I will tell you a real anecdote.A close relative submitted their final master's thesis a month ago. The center ran the thesis through Turnitin, a program that calculates the probability that a thesis was generated by AI. If it's very high, they reject it and require it to be redone. It came out at 8%.The thing is that when we received the correction from the tutor, it seemed a bit strange to us. It was too impeccable and with somewhat baroque expressions. Strangely precise points to correct. How strange, we said to ourselves! We uploaded it to a detector. Result: 80% probability of being AI. I promise it's true.It's no reproach! I think what the tutor did is perfect, as long as he had read the work and polished the AI's correction. But the anecdote shows us that it is absurd to try to expel artificial intelligence from the educational system.It happened with the calculator, and with the internet. And with Google! Do you remember when it was forbidden to use search engines for schoolwork? Little by little we saw that the issue was not spending three hours looking for a piece of data that could be found in thirty seconds. The important thing, from an educational point of view, was to know what to do with that data.Introducing AI into the educational system is quite a challenge. We must differentiate between using a tool and using thought. A student can use AI and force themselves to order ideas, compare approaches, correct an argument, imagine examples, or improve a complex theory. To think!The problem is the student who doesn't think, only copies or pretends a competence and delivers an empty text.So before he will have to learn to write, of course. You cannot put a child in front of an AI before they learn to write well, to read, to order ideas, to sustain arguments... Just as you don't teach to use a calculator before knowing the times tables, we will have to introduce AI after the student has achieved these skills.Once this threshold is overcome, introducing AI will allow us to develop in our young people the skills that are already most in demand in the labor market: judging, refining, correcting, reinterpreting, contrasting, personalizing…Therefore, this is not about whether the student can use AI. The question is at what age they start using it, how they should use it, and to develop which skills. We will have to design exercises in which AI raises new skills: asking for processes, demanding intermediate versions, introducing oral debate, having a student explain why they have accepted a suggestion and why they have rejected another.This is the future. And it will require great masters... in person!

This week Anthropic has presented an AI that it has decided not to sell because it considers it too dangerous; Meta has launched the first closed model in its history and has abandoned the commitment to open source that it had preached as a moral obligation, and OpenAI is preparing GPT-6 while a journalistic report uncovers very questionable practices by the director. All three together paint the portrait of an industry in a race without brakes, incapable of self-governance or not interested in doing so.

1. “It’s a 3D view accompanied by personal experience, which gives us a much better interpretation than many images obtained by robotic probes”. I found it gratifying that a NASA administrator, Lori Glaze, recognized the human gaze, at a time when AI seems poised to overwhelm natural intelligence, risking domination over collective experience, ridiculing our condition. Victor Glover, one of the four astronauts who orbited the Moon, said during the eclipse: “It’s the strangest and most surreal sight we’ve had today, with the Earth’s glow illuminating the Moon”. And the spacecraft continues its adventure, which no one can explain better than the four people who have lived it. Without them, there would have been events but not an experience. That is, a lived reality transmitted by beings of flesh, bone, blood, and natural intelligence, who will be able to explain it to us (with the limitations imposed by the relevant authorities, it must be said, which is also a human condition) with the intensity of the lived experience. That is, a perception, from reason, accompanied by sensations, feelings, strengths, and weaknesses, that shape our species and that AI, capable of generating infinite accumulation and combination of data, will never be able to transmit with the singularity and sensitivity that, for better or worse, constitutes the human condition.

In 2019, a Chinese developer created a GitHub repository, the most important social network for programmers worldwide, called “996.ICU”. The codename denounced the exploitative labor conditions of Chinese tech companies: if you work 996, from 9 AM to 9 PM, 6 days a week, you’ll end up in the ICU. The repository went viral, with more and more Chinese programmers supporting it and creating blacklists of companies that applied the 996, such as Alibaba, Huawei, or Tencent. Chinese tech companies and the government were alerted, but the wave was difficult to stop: GitHub, a large open-source library, is the only major Western social network that is not censored in China. The 996 became a major national debate. Two years later, this spontaneous labor movement managed to get the Chinese government and courts to declare these working conditions illegal.In recent years, there has been much talk about open Chinese AI models. The most famous was Deepseek, but Qwen, Manus, or Kimi are also well-known in the programming community. Unlike the closed and private AI of American tech companies, Chinese companies have opted for open, free, and highly modifiable models. This has made them popular among startups or developing countries. Right now, open Chinese AI models are the only ones that can rival American ones. They have become one of China's most cutting-edge technological sectors, and a great source of national pride.China's technological success is often explained through state subsidies or long-term plans of the Communist Party. But, as analyst Kevin Xu explains, the Chinese AI models that are now so successful have not emerged from government initiatives, but from a spontaneous movement from below: the social mass of Chinese open-source programmers who launched the campaign against 996 is the same one that has created the AI models that fascinate the world.

Escalation is usually understood as a simple increase in violence, but it is actually a more complex phenomenon: a crisis escalates when it intensifies, expands to other geographical areas or domains (land, naval, aerospace, or cyber), or affects increasingly sensitive targets, thus becoming more difficult to control. Today, this risk is greater because war increasingly depends on a set of technologies that change its logic: the use of precision and long-range weapons, the proliferation of drones –and, above all, the possibility of using them in swarms–, artificial intelligence, cyberspace, electronic warfare, satellites, and, in some cases, hypersonic weapons. But also because many of these capabilities are more accessible and spread beyond major powers. Not all have the same weight, but they share a common underlying effect: they accelerate events, increase ambiguity, and make it more likely that a limited action will be perceived as a more serious step than perhaps intended. The first problem is speed. Precision capabilities –the combination of guided weapons, sensors, command and control networks, and navigation tools that allow for the precise location, tracking, and engagement of targets– can produce profound effects in a very short time. The novelty is that the time between observation, decision, and action has been radically shortened, between the sensors that detect, the command centers that process information, and the vectors that execute the attack. Technology compresses decision cycles and favors premature reactions or those based on an incomplete reading of the situation. A well-coordinated attack can not only destroy physical targets: it can degrade command and control, disrupt communications, degrade radar coverage, or hinder target detection and tracking. To all this must be added AI, which accelerates data fusion, target prioritization, and the pace of decision-making, as well as hypersonic weapons, which, due to their speed, maneuverability, and shorter warning times, can further reduce the margin for interpreting signals and calibrating responses. The strategic problem is evident: as time shortens, the risk of reacting not so much to what is known as to what is feared increases.

Beth Noveck (New Jersey, USA, 1971) makes it clear in each of her answers that she refuses to fall into defeatism. She offers solutions, gives examples of success, and argues that AI can help improve the functioning of institutions. And she knows what she's talking about, not only because she is an expert in how to use technology to make governments more open, but because she knows institutions from the inside: she was the first Chief Technology Officer of the United States during the administration of Barack Obama. She visits Barcelona to sponsor the presentation of the Pompeu Fabra University's 2026-2037 Strategy.

Over the last decade, the technology sector has experienced a boom in Software as a Service (SaaS), which became one of the leaders of digital transformation in the business world. This cloud-based software, with user subscription programs and employee interfaces, largely defined how SMEs should approach their digital transformation. Traditional SaaS marked a turning point, especially for business cost management, particularly in the area of ​​human resources. However, the emergence of artificial intelligence (AI), and especially AI-powered autonomous agents, has changed the rules of the game and led to a situation that the ecosystem has dubbed as SaaSpocalypseOr the SaaS apocalypse. It's not so much the end of SaaS as a reinvention of the sector, which is realizing that it won't thrive unless it shifts towards an agentive version—that is, integrating agentive AI. The model now requires autonomous agents capable of executing end-to-end workflows with human oversight, but with a degree of autonomy. An example of this shift is the Catalan company Nodus, an operational architecture that allows for the structural deployment, governance, and scaling of digital workers within a company, preventing the fragmentation that arises when AI is implemented in a scattered way. Founded in Catalonia, the company began developing its technology two years ago and will be working with clients by 2025, aiming to integrate AI into the operations of SMEs to help them grow in an integrated manner.

Factorial, the platform dedicated to AI-powered business operations, announced on Tuesday the launch of the AI ​​Acceleration Fund, a €10 million fund to boost the adoption of these technologies and the digitalization of business operations in European companies. According to CEO Jordi Romero, "the goal is to lower the barrier to entry for AI adoption for SMEs" and facilitate companies' transition from pilot programs to real-world deployments of automation and data usage in areas such as human resources, finance, technology, and operations throughout the entire employee lifecycle: performance, workforce analytics, and planning. The program, operational from Tuesday and with applications open until April 30, has a budget of €10 million structured in two equivalent components. The first tranche, worth €5 million, is allocated as direct cost reductions applied immediately upon signing the contract, without impacting the commitment to the platform. The second tranche, also worth €5 million, is offered as AI credits within the Factorial platform itself, so that companies have additional resources as their transformation program matures. Furthermore, participating companies will have access to a transformation support program that includes training and development webinars, access to events, and thought leadership reports.

It has been ten years since a phenomenon that revolutionized the streets and phones of thousands of people around the world. The Pokémon Go app, an augmented reality game in which players have to go outside and search for creatures with their mobile camera, became a huge success. But the mobile cameras of users have served the game's creator, the company Niantic, to achieve a more ambitious goal. It has used the photos, without players knowing, to create a three-dimensional map of the world that autonomous robots will now use to move through the streets, especially when delivering packages and orders.

For years, sociologist Elisa García-Mingo has immersed herself in the study of the machosphere, online communities that spread misogynistic content, whether through ridiculing comments and messages or by asking artificial intelligence to reinvent photographs of women to undress them.

Did you know that algorithms decide what ads we see on social media? It might seem insignificant, but it's not. "If you're a young person about to decide on a university degree, it's very likely that if you're a guy, social media will show you degrees in engineering and computer science, and if you're a girl, degrees in education, nursing, and caregiving," says Liliana Arroyo Moliner, PhD in sociology and director of the Chair for Socially Responsible Digital Innovation. Milagros Sainz, a researcher and professor at the Open University of Catalonia (UOC), explains that the popular Google Maps app "uses a man's pace to calculate walking distances, which often makes women or people with mobility issues take longer." There's also discrimination when looking for work, in healthcare, and with the algorithms used by banks to decide whether or not to grant a loan, because the mathematical models of these algorithms are applied in very different fields, and many decisions are made taking them into account.

It seems increasingly urgent to provide our students with tools that allow them to differentiate rigorous information from falsehood and to foster critical thinking. The problem is that we do not have infallible tools that save us the effort of thinking. This is why I want to talk about the tree pop of the Pacific Northwest (Pacific Northwest tree octopus).

The image of Chinese robots dancing like professionals has gone viral. It makes us wonder if machines will replace us in many tasks that humans currently perform. Artificial intelligence (AI) should boost productivity growth by automating the production of goods and services and, more surprisingly, the production of ideas and discoveries. The paradigmatic example is AlphaFold2, the AI ​​model that "solved" the problem of determining the 3D structure of more than 200 million proteins from their amino acid sequence. This has transformed structural biology, reducing analysis time from months or years to minutes, and led to the awarding of the 2024 Nobel Prize in Chemistry to Demis Hassabis and John Jumper of DeepMind. In addition to facilitating imitation and learning, AI can improve itself in a feedback loop. It has evolved from "predicting the next word" to ever-increasing sophistication.

I'm aware of how foolish it is to want to talk about artificial intelligence right during the week of Mobile World Congress, the year in which AI is the main focus of attention. And even more so if the point is to say that one of the best things it offers us right now is that it not only forces us to better define its scope and limits, but above all to consider what differentiates it from human intelligence. Or, to put it without so much self-promotion, what the true specificity of the human condition is. A few days ago, in SubtrackIgnasi Llorente argued, quite rightly, that AI needs more philosophers and fewer computer scientists. It's a good way to put it. Llorente wrote this in response to the news that Google is looking for philosophers to join its AI teams, and that it doesn't want them just to make things look good.

Whenever an invention capable of simplifying more or less costly tasks emerges, a significant portion of society's imaginative workforce is concentrated on developing this invention and finding a social place for it. In other words, many other issues are neglected in favor of focusing on the one that seems to promise the most dazzling future. There is, therefore, an expansion of imagination on the one hand, and considerable pressure on the other (some disciplines are filled with disciples; others are emptied). However, this shift in imaginative focus occurs amidst a kind of revolutionary explosion in which it becomes difficult to assess whether we will find "progress" in the direction indicated by the new invention.

A study by Pompeu Fabra University (UPF) shows that the most popular generative artificial intelligence (AI) tools spread errors in Catalan because they are trained in Spanish, in addition to English: non-standard forms in Catalan grammar and lexicon are spread due to the influence of these global languages. The study, led by Professor Thomas Brochhagen from the UPF's Department of Translation and Language Sciences and published in the journal "Linguamática", is pioneering in demonstrating the bias towards Spanish. According to researcher Mireia Almena (UPF), "AIs not only reproduce language, but also influence its evolution and can have a much greater impact on languages like Catalan, with less written content in digital media, than on other languages with more speakers and text production capacity such as English, Spanish or Chinese." Therefore, they call on institutions to work to improve these biases. In fact, the Accent Obert Foundation has already announced that they will "take the official exams of Catalan students for the most popular AIs" to evaluate their knowledge of Catalan language and culture, in order to objectify the shortcomings in this area. Spreading Errors

I was impressed, and not only me, by the spectacular Chinese exhibition last Monday of humanoid robots dancing with a skill we used to associate with expert human dancers. This prompted three comments.

Every day, headlines and sweeping statements about artificial intelligence and work circulate, fueling collective anxiety: that AI will destroy millions of jobs, that it will replace people, that it will end up controlling us, or that it is a threat to society. These are expressions of a diffuse, often irrational fear, characteristic of what psychology identifies as confusional anxieties: when the risk is not well defined, the fear is magnified. The only way to reduce this anxiety is to confront it with data and sound judgment, distinguishing what is true and what is not regarding the real impact of AI on the world of work.

The booing of JD Vance at the opening ceremony of the Winter Olympics in Milan has become the audible representation of a growing sentiment within the European Union. The arrogance of Trumpism has awakened a imperceptible, wounded European pride that is beginning to surface. The "rupture" that the Canadian Prime Minister Mark CarneyThe debate that began in Davos has also started on this side of the Atlantic. It is still embryonic, more declarative than real. But the EU has already defined the playing field for the confrontation with the United States that it is prepared to fight: the digital space.

It was only a matter of time before the Spanish government announced restrictions on social media. This was foreshadowed by the interest in examples like Australia and, soon, France. The first spark of the debate on age restrictions ignited in the summer of 2024, with the controversy surrounding the consumption of pornographic content at very young ages. At that time, the government announced the Beta Digital Wallet, a digital passport that includes age verification—commonly known as pallaporte–. The restriction is also mentioned in the draft bill for the protection of minors in digital environments, which is currently being processed.