programmed

The AI ​​Claude Code “only” programmed. With Cowork, Anthropic wants its AI to take care of everything else

by

Claude Code has become a revolution for programmers, but at Anthropic they are not satisfied with that, and now they want their Claude family AI models to serve much more. And that’s why have created Coworka different agent, especially ambitious and who opens the door to fantastic options… if you trust him. What is Cowork. Those responsible for this project have taken the foundations of Claude Code and applied them to the Claude desktop application (for now, only the macOS one). But they have also done something equally special: giving Claude permission to access a specific folder on our computer and, from there, he can take control of those files and work with them as we want. Hello, robot-secretary. Instead of access to the “vibe coding” we will have access to a kind of “vibe working”. Thus, we can ask Cowork to do all kinds of operations with those files: If we have a folder full of disorganized icons, we can ask you to ordered them to us and reorganize them all into folders by file type or theme If there are a lot of photos of receipts in that folder, we can tell you to create an expense report If what we have is a bunch of digital voice or text notes, we can ask you to write a report summarizing and combining them all. If we have a folder full of podcasts, we can have it go through it, analyze it and summarize the top 10 points of all of them or transcribe them If you have all your financial trading and investment reports and data, you can ask them to create a final report for you. help you declare them If you have videos and want to find one of a squirrel and then convert it to another format, also does. Full autonomy. We are therefore faced with an AI agent capable of accessing our files, analyzing them and working with them to generate new information and useful content from all that data. And we only have to ask it with natural language, because the agent is capable of understanding it, asking us questions if it needs more details, and then solving the task autonomously even if it involves several steps. Cowork operates in a container. The way CoWork works allows you to grant permission to certain folders, but when the AI ​​operates on said files it does so in isolation. As explains Simon WillinsonClaude uses a virtual machine and downloads and boots a custom Linux file system to operate on those files independently and isolated, which theoretically guarantees that our files are theoretically safe and Cowork does not access anything that we have not given permission to. Connections to other apps. In addition to being able to work directly with your files, Cowork benefits from its ability to connect with other applications that you have installed on your computer. You can use ffmpeg to convert the squirrel video, Asana if you want to organize your notes into projects, or an office application if you need to create a spreadsheet. But we will have to trust. Willinson himself warns that these types of systems have the danger of someone “hacking” them with jailbreaking or prompt injection techniques that now become more dangerous because, as we say, what Cowork does is work on our files. And of course we have to be careful with the information and data we share with CoWork: those responsible for Anthropic themselves have a document to “use it safely“. Limited release. Cowork is available as a “research preview”, and is only available to users of the Claude Max subscription which costs between $100 and $200 per month. It is clear that at Anthropic they prefer to go step by step with a very powerful but also delicate feature if we do not use it with caution: in the end we are giving access to our files to an AI, and we know that AIs can make mistakes. An AI on your computer. This release from Anthropic points to what all AI agents that want to conquer our computer should theoretically point to. Since that Computer Use that Anthropic launched in October 2024, things have come a long way, and little by little we are getting closer to that future in which we will be able to work with our computer in a very different way than we did until now… if we want and trust AI, of course. In Xataka | Operator also “looks” at the screen and moves your mouse for you like other AI agents. It does it better thanks to CUA

A teenager discovered the ‘Málaga’ virus and ended up founding VirusTotal. The enigma that remains is the same since 1992: who programmed it

by

Bernardo Quintero (@bquintero) was 14 years old and his first PC, an Amstrad PC-1512, had just arrived home. It was 1987, and the co-founder of VirusTotal He was excited by this machine that allowed him to exploit his computer curiosity. His hobby ended up being trying to circumvent the copy protection systems of some games, and he was there one day when something suddenly happened. A little white ball moved on your screen. By itself. Without him having done anything. He soon discovered that it was a computer virus. One that he ended up studying to know how to detect and eliminate it. He succeeded, and over the next three years he ended up improving his first antivirus, a tool that allowed him to recognize and eradicate seven different viruses he had encountered. It didn’t seem like that project was going to go much further, and Quintero began his studies in Computer Science at the Polytechnic University School of Malaga. In one of the first classes, a professor asked if anyone wanted to raise a grade with a Pascal programming project. He signed up, and when talking to the professor, he asked him if he had done any previous projects. “Well, yes,” he replied. “An accounting program, disk utilities, an antivirus…”. The teacher cut him off. “Did you say antivirus?”. When he answered affirmatively, the professor asked him to accompany him to his office. There he showed him how the entire IT department had been infected by a virus that the antivirus did not recognize. Fragment of the code in Turbo Pascal 5.5 of the antivirus that Bernardo Quintero developed to eliminate the “Málaga-2610” virus (1992). Source: Bernardo Quintero. Quintero soon detected where the problem could be and went home with an infected disk to work on an antivirus. It took him more than he thought, but after a few hours he managed to figure out how to detect it and delete it. That helped him pass the subject, but it also ended up being the definitive seed of the professional project that would end with the founding of Virus Total. He tells it all in more detail in his novel, ‘Infected‘, which he published at the beginning of the year and in which he narrates those beginnings and how that ended up leading him to create VirusTotal, the Malaga company that would later end up being bought by Google. That virus in his faculty was called “Málaga”, and Quintero spent years without paying much attention to it again. So, three years ago, this expert posted a message on Twitter (X) to try to solve the mystery of who would have created it. Already then he discovered that according to several sources the virus had been created at the Polytechnic School of Informatics. The objective, I counted thenit was not about bringing the name to light, but about chatting with that person and remembering those times. He failed to reveal the mystery, and that mystery remained unsolved again. But Bernardo Quintero never forgot that and returned to the fray with a new attempt a few days ago. After first publishing a message on X, the next day he published a summary of that story on LinkedInand asked for help in that post to try to solve the mystery once and for all. We contacted him, and he told us how while in the past he had focused on discovering how it infected and creating the disinfection tool, he never tried to find out who had created the “Malaga” virus. But he told us that “now, looking at it with new eyes, I have seen a couple of interesting details and I have discovered the motivation.” In fact, he adds that thanks to those messages on X and LinkedIn “I have received stories from several people who studied those years at the Polytechnic of Malaga and who believe they know the author.” Of those candidates, he explains, “I have ruled out 3 or 4, but there is one that fits very well with the new data I have.” The mystery seems to be close to being solved. “I just need to clear up one unknown to confirm the author.. And if it is confirmed, there is a beautiful and sad story that will be worth telling.” Everything therefore indicates that it will finally be known who was the author of that virus, and Quintero has promised to tell more details these days. We will be attentive. Image | Mika Baumeister In Xataka | The computer with the most malware in the world: this is MICE, the challenge of Bernardo Quintero and VirusTotal

This genius has programmed animal crossing characters to play alone. The first thing they have done is rebel against the landlord

by

The use of AI opens new exploration (and exploitation) routes for the Modders that manipulate the code of games so that they run into new expressive routes. The last test: Thanks to an advanced memory hacking technique and artificial language models, the residents of ‘Crossing animal‘They comment today, conspire against Tom Nook and explore new dialogue styles, in which there are no lack of social criticism. What have you done to you, animal crossing. The josh fonseca hacker used the Dolphin and Scripts emulator in Python to replace the text messages saved in the RAM of the Gamecube and that function as dialogues during the games (and it was documenting on YouTube). Since the game uses its own text coding system, it developed decoding tools that allowed information to an understandable format for a couple of extensive models of GEMINI -based language (one to write the dialogues themselves, another to give them intention, emphasis and “direct” the charts of the characters) that began to generate new dialogues. In order for the dialogues to be consistent with the game, the model was trained with specific examples and the animal crossing wiki was consulted to create character sheets of a certain complexity. It is an experiment that made, at a minor scale, in ‘Stardew Valley‘. And what happened. In a few moments, the villagers did what they usually do in the game: conversing with naturalness and humor, but without registered in the real world. The thing changed when a Feed RSS, which allowed them to discuss political and current issues. Soon they started talking about Trump very positively, comically positive in fact. The reason: the feed that Fonseca used was of Fox News, simply because he was the one who had the most at hand. This behavior of the characters makes sense: if the knowledge they acquire includes those that belong to the real world, the machine has no way to distinguish them from their fiction. But the next step was even more unexpected. To the barricades. Fonseca also installed a shared memory space thanks to which the characters could gossip. That is, remember the other characters, what they had said and what they felt for them. They began to plant rumors and criticism against Tom Nook, the iconic and discussed Magnate Tanuki of the game. The seed of the rebellion was soon planted: the characters realized that one of the key elements of the game, the high prices imposed by Nook, are an economic spring that forces them to work and trade continuously to pay debts. Suddenly, ‘Animal Crossing’ became a spontaneous reflection on undeniable traps (because if they end, there is no game) of capitalism. It is not the same, but it is the same. In one Interview with 404MediaFonseca seems to recognize that the experiment can alter the original spirit of the Nintendo game, but considers that its experiment is a powerful exercise inspired by hauntology (philosophical current that studies the possible future that never came to complicate). In addition, it states to be convinced that well used is a tool that can enhance art and creativity, as long as human screenwriters continue in the center of the process. The power of the past. Fonseca thinks that there is nothing like the games of yesteryear (something that is perceived on his YouTube channel, overflowing with nostalgic videos): “As a child I thought: ‘Video games will improve every year’. But after twenty years playing, I have become a little skeptical and think: ‘In reality there has not been so much innovation. Technologies of the future that interest me. Header | Josh Fonseca In Xataka | Someone has cut in half a playstation to create a portable version. The result is fantastic

Tesla Robotaxis have been programmed to drive as a human. So when they see the police hit a brake

by

In addition to taking the first steps in a new market, the appearance of Tesla’s robotaxis in the streets of Austin leaves us many other readings. Among them how the company has approached autonomous driving, the use of artificial intelligence and how it takes advantage of its millions of cars in the street to advance faster and spending less money than competition. And that has its consequences. An unexpected brake. A Tesla Robotaxi that exceeds a crossing, which marches at the expected speed and that, without prior notice and without apparent reason, slows hard until it almost stops completely. What makes the car doubt? To the right of the intersection, in a street perpendicular to the march of the Robotaxi de Tesla, a police patrol is stopped. This does not hinder the march of the autonomous vehicle but, despite this, the car stops strongly. The video can be found on YouTube After the up of a user and Antitesla accounts have replicated them in X. Who has uploaded the video is Edward Niedermeyerjournalist specialized in the motor world in the United States and author of a book in 2019 on the birth of Tesla and the figure of Elon Musk, especially known for its critical positioning with the brand. Click on the image to go to the original tweet Fear of police. In Techcrunch They echo the video and put on the table the possibility that this type of behavior is due to learning that Tesla robotaxis have made of their own drivers. It is likely that, learning with the Huge database Of millions of cars on the road, the artificial intelligence that supports the movements of the autonomous car replicates the behavior of a good part of the humans: stop before a police car. This is what DGT defines how “Miron effect”. “Live millions of lives”. For a long time, Elon Musk presumes that the artificial intelligence used in its drinks from the vehicles that the company has in the streets. “It’s like living millions of lives simultaneously And see very unusual situations that a person in his entire life would not see, “Elon Musk said in the presentation of Tesla Cybercab. The company trusts not having to invest billions of dollars in the development of its total autonomous driving such as Waymo and Cruise because they use the data that their drivers yield with their vehicles. Thus, artificial intelligence learns from their behaviors and makes decisions that could be considered more human and less robotized. As human as stopping a police car or cover an intersection. A differential value. Tesla is confident that this approach is key to positioning itself as the best operator in the market. Although there is only a dozen robotaxis circulating through the streets, the data collected are not only obtained with trial vehicles, real data with real conductors are obtained, which should save enormous amounts of money and hours invested. Not only that, the company also aspires to have cheaper vehicles on the street. The combination of Artificial intelligence cameras It is a combo that, according to the company, allows them to save money on radars and lidar sensors, which make vehicles that are put on the streets a lot. Of course, it is not entirely clear that this is the right act because, as has been seen in some tests, The lidar sensor is much more effective in front of the cameras in concrete situations such as low visibility or with obstacles that can generate confusion to the cameras. A shadow. The latter is relevant. Mark Rober’s test showed that a car with cameras cannot discern safely If what is in front is a painted wall like a road or a real road. The lidar sensor, however, does detect the obstacle. Although this situation is not going to be given in real conditions, it does show us the effectiveness of the system to discriminate when it is facing a situation or not of danger. For example, when a Tesla Robotaxi confuses a simple shadow with an object On the road. This case, which is in A Reddit post Where all the errors of the robotaxis are being collected, it could be solved with the use of this system instead of trusting everything to the cameras. Photo | Tesla and Remi Gieing In Xataka | They collide with “clearly visible objects”: Tesla wanted to save to the maximum in autonomous driving and already know the consequences

I have no idea how an app is programmed. I just created one in 20 seconds with Google Firebase

by

I have tried Learn to program a few times. Each and every one of them I have finished leaving it. Beyond the most basic foundations, I have no idea how to start building an app. My partner Iván Linares, from Xataka Android, has managed to create an application with the new tool from Google, Firebase. And, according to your experience, it is something that is achieved in five minutes. So I wanted to test This tool To check if, the King of Dummies in programming, I could create my own applications in a matter of minutes. It would be an achievement. Google Firebase. Google Firebase It is a cloud platform for the development of applications. It is not a new service, but now it has a new ally that allows you to even create an app: Gemini. Thanks to the implementation of its artificial intelligence, one of the most powerful and covered of the moment, it is possible to create applications in Google Firebase. Best of all, we do not need to have a code idea. We just have to go to Firebase Studio. Point that the process is completely free, and that we have up to three work spaces available. More than enough to to play beautifully with the tool. Gemini’s integration is still in the test phase, so we are warned that there may be errors. Testing Google Firebase: Open Firebase is to open an interface that leaves no margin of error. We have a box to write the prompt, in which some ideas already preferred appear. I have not wanted to complicate or the least: “A CALCULATOR APP”. Gemini has not taken five seconds to generate the keys of “Calcify”. It does suggest the name, and you can change it if you don’t like it. In the summary that has offered me, the following data appeared, information about the app that we are going to create. The UI (the interface). Mathematical Operations Algorithm Dynamic screen to show the results Errors calculation Color palette Wow. In case I did not convince me some part of this return on how to make the app, you only have to describe the changes. Now just play “Prototype This App” and expect magic to happen. It is impressive to see how, in real time, everything is created. I have been especially attracted to checking how errors automatically self -correction, something that takes more than a few seconds. It has not taken a minute, in just 20 seconds the app. Yes, it is nothing more than a calculator app. But it has created it an AI with a prompt of a line in less than 20 seconds. It is quite impressive. So far we have our app, but I don’t hide if I confess that I have released a “jo*er” just after. I wanted to try how long it takes to make changes. With a simple prompt. I want the color of the “ac, c, and +/-” keys to be network, so there’s more distinction. I also want it to have the name “Calcify” Written in the Top Left Corner. He has got to work, changing the necessary aspects to make the modification that I have told him and … It has not taken five seconds to return the app with the relevant modifications. Firefly’s main limitation is in the publication, although it helps us a lot. We cannot directly ask you to generate the APK, and we will have to help ourselves more or technical knowledge to import the necessary files in Android Studio and end up generating the app file to use it outside Firebase. However, it is an impressive tool as much as entertainment and to take our first programming steps. Combined with Chatgpt or Deepseek it has enormous potential. Image | Xataka In Xataka | Jensen Huang is clear: At this point no one should learn to program, AI will do it for us

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.