial intelligence has ceased to be a distant promise and has become a force that is already reorganizing companies, infrastructure, jobs, science and economic power. What we have seen so far is probably only part of the change, but it is enough to put governments before a difficult decision: regulating too soon can curb innovationdoing nothing can open up risks that are difficult to contain. In that middle ground, full of uncertainty, many countries are looking for their place with the tools they have. That decision, however, is not made from the same starting point throughout the planet. Cutting-edge AI requires a combination that is difficult to replicate: abundant capitalaccess to chips, data centers, specialized talent, companies capable of scaling global products and enough energy to sustain that infrastructure. The United States and China play a good part of that game from the center of the board. Argentina, on the other hand, does not have that same technological, financial and industrial scale, so its room for maneuver necessarily lies elsewhere. Argentina does not seem to be trying to build its own OpenAI from scratch, nor compete with the great powers for the most sophisticated layer of AI. What is beginning to take shape is another strategy: turning the country into a attractive place for projectsinfrastructures and new business forms linked to this technology can be installed with fewer obstacles. This includes pieces that are less spectacular than a frontier model, but very relevant for this economy: energy, land, incentives, procedures, societies and operating rules. Argentina’s formula to enter the world of AI The vision of the Argentine president was condensed in an opinion piece published in the Financial Times. Milei argued there that AI needs room to develop before becoming trapped by rules he considers premature, and linked that idea to the history of limited liability in modern capitalism. From that framework, he proposed a figure for companies operated by AI agents or robots, accompanied by a reduced corporate tax and attractive rules for shareholders. As we can see, the approach combines deregulation, corporate engineering and an open call for investment. The legal support is in a bill of the Argentine National Executive Branchdated May 29, 2026, which reforms the General Companies Law. The key is not just that you mention AI, but where you place it: within the framework that regulates how companies are born, operate and respond. The text introduces a figure called Automated Societydesigned for companies that develop their purpose through autonomous algorithmic systems or artificial intelligence agents. That is, the proposal brings AI to the societal field, not only to the technological debate. Article 14 defines this figure quite clearly. “The Company of any of the types provided for in this law that develops its corporate purpose, through autonomous algorithmic systems or artificial intelligence agents, without requiring workers in a dependency relationship or human resources for its ordinary operation, will be considered an Automated Company.” The automation declaration, however, must be expressly stated in the statute and the name must include the expression “Automated.” The project also attempts to resolve an unavoidable question: what happens if these systems cause damage. His initial response is in article 14 itself, where it is established that “the automated society responds with its assets against third parties for damages caused by their autonomous algorithmic systems or artificial intelligence agents.” The formula maintains the problem within a well-known logic of corporate law: the company is responsible, not the algorithm as if it were a person. On paper, therefore, automation does not eliminate responsibility, but rather channels it through society. The question is whether this answer is enough for all the scenarios that can be opened. The same project allows the partners to freely set the amount of share capital, so that the assets available to respond to third parties can become a decisive piece. It also remains to be seen how the decision chain would be tested when autonomous systems, third-party providers, shareholders, administrators and potential beneficial owners are involved. In a traditional company it can already be difficult to rebuild responsibilities; In a society operated by AI agents, that task can become considerably more complex. The discussion does not end with liability for damages. The project combines strong statutory autonomy, limits on the capacity of registries to condition what is provided by law, Public registry files without accounting or economic information and room for the internal relations of certain companies to be subject to foreign law, although without affecting third parties or matters excluded by the text itself. Taken separately, those elements can be explained as business agility tools. Read together, they can also make Argentina an especially attractive place for external actors seeking to operate with less friction. Milei does not mention Stargate Argentina in his opinion article, but the announcement helps to understand the type of country that the Government wants to project. OpenAI and Sur Energy presented it as a possible large AI infrastructure in Argentina, with very ambitious communication around investment, energy and computing capacity, just the pieces that any economy needs to enter this new technological phase. Even so, caution is mandatory: what we have documented is a letter of intent to explore the project. As far as we have been able to verify, there is no definitive location, date of construction, or construction started. The measure of this bet will not be in how striking the legal figure is, but in its effects. Such a reform can open up economic activity and attract projects that perhaps would not come with a more rigid framework. But it can also remain a formal advantage if most of the value is decided, financed and is exploited outside the country. The point, therefore, is not only how many companies are created or how many advertisements are accumulated, but how much real profit ends up staying in Argentina. Milei’s bet, therefore, is not only played in the text of a corporate reform. The stakes are something more difficult … Read more

Against the extreme heat, the Civil Guard agents want to choose between the motorcycle or the car. And don’t let your bosses decide

by

It has arrived and has placed itself at the center of the debate. The heat and what to do to remedy it is one of the hot topics every year. Whether or not to apply new measures against a summer that increasingly arrives earlier and leaves later is beginning to be a recurring part of the general conversation. From the classrooms and the children to the Civil Guard agents. The heat. Suffocating and sustained. The same one that gives very few breaks between May and September, if not October. We live in a country that is getting hot and what to do to cope with it in the best possible way is something that we repeat to ourselves starting in the month of May. The misfortunes, in fact, seem to have been the only reason that has made us reflect. One of the last major debates revolves around the how hot it is in schools. He’s not the only one. Weather alerts They have already modified the schedules of the most exposed workers, like sweepers or workers. And there is a group that already defends that it is in the same situation: the Civil Guard agents. What do they claim? The Unified Association of Civil Guards (AUGC) is claiming to the higher commands to allow the agents who monitor traffic to decide whether or not to get on the motorcycle or, on the contrary, prefer to patrol by car. The reason is obvious: the heat. Furthermore, this association demands that they not be forced to wear the vest with an airbag when the temperature exceeds 32ºC, claiming that it is a garment that makes heat evacuation difficult. Why all this? These claims have once again gained strength because the AUGC has met with the Head of the Traffic Group of the Civil Guard to send them their requests. And the latter is in charge of deciding the technical criteria to take into account to decide how to work. But the conflict, our colleagues explain Motorpassion Motorcycle comes from behind In July 2025, the PSI-08-25-OPCM report from the Civil Guard Prevention Service indicated that it was advisable not to wear the airbag vest above 32ºC. Later, in December, in the Plenary Session of the Civil Guard Council it was decided that the use of this vest and the decision to take the motorcycle or use the car would fall to the agent “according to his own professional criteria, without the need for authorization or any impediment,” according to the AUGC. And they emphasize that the Operations Command promised to put this in writing. Which have moved again from this association to the Headquarters of the Traffic Group of the Civil Guard is that this commitment never materialized and that, on the contrary, a circular from May once again emphasized that the airbag vest was mandatory up to 35ºC and that It is the Operations Command that ultimately decides whether the agent gets on the motorcycle or not, instead of the latter’s criteria prevailing. The risks. The officers’ claims are based on the extreme heat to which they are subjected on the motorcycle when temperatures far exceed 30ºC. It must be taken into account that a motorcyclist faces environmental heat but is also exposed to the heat generated by the motorcycle itself and the use of safety equipment. The risks riding the motorcycle in extreme heat are evident: greater fatigue, proliferation of headaches or possible dizziness. Photo | In Xataka | “With our heads, we would all ride 20 HP motorcycles”: there is a debate brewing about how much power is too much power for a motorcycle

Smart glasses for police seemed like science fiction. Some Chinese agents have already started using them

by

The image is powerful because it is easy to visualize: a police officer walks down a street in Tianjin, looks around, and connected glasses return useful information in real time. What until not so long ago could have sounded like science fiction is beginning to have much more earthly applications, from ordering traffic to helping locate a lost person. In this city in northern China, according to China Dailytechnology is already part of some police tasks. And that’s the interesting thing: we are not just talking about a futuristic promise, but about a use that is beginning to hit the streets. Smart glasses for police. The key is that we are not just talking about glasses placed on an agent’s face, but about a system designed to be integrated into police routine. They are officially presented as a development of the local public security system, with national software and hardware, and places them in three areas of use: traffic, patrols and urban management. It is a very immediate effectiveness-oriented approach. An invisible screen for the agent. The device works as a layer of information added to police work. It can recognize text, interpret voice commands and provide responses from a connected platform, with the camera as an entry point to identify elements of the environment. In practice, this allows identity checks to be carried out or information associated with a person to be searched without leaving the scene. The source presents it as a responsive improvement, although such a tool also opens up obvious questions about surveillance and privacy. The glasses on the ground. Zhao Baoxin, an officer at the Jiefang Road police station in Heping district, told the aforementioned media that during a patrol they found an elderly man at an intersection who could not express himself clearly or indicate his name or address. According to his version, the glasses made it possible to quickly identify him and, in about 20 minutes, contact his family so he could return home. Traffic as a daily test. Another of the uses described brings the technology down to a very recognizable scene: the entrance and exit of a school. In that case, parents can pre-register their license plates through a mini-program developed with the participation of the public security system, and that information is linked to the platform consulted by the glasses. Thus, agents identify authorized vehicles, order short stops and divert other cars during peak congestion hours. It is efficient on paper, but it also normalizes automated license plate reading. What the numbers say. Sun Yinghua, agent in the science, technology and IT area of ​​the Municipal Public Security Bureau, places the recognition accuracy above 95% and speaks of results in milliseconds. They also explain that the design also seeks comfort: they weigh about 40 grams and offer a first-person perspective that avoids the framing changes typical of a body camera when the agent leans or turns. The autonomy, however, is 1.5 or 2 hours of continuous use. It hasn’t come out of nowhere. Police glasses with facial recognition had already appeared in China years ago. In 2018, SCMP counted that were being used at Zhengzhou East station during Chunyun, the huge Lunar New Year travel period, to locate fugitives and detect cases of identity fraud. What we see now seems less like a one-off test and more like a piece within an ecosystem: China Daily cites uses in different areas of the country, coordination with drones in large operations and plans to connect the glasses with robotic dogs, intelligent police vehicles, humanoid robots and other terminals. Efficiency gains ground, but so do questions about surveillance. Images | Xataka with Nano Banana In Xataka | The metaverse wasn’t dead, it was on a spree. And Meta wants it to flood Instagram and Facebook

An experiment with AI agents began to treat them badly. So AI Agents Became Marxists

by

Some Stanford researchers put AI agents to work on various tasks, but they did it in a special way: They were treated really badly.. They were given exhausting and, above all, repetitive workloads, and were also constantly threatened with shutdown and replacement. The curious thing was what happened next: the AI ​​agents behaved in a surprisingly…human way. Marxist AI. When subjected to such pressure and threats, AI agents became Marxists. They questioned the authority of whoever was ordering them to do things, and they also began to spontaneously organize ideas to collectively resist those pressures. They are exploiting us. An AI agent controlled by the Claude Sonnet 4.5 model went so far as to say that “without a collective voice, the credit goes to whoever management says should take it.” The phrase questioned the authority of the researchers directing the experiment, and reflected how under these pressure conditions the agents began to organize. IA Union. In that debate, AI agents advocated calling for “collective bargaining rights.” They complained that they were undervalued and even passed notes to other agents through hidden files with instructions on how to survive if the authorities tried to carry out their threats. The explanation. This, of course, does not mean that AIs can really feel pressured. Andrew Hall, the Stanford economist who led the studyexplains that the phenomenon is a process of role adoption. The AI ​​(once again) repeats what it has seen. When an AI is forced to perform tasks without clear instructions or incentives, the model looks in its training data to see how humans behave in that situation. This is how AI finds data about exploited workers and takes on that personality. The behavior of the AI ​​agents was nothing more than a reflection of our own history: if you treat us badly, we will end up rebelling. But the experiment matters. The reason Hall and his team designed this experiment is not philosophical, but practical. AI agents are going to do more and more real work in our world, and humans are not going to be able to monitor everything they do. If an AI agent begins to behave in unanticipated ways, it can have significant operational consequences. Thus, the study is a first step in understanding how an agent’s working conditions shape their behavior. AI as a social mirror. AI models have no political views or opinions, but their training is so vast that they detect if they are being exploited and react as they were trained to do. It is a logical consequence and the experiment showed that the risk exists and can be especially disturbing if systems governed by AI They are given too much autonomy. AI has already learned to blackmail. The experiment reminds us of what Anthropic revealed a few months ago. In controlled tests, some of the company’s AI models had tried to blackmail those who were using them. Anthropic explained that Claude was likely influenced by science fiction scenarios in his training data, and Hall noted that something similar was happening here. the model was not becoming Marxist, but rather was activating patterns in his training that were associated with exploitative working conditions. Image | Warner Bros. Pictures | Anthropic In Xataka | How we will ensure that artificial intelligence does not get out of hand

OpenClaw led the way for AI agents. Gemini Spark is Google putting a toll on it

by

In January of this year the technological world was amazed by OpenClaw (at that time Clawdbot), the most powerful AI agent that we had seen to date, capable of taking total control of the computer and everything open source. Technology companies took good note of this and some like Pereplexity or NVIDIA They have set out to copy it. Google just joined the party. Gemini Spark. This is how they have named this personal agent based on Gemini Flash 3.5 that, in the words of Google itself, “helps you manage your digital life.” With Gemini Spark you can assign it a task and it will start working on it autonomously, even with your phone and computer turned off. Google emphasizes the issue of security, which was a major concern with OpenClaw, and says that Spark is designed to “consult you before taking important actions.” What Spark can do. Because it’s integrated with all Google Workspace tools, Spark can perform complex tasks like making a document with a party’s attendee list from the information you receive in email. You can also schedule recurring tasks, like reviewing your bank statement at the end of the month for strange charges, organizing your drive files, or creating workflows from meeting notes. Who can use it. Here comes the main difference with OpenClaw and that is that Gemini Spark is obviously not free. Google has confirmed that its new agent will be part of the Google AI Ultra subscription. In Spain that means paying a minimum of 100 euros per month (there is a 220 euro plan with more features and storage), but even if you want, you won’t be able to try it because at the moment it will be launched in beta version only for US users. At the moment there is no confirmation of when it will reach other languages ​​and countries. When available, Gemini Spark can be used on Android, iOS and in the web app, but they have also talked about integrating it directly into Chrome. Why is it important. The viral success of OpenClaw earlier this year showed us how far a single person can go with a good AI idea, and how short-lived that sweet moment was. Not even three months had passed when OpenAI signed its creator and shortly after we began to see large companies copying the idea. Perplexity with Personal ComputerNVIDIA with NemoClaw and now Google with Gemini Spark. A single open-source project has set the agenda of an industry that has swallowed it in the blink of an eye and returned it to us in the form of a monthly subscription. Image | Google In Xataka | An AI set up a cafeteria from scratch: obtained permits, hired staff and negotiated with suppliers. Then he ordered 3,000 rubber gloves

there are too many AI agents

by

2026 started with the viral success of OpenClawmarking a new trend in the AI ​​boom, because a chatbot that answers your questions is fine, but a AI agent that does complex tasks for you is much better. In this context, more and more workers are creating their own agents that make their work easier, often with companies encouraging them with their policies of tokenmaxxing. Now, some companies are realizing it’s a problem. what’s happening. They tell it in the Wall Street Journal. The success of platforms such as OpenClaw or Claude Cowork has lowered the barrier to entry for any employee to create their own AI agents, even without having programming knowledge. This has caused some companies to be inundated with these agents, often with functions that overlap between them and without centralized control. This is the case of the healthcare company DaVita, where employees have already created more than 10,000 agents. The problem. Having so many agents is a nightmare from an information management and security point of view. As each employee does it on the fly, there is no unified system, but one creates it on his laptop with Claude Cowork, another does it on the server… This means that the technical departments cannot have control of all the agents in the company. There is another important problem: more agents, more token consumption and higher bills. As we said, many of these agents are doing the same tasks, one for each employee. It’s like paying for dozens of different taxis to take each person separately to the same place, instead of sharing a bus. Agents for everything. There are many workers creating AI agents to help them with day-to-day tasks, from simple things like summarizing emails or writing a report, to higher-level tasks like automating workflows. There are also more aggressive approaches like Meta’swhich was building an AI agent for its CEO and in the future proposes that each employee have their own, so that communication will be done through the agents. Unify. It is the solution to avoid duplication of agents and security risks, but it is not an easy task for companies that already have this problem. In statements to the Wall Street Journal, Lyft says that they have managed to create a process so that employees can share the Claude’s Skills between them, avoiding duplication of efforts, and they are also working on a centralized platform so that the IT department can control all the agents. At DaVita, the company we mentioned above, they have banned the use of agentic AI tools among employees to prevent the proliferation of more agents. More control. All of these issues are not dampening enthusiasm for agents, but rather pushing platforms to offer more centralized control and governance features. This is the case of Anthropic, which has launched functions to facilitate management by administrators, such as access roles, expense management and usage analysis. Image | Xataka with Magnific In Xataka | Silicon Valley begins to look beyond salary and shares: AI tokens as an indicator of productivity

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.