tool

The Captcha had become an excellent tool to fight the bots. Until Chatgpt Agent arrived

by

In 2003 a young Guatemalan named Luis von Ahn published a unique study along with two colleagues from the Carnegie Mellon University and an IBM researcher. That project described an automated test that was easy to solve for humans but practically insurmountable for artificial intelligence systems. Those researchers called that test Captcha. The concept was simple and focused on the already famous Moravec paradox: there are things that humans do effortlessly – such as solving the visual puzzles that the captcha propose – but that the machines fail to solve. The idea turned out to be one of those between one million. Von ahn He ended up creating an improved version to which He called recaptcha that not only verified that you were human: I did it helping Train and perfect OCR systems. That other complementary idea was another unique moment “Eureka!” De von Ahn, and in fact he ended up making him a millionaire in 2009, the year Google decided to buy his service. Then he would dedicate himself to another equally striking project (or perhaps more): Duolingo. A dizzying (and juicy) evolution While doing so, the captcha continued to grow and evolve. Putting it more and more difficult for machines that gradually stated that perhaps those tests were no longer so valid. From those basic captchas we ended up moving to recaptchas of all kinds in which visual puzzles not only challenged the abstraction capacity of machines, but also helped to train no longer OCR systems, but artificial vision systems to better recognize cars, buses, zebra crossings or, how not, fire mouths. But artificial vision and intelligence systems also improved, and that struggle between these tests (Captcha comes from Completely Automated Turing Tests to Tell Computers and Humans Apart) and the machines became more and more interesting. It was a Singular cat and mouse game with spambotsand when some AI system managed to overcome a captcha or any of its variants, the puzzles became more and more difficult. The story has repeated again. He did this Friday. It was then that a User of the R/OpenAI community in Reddit published captures of Chatgpt agent overcoming without apparent problems one of the recaptchas more popular and used today on the Internet. This is the system TURONSTIL from Cloudflare, which presents a small box with the text “I’M not a robot“(” I am not a robot “) to click on it. It seems very simple and simple, but it is not so much for the machines. As indicated in Cloudflare, this recaptcha variant analyzes various signals such as the mouse movement, the time we take to click, the “digital footprint” of our browser, the “reputation” of our IP or some Javascript execution patterns. With them determines whether the user is a human being or is suspicious of being a bot. And if there is suspicion, the system presents after that first captcha another in which we do have to solve some type of visual puzzle. AI does not know if it is human, it only tries to operate as such The funny thing here is that Openai’s agent solved the problem in an obvious way: seeing what was on the screen to act accordingly, something that had not been easy until now. The agent was even narrating what he was doing, and while doing that step he showed the following text: “The link is inserted, so I will click on the” verify that you are human “click to complete the verification in Cloudflare. This step is necessary to demonstrate that I am not a bot and be able to continue with the action.” Or what is the same: The machine was self -healing like a human being. It is something unusual, but perhaps not quite strange considering that 1) AI does not really know what it says and 2) has been trained to speak (and act, at least in a limited way) as a human being. Operator, Openai’s previous agriculture, passed it really bad With these systems. Does this mean that captcha are threatened with death? Probably not. This is nothing more than another battle of that war between the bots and the captcha. One that for example we saw Another AI victory In October 2024 but it did not involve the debacle of this type of user verification tests. As they point out In Ars Technicathe captcha systems have not stopped evolving. From those blurred and deformed texts we go to recaptcha in which we had to solve visual puzzles of all kinds that lately force us not to identify traffic lights, but to place an image in a specific orientation – a increasingly popular system called Arkose MatchKey– Oa having to identify some element of an image that does not agree on it. In fact, the most recent captchas are no longer so much to prevent bots from exceeding that barrier to slow them so much that making brute force attacks with bots does not compensate. Captchas not as a barrier, but as a bots brake An article of those responsible for Arkose Labs, creators of Matchkey, made it clear that “There is no completely impenetrable captcha“, and that with their proposal what they intended was” to introduce an economic deterrence or cost proof for the malicious behavior of the bots. “Or what is the same: than to develop a bot that exceeds those captcha was so expensive that It was not worth it. Thus, we should not worry much in case the agents of AI can overcome this test, because surely they will appear captchas that continue to assume an almost impassable barrier for these systems. It is precisely the same concept with which it works THE ARC-AGI 2 TESTwhich measures visual understanding and abstract reasoning of AI systems and that is so complicated that the best AI models, which are also very expensive, do not exceed 4% of cases in the best case (O3-Preview). Will there come a time when those agents of ia get … Read more

Many times art is inexplicable, and a retiree from Malaga uses the most illogical tool to create it: Excel

by

Explain what art is as complicated as sometimes get the weather forecast. Few things are more subjective and, while for many a banana is a simple fruit, for others it is a work that costs 6.2 million dollars. But within that subjectivity, reflecting reality in a painting is one of the most “accepted” ways to create art. What if that painting is done not with Photoshopbut in Microsoft Excel? That is precisely, with what Jesús Villanueva enjoys. No tutorials. “Normal,” you may think, since who would have occurred. Well let me tell you … it’s not so weird. We are going to see it throughout this pictorial trip with such a boring program (and useful) as Microsoft Excel. The program, which has almost 40 years behind himit was launched as a system of exclusive spreadsheets for Mac In 1985 and its bases have not changed: it is still a calculation program. However, each version became more powerful and versatile. In it Microsoft blog We found the story of Jesús Villanueva, 78, who entertains himself not playing ‘Skyrim‘, doing Streaming of ‘Leage of Legends‘ either Catching Pokémon: It entertains with Excel. He studied a construction eyeliner course and is what has been dedicated, but after retirement and pandemic of the COVID-19something changed Miguel Ángel del Excel. With free time (because you have to have it to create such works), Jesus began using Excel as canvas. “I never found a tutorial to explain how to do it. It has been trial after proof, drawing after drawing,” he says. He began recreating buildings in a format without perspective, but as he learned to master the tool, he went to more complex compositions. Image | Jesús Villanueva Image | Jesús Villanueva The most impressive thing is that it is not dedicated to borrowing external resources or hitting images, but to draw with colors directly in the grid, combining complex and simple shapes to give life to your idea. It is based on photographs, yes, but draw all forms manually in the application. Your secret? “Everything is a matter of patience, even in works that can take up to three months, such as the Siena cathedral.” Other religious buildings in its porpholio are the Mosque of Córdoba or the Cathedral of Florence, for which he had to draw the little characters large and then reduce their size to be able to embed them on the general image. The “Insert Forms” brush. That is the basic tool for Jesus. Choose a figure, calculate the proportions to case in a reference image and there begins to adapt, color and combine with others to give life to the object you want. When he has it, he unites it to the rest and the scene is formed. He also plays with vector strokes and the change of scale of elements. Image | Jesús Villanueva Japanese landscaping. In it Instagram of Jesus You can see his work, but he is not the only passionate of this particular brush. Tatsuo Horiuchi, an 84 -year -old Japanese man, also decided to unleash his particular vision of art when he retired. He says he took 10 years to create something that he felt that he could proudly teach people, but also that there were people who did not understand their hobby or that they laughed at it. “Why do you put so much effort on something that is not useful? Are you crazy?” “Well yes,” he replies in a video in which we can see part of his collection. Excel has a lot of life. Horiuchi gives a very important key. “I think, even if you don’t have talent for painting, you can paint something whenever you have Excel.” And this is interesting because there are people – as I – who are denied to paint even if he would like, and with a grid, it may be easier to start drawing. In any way, it is always curious to see that Excel is not just the tool in which Sustains the world economy and even formula 1 equipment: Excel too It can be a sport. And to gamify a tool as boring as it has all my respects. In Xataka | Thousands of people have been living from art for three years. Ireland pays them 1,300 euros per month to fulfill their dream

To carry out its greatest attack on Russian soil Ukraine used a tool as old as wars: bribery

by

On June 1, what seemed several mobile houses located at various strategic points of Russia were revealed as what They were really: camouflaged trucks with an army of drones prepared for covert operation. The so -called Spiderweb began, the greatest Ukrainian offensive on Russian soil, a devastating attack on the fleet of Moscow bombers. Now, almost a month after the attack, Ukraine has responded to one of the great unknowns: how the hell swarms entered into Russian territory. Redefining modern war. For more than eighteen months, Ukrainian intelligence services They designed meticulously an operation that culminated in an unprecedented attack on thousands of kilometers of its borders: An army of camouflaged drones in those prefabricated mobile houses, which were transported across international borders until they reach Russian strategic air bases. Now, Vasyl Maliuk, head of the Ukraine Security Service (SBU), has revealed the details of the operation in an extensive Interview with Washington Post where he assures that the attack destroyed at least 12 Russian aircraft, including Tu-95 bombersand damaged a total of 41 devices, some of them located More than 3,400 kilometers east of Ukraine, in Siberia. The satellite images analyzed by the post confirmed Partial destructionalthough some visual tests were limited by weather conditions, which suggests that part of the remains could have been removed by Moscow before being captured. How it was created. Maliuk explained That the design of the operation began in November 2023, in response to a new wave of Russian bombings on Ukrainian cities. The idea was to hide drones inside those houses Mobiles that simulate being standard housing units. These structures had to have autonomous energy systems (Solar panels and batteries special) to keep the vehicles unmanned loaded, even in extreme winter temperatures, waiting for the final order. Russian restrictions on the importation of technological components made a tactic as old as the wars themselves: The briberyin this case to customs agents to introduce the materials. No one knows (almost) nothing. Secretism, apparently, was total: each group worked with compartmentalized informationthe engineers who manufactured the drones ignored their real purpose, as well as those who assembled the prefabricated houses they did not know that they housed weapons. In Maliuk’s wordsit was “a tactical symphony” with multiple critical but autonomous parts. Surgical precision. Thus, the morning of the attack, 117 modified FPV drones with double explosive load They were activated from the mobile roofs of the camouflaged houses. Each artifact was directed by A selected operator Among the best in the country, who was assigned A specific objective: A specific plane, whose location and environment were studied with detailed models. Loads They were designed To penetrate the fuselage first and then detonate inside, maximizing damage to key points such as fuel deposits, missile launchers and electronic systems that Russia cannot easily replace. Although Maliuk avoided detailing the communication systems used to The remote guidanceconfirmed that they were multiple and sophisticatedspecially designed to avoid interference and guarantee success. Strategic consequences. The Spiderweb Operation It was just the first blow. Two days later, Ukraine He executed another offensive of wingspan against the Kerch bridge, symbol of the Russian occupation of Crimea, using submarine explosives of 1,000 kilos. Although traffic was restored, the coup had a clear symbolic and strategic effect. Russia replied With a massive wave of drones and missiles that reached civil areas of kyiv, leaving dozens of victims. Far from causing a decalized, the Ukrainian action confirmed A new level In the technological war and demonstrated kyiv’s ability to achieve very much within the enemy territory. According to MaliukSpiderweb is just a sample of the potential of Ukrainian intelligence to alter the rules of the conflict, and represents a logistics, technical and human effort “unified in a single structure.” The future of wars. It We have slipped On other occasions. In full Russian offensive, and given the constant pressure on Ukrainian cities, the Spiderweb operation Mark a milestone In the modern war. Not only because of the depth of scope (from bases in Ukraine to Russian facilities in Siberia), but also for The operational model: Fragmentation of knowledge between operators, bribes, use of civil camouflage, innovation in light armament and ability to launch high precision attacks without resorting to large conventional platforms. In words of Maliuk himselfthis is just “the tip of the iceberg” regarding the work of your agency with drones, networks of agents and undercover operations. In a conflict that is increasingly freed with bits and algorithms than with armored divisions, the offensive showed that the 21st century war is defined by who dominates the stealth, well above the brute force. Image | UKRAINE’S 93RD MEChanized Brigade In Xataka | We suspected that Ukraine drones attack had been destructive. Space images have revealed how much In Xataka | In 2024, Ukrainian trucks disguised as “house” entered Russia. Now they have dynamited their main air bases

Scientists have a new very powerful tool to fight Alzheimer’s and Parkinson: quantum computers

by

To solve most of the problems that scientists expect to be able to address in the future with Quantum computers capable of amending their own mistakes, such as optimization, those in the field of cryptography or artificial intelligenceit will be necessary have several million cubits. It can even, that hundreds of millions of cubits. The most advanced quantum processor currently IBM has itand it has just over a thousand cubits, so it is evident that many technological challenges remain that it is necessary to solve. The interesting thing is that there is no single way to go through this path. Organizations that are investigating in the field of quantum computing work in several different cubits technologiesand each of them is in a different degree of development. IBM, Intel and Google are some of the big companies that have opted for superconductor cubits, but also do so smaller ones, such as Atlantic Quantum, IQM, Anyon Systems, Rigetti Computing or Bleximo. In fact, if we stick to the number of companies that are working on this type of quantum bits it is reasonable to conclude that this is the technology that has greater support and greater investment, so, in some way, it is the one that goes in the lead. This strategy is probably what will help us to have more cubits, but it is also more prone to make mistakes than ion traps cubits, which are one of the alternatives to superconductors. In addition, these last cubits are characterized by working at a temperature of about 20 millikelvin, which are approximately -273 degrees Celsius, with the purpose of operating with the greater degree of isolation of the possible environment. Ion traps are given great proteins Ion traps are currently the main alternative to superconductor cubits. This is the technology in which they are working, among other companies, Ionq and Honeywell, and is characterized by using ionized atoms, and, therefore, with a non -neutral global electric charge. This property allows them to be isolated and confined inside an electromagnetic field, although this is only the starting point. From here Ionq acts on the quantum state of its cubits with ion traps cooling them to reduce the level of computational noise and uses lasers just then to operate with them. However, it does not use a single laser; Use one for each ionand also a global laser that acts on all of them simultaneously. Honeywell also uses ionized and laser atoms, but the procedure used to establish the intertwining between two ions and act on them with a laser is different from that used by Ionq. For scientists it is crucial to understand the folding of proteins that triggers Alzheimer’s or Parkinson Parkinson Precisely a team of researchers from this last company and the German emerging company specialized in quantum quantum computing To do something amazing: Solve protein folding problems with up to 12 amino acids. To carry it out they designed a quantum optimization method that seeks to find the optimal configuration of protein folding. Expressed in this way it seems complex, and it is, but the really important thing, and with what is worth it, it is that these quantum computers thanks to the appropriate algorithm are able to help scientists Understand the protein folding mechanism that triggers diseases such as Alzheimer’s or the Parkinson. And understand this phenomenon well is the first step towards the elaboration of effective treatment. This result is very promising, but there is still a lot of work to do so that quantum computers help us in front of these diseases. On the one hand, folding models must evolve to be more reliable and realistic. And, in addition, the classic algorithm that is responsible for refining the results delivered by the quantum algorithm should be more precise. Even so, the work of these researchers is an exceptionally promising starting point. Image | Ionq More information | Arxiv In Xataka | Quantum computers threaten encryption technologies. This is the reason why we do not have to panic

The Eurovision voting system is broken. And that has led to becoming a propaganda tool

by

Eurovision has attracted the eyes of the entire continent this weekend (the audiences in RTVE were stratospheric), although this year a series of lines that place the contest in an awkward position have been crossed. The participation of Israel, controversial from the first moment, and the overturning that the scores gave at the last moment have put on the table the idea of ​​the Use of the system for political propaganda purposes. And the fault is, in part, of the way of scoring. What happened. That a series of events took place that, separately, could not attract attention, but that added to an unusual edition, with cross accusations and, more than ever, a festival in which music has been relegated to an absolute background. Israel starts lazy → Israel received few points from the jurors of the countries. Only Azerbaijan gave him 12 points, which left him in the tail of the scores, with just 60. Robo in the scores → but with the points of the televoto there was a dramatic turn: he received 297 points of the public, with 12 points in 13 countries, including Spain. Many others gave him the 10 points of the second position. Israel led the classification until the last moment, when Austria received the points of the Televoto and surpassed it for just a dozen advantage. How to vote. The voting system goes beyond a mere “first the jury, then the public”, and carries with him A account of a certain complexity: The European Radiodifusion Union computes the votes of its spectators, collected by call, text message or internet. Each person can vote up to 20 times, but not by their own country. The spectators of non -participating countries (“Rest of the World”) group their votes and count as if they were one more country (which also allows to vote to countries like Israel). Next, a list of the most voted options in each country is prepared and the same points as the jury: 12, 10, 7, 6, 5, 4, 3, 2 and 1 points, regardless of the number of people you have voted are awarded to the first 10. That is, it is not about how many total votes are those who remain above but simply that they are The most voted In each country. These points of the public are worth half of the total, which enables radical modifications in the jury’s decision, as has happened this year. Made the law, made the “trap”. That system explains that it is relatively simple to mobilize certain political positions to vote in one direction or another. David Saranga, acting director of the Public Diplomacy of the Ministry of Foreign Affairs of Israel, admitted to the Ynet portal, as stressed See youthat his body intervened in previous years to promote the vote to Israel, achieving 12 points by the public. Influential profiles on social networksmatches of right and means opposed to government They served as a speaker for a positioning related to Israel that resulted in votes. Not so much televoto. The voting system It has evolved With the passage of time. Until 1997, the winner decided on national jurors, without direct participation of the public. Between 1997 and 2008, the televoto by telephone and SMS was the main system and the jury was only used as support. Since 2009, the current mixed system was established: 50% of the result comes from the televoto and 50% the professional jury. The intention was to avoid precisely situations such as those that have been seen in this edition, although they have led to mass votes in the opposite direction, as happened with the Victory of Ukraine in 2023. Spain protests. This propaganda work that, despite the statements of Saranga, has been camouflaged as spontaneous movements by high Israeli charges as Israel Minister of the Diaspora, Amichai Chikliit fits perfectly to Eurovision rules. That has not prevented RTVE from asked the European broadcasting union open a debate about “the appropriate” of the Televoto and if it is being manipulated by political interests. In addition, as announced by the public channel itself in its news, it will request an audit that specify and break down the voting of each country. TVE against Eurovision. This audit, of being held, will be the last confrontation of the public channel and the European broadcasting Union. Before that, last Thursday, in the presentation of the second semifinal, The presenters Julia Varela and Tony Aguilar said that “this year RTVE has asked Eurovision for a debate on Israel’s participation in the festival. The victims of Israeli attacks in Gaza already exceed 50,000, and among them, more than 15,000 boys and girls, according to the United Nations Organization.” Those words earned a warning to RTVE, who responded with a poster before issuing the contest positioning in favor of Palestine. Melody’s failure. It has been suggested that the low classification of Spanish representation could have to do with this conflict between RTVE and Eurovision. And that his position 24 is not something that we are precisely unusual in Spain: in the last twenty years, We have remained in positions below 15 times 16 times. This year the perspectives were better: the forecasts put it among the first 15. Not having an apparent explanation for the debacle, conspiracy theories were launched blaming these bad results to the pro-palestine position of the government and RTVE. It was then when He put on the table The idea of ​​manipulation in voting. Header | RTVE In Xataka | Film rooms are becoming “show rooms” of all kinds. The best example: Eurovision

Overcoming our brain when making the purchase is not easy. Some researchers have developed a tool to achieve it

by

“Codazo theory.” It’s how we could translate Anglo -Saxon expression, “Nudge Theory”, Which refers to the study of actions focused on giving us A “little push” Towards a specific decision. The concept is especially used in the context of the behavioral economy, That place where the economy is mixed with psychology. We know that companies often resort to these “pushcins” to lead us to purchases that generate more benefits but can we use this tool to improve health? Digital tool. The answer is yes, and there are many researchers who work to develop this type of tools. The last It has been developed By a team of Duke-Nus Medical School and it is a digital tool that aims to facilitate the task of choosing healthier products in our purchase on-line. An imperfect system. Choosing healthier food products is not always easy. Yes, we can always make sure that our purchase car is full of fruits, vegetables and other basic foods to cook at home, but this option is not always realistic. Labeling systems such as Nutriscore They can help us choose better, but they have Important limitations. The first is that evaluating the healthy of a product in a simple index is not always easy and sometimes leads us to important inconsistencies. The second, that this index must compete with marketing strategies that seek to buy a product, regardless of how healthy or ceases to be. Beyond labeling. To exceed these limitations, Duke-Nus’s team designed a digital tool destined to facilitate a purchase on-line better informed at nutritional level. This tool complemented the information available on the Internet purchase page, adding additional information. This additional information included a traffic light based on the index Nutriscore of products, complemented with other tools. First, the page ordered the products based on this index, first showing the healthiest products instead of doing it alphabetically. Second, the tool incorporated the follow -up of the products in the purchase car, indicating what proportion of food was labeled as green, yellow or red. Third, the digital tool also showed healthy alternatives (with similar prices and characteristics) to selected products. Studying alternatives. The team tested the tool through an experiment. They asked a group of participants to make three purchases on-line over three to six weeks. Some of the participants used this tool, while others used the conventional portal. The team observed that the purchase cars of the people of the intervention group scored better in the nutritional index. They also contained less calories, less fat in total and less saturated fats, less sugar and less salt. The details of the study were published In an article In the magazine American Journal of Preventive Medicine. Public health issue. By will or necessity, there are many people who seek to eat healthier, and that begins in a healthier purchase. Achieving it for oneself is not easy, but it is not to help these people without falling into prohibitions and Tax Methods. Here is like the behavior economy and Nudge Theory They can help us: the same tool often used to take us to more advantageous options for sellers can also be used to improve the health of consumers who wish to improve their diet. In Xataka | The three expert tricks so that they do not lead you with food labeling, according to one of the country’s biggest specialists Image | Lothar Boris Piltz

is to use AI as an oracle instead of as a tool

by

A century ago, being illiterate was not knowing how to read or write. In developed countries That problem today is residual. But another type of illiteracy is emerging. More subtle, more difficult to detect, with more gray and perhaps equally determining: Not knowing how to interact with AI. This new literacy does not know how to program or understand how models work. It is something more basic: know how to ask good questions, know how to read the answers, and above all, know how to distrust. Not in a paranoid way, but with criteria. Distinguish when we are using AI … and when AI is using us. It is the difference between being a passive AI user – some who swallows without chewing – and using it as a lever for thought, as an extension of our analysis capacity. Because, well used, it can be that: a cognitive multiplier. There you play A huge difference: There are those who use these systems as if they were a vitaminated google or a estroid calculator. He throws a question, copy the answer and voila. Other people – even more – are learning to talk with them. To stretch its limits. To generate ideas that neither the machine alone nor them could have produced. The key is not the tool, it is how you use it. And for that it is necessary literacy in AI. The thing goes beyond who does what with Chatgpt. Systems like Deep Research They are beginning to automate tasks that, until recently, were the point of entry to many professions. Reports, summaries, preliminary analysis … just that type of work that served to train, to understand the trade from within. If you give that to a model, how do you learn to think like an expert? That is The black hole that is coming in many companies. If you automate the formative tasks, how are you going to train the new ones? If we do not redesign well – and fast – how the experience is transmitted, we could have entire generations without a real basis. People with titles but without criteria. And not only that: this new illiteracy can be hereditary. Like parents who did not read they did not raise reading children, those who do not know how to use these tools will hardly teach to use them. Learning will remain in the hands of school … or algorithm. The paradoxical is that All this disguises himself very well. Someone can generate a brilliant report, a perfect presentation, a seemingly solid analysis … without deep understanding, much less. It is enough that you know how to ask for it well. The risk is not just that mediocrity is imposed. Is that nobody realizes. As Antonio Ortiz has been notifyingthe real problem is not that AI thinks for us. It is that, little by little, we stop thinking about ourselves and our atrophy begins. That is why the true digital literacy of the future will not be technical. It will be ethical, critical, cognitive. Know when to ask the AI ​​to think about you. And, above all, when to say no. In Xataka | Bill Gates has a favorite book about AI for a reason: predicts better than anyone what will happen to jobs Outstanding image | Xataka

Chatgpt has been a tool. If you start remembering all our conversations, it will be something else: a relationship

by

A few days ago, Openai announced an “memory” update for chatgpt: You will begin to remember all our conversations Already take them into account when giving us an answer. No four inferred pieces as before: All our history. Well, it wasn’t even Openai who announced it. Was Sam Altman in Xas who does not want the thing. Openai focused those days on giving a court to a GPT 4.1 Much less transcendent for the user. And a few days later, Grok announced the same movement. The new super memory (not yet available to the European Union) is a total change in the use we give to Chatgpt. And to a lesser extent, to Grok, who does not have a lower professional utility. We are no longer facing a tool that we use and abandonbut before a digital entity with which we have an evolutionary conversation. A relationship. Let’s think about how we use a hammer, a calculator or even Google: we use them to solve a problem and then forget them until the next time we need them. There is no evolution in our interaction with them. On the other hand, our relationship with another person is contrasted on a cluster of previous interactions. We do not expect to have to remind a friend what team we are, or our partner what music we like. There is something deeply human in wanting to be rememberedto desire continuity in our interactions. The IAS that offer this experience will have an advantage not only technical and functional, but also psychological. Each conversation with chatgpt – at the margin of the utility of the GPTS– It will no longer be an eternal first day, but the continuation of a thread of shared knowledge. An assistant who remembers your allergy to nuts. Which understands that you like explanations with sports analogies. Who knows that you are working on an important personal project. On face B of the album, the concerns that it provokes to deposit so much about ourselves in an entity controlled by a company. Persistent memory offers extreme customization, but it also costs us A privacy toll. Openai says that you can deactivate this function, but the value of the service decreases if you do. It is a usual dilemma for any user of modern services. It will also exist for some The subtle temptation to replace human interactions – Sometimes frustrating – for more predictable interactions with an AI designed to please us. The AI ​​never tired, never has a bad day, never judges our repetitive questions or laughs if we ask something too basic. It is an idealized company version that could be too attractive to some, especially for those who feel alone. It is the first stone of A new type of software that asks for another type of relationship. And we are not accustomed to something like that. It is not useful to treat it as another human, but neither as the classic tools. Another conceptual category will need. Chatgpt will know us better than many of our friends and family. It will be something that maintains the thread of our thoughts throughout weeks, months and years. A constant presence that will evolve with us, which will even anticipate our desires. Chatgpt will no longer be a tool, but something much more intimate and personal. Almost alive. Outstanding image | Xataka In Xataka | Openai’s hypothetical social network does not want to connect people. Want your data to train your AI

AI is also a tool to get a job: “No one will notice”

by

The curriculum is the only cartridge of the candidates To highlight at the beginning of the selection processes. Therefore, it must be optimized enough for ATS automated selection systems (Applicant Tracking System), but also Offer clear and concise information In case you check them a human recruiter. The use of AI in the elaboration of curriculums is not only an increasingly common trend, but already It is a common practice among those who seek employment efficiently. He 2025 market trends report From the Career Group Companies personnel selection platform, it points out that almost 65% of employment applicants already use AI in some phase of their application process. This figure shows how AI is not only transforming the labor marketbut also the Personnel selection stages. Far from be a “trap” In the process, the correct use of AI It can make the difference between unnoticed or capture the attention of recruiters. It is not enough to be good at your job, too You should know how to tell. It’s okay to use in your curriculum Kathleen Nolan, a senior recruiter at the Growthloop technology company, has reviewed hundreds of curriculums and stated in statements to CNBC that has no problem that candidates use AI to write their credentials. “Nobody cares if you use AI. If you use it well, nobody will notice,” said Nolan, who has been in the sector for seven years and personally reviews each curriculum he receives. Nolan stressed that LMost people already use AI for tasks like this and that the important thing is to know how to use it correctly. In fact, the recruitment expert recommends that candidates Use tools like chatgpt To generate a first draft for the presentation letter or for the curriculum, and then customize the result before sending it. For the recruiter, use AI to customize the curriculum is to take advantage of a tool available to everyone. No one would be surprised in the 21st century someone use a computer and a word processor to write your curriculum or your presentation letter. “It seems very hypocritical when I see the recruiters talk about that, because, come on, if I were in the labor market, I would also be doing it,” says Nolan to CNBC. The AI ​​is nothing more than a tool that facilitates and speeds up the adaptation of the curriculum to each offer, incorporating keywords and highlighting specific skills that appear in the description of the position. “I don’t know why it has become such a controversial issue. While it is used as intended, no one can know, and anyway should not matter,” says Nolan. This increases the chances of curriculum passes automatic filters and arrive at the hands of a human recruiter. After passing that first filter, the document must be clear enough to Show the information clearly and consistent so that the recruiter can get a clear idea of ​​your skills with a simple look. It should be taken into account that these people check hundreds or thousands of curriculums, so, in many cases, they can only dedicate them more than a few seconds. AI can help you identify the most relevant terms and to rewrite your achievements to give priority to the most prominent for the position to which I postulate. For example, you can ask you to summarize your qualification and experience using action verbs or that stands out Quantifiable results in previous jobs. Thus, your curriculum will be more attractive and will be better aligned with what companies are looking for, without losing authenticity or precision. The importance of reviewing the AI ​​text Although AI can save you a lot of time and improve the quality of the curriculum, it is always necessary to check the final text and give your personal touch. IA tools are trained with large volumes of data and, sometimes, They can hallucinate and include incorrect information. Therefore, it is essential to thoroughly review the result and correct possible errors to ensure that everything that the document appears is true and consistent with your experience. “I would not expect someone to write a personalized presentation letter for every opportunity, because that would consume a lot of time,” said Nolan. However, the recruiter remembered that, if it is going to be sent to one, it must ensure that it is well written and that it reflects the candidate’s enthusiasm for vacancy. Reading a “type” letter can generate the opposite effect on the recruiter. In Xataka | If your chair holds in a job interview, it is no accident: they are evaluating more than your curriculum In Xataka | Someone used AI to overcome an interview on Amazon. His success has made Google see his candidates face to face Image | Pexels (Edmond Dantès, Beyzaa Yurtkuran)

What is Google Firebase Studio and how this tool works to create artificial intelligence applications

by

Let’s explain what Google Firebase Studio is, the new tool for Create applications with artificial intelligence. You can do it without writing a line of code, just describing Google’s AI what you want, so that it generates the code for you. At the end of the process, you will have a functional web application that you can use and publish to see the light. Obviously, this is only prototyped, because you will also have the option to edit and review the code on your own to make sure everything is fine. What is Google Firebase Studio Google Firebase is the Webs and Applications Development Platform Created by Google. The idea is to cover most of the needs of developers, being able to store software, provide notification system for apps, necessary databases and authentication services or use analysis. And in 2025 Google has created Firebase Studio, which is a section within this platform that allows you Create webs and apps with artificial intelligence. You do not have to do anything, it’s something like Vibe Codingwhere AI generates the code and you just have to check it. The idea of ​​this service is that you simply have to Describe the type of application you want to create. For that, you must define what you want me to do in detail with a natural language. You have to describe what you want to do with the app, how it should be, its functions and anything else you can think of the same as if you will explain to a person. For this, Google Firebase Studio uses Geminithe artificial intelligence system of the company itself. Specifically use Gemini 2.0 Flashwhich will be the one that is responsible for interpreting your description, and to generate the application code in a few minutes. Of course, remember that what this tool will generate They are applications prototypes. Once you have it, you may have errors, and you will have to check the code, make the corrections you want and click on Post so that Google is in charge of hosting her. How to use Firebase Studio The first thing you have to do is enter the Firebase website, whose address is Firebase.google.com. Once inside, click on the button Begincontinue with the introductory steps that appear when you are going to use this program for the first time. This will take you to the Firebase Studio page, whose address to enter later is Studio.firebase.google.com. The tool interface is similar to that of an artificial intelligence chatbot, you have a writing field where you should Define what you want the application to do. Then click on Prototype with AI. This will take you to a Gemini page, where you are shown Suggested options and design For this application, as well as a name. On the right you have the icon of a pencil in case you want to make changes in the suggestion. When you have the app as you want, click on Prototype This App To create it. If you click on the edit button with the pencil iconyou will go to a page where you can manually change what you want within the description, including colors or functions. You also have a field to write the changes you want to make and that the AI ​​is in charge. When you finish click on Save. You will also have a button Prototype This App To start creating it from this window. And that’s it, then you’ll see in front of your eyes how the application code is createdand how after doing so, possible errors are sought and corrected. And that’s it. When you finish you will have a demonstration of the application to try it, and below you will have the button Edit of Code to review and make changes in the code. Up to the right you have a publishing button so that the web application is accessible. In Xataka Basics | Gemini Advance functions that become free in March 2025

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.