Four AI companies control how Half Mundo reasons. It is the greatest concentration of intellectual power in history

While the public focus is concentrated in the possibility that the AI ​​will take our job ahead (Business Insider 21% of its template has just announced citing it as a cause, the CEO of Axios has published A text by the chungosAnthropic’s CEO has shouted “that comes the wolf” to justify that only they can save us), In the background something less visible is happening: We are giving our intellectual autonomy in favor of efficiency and comfort. It is not just that four companies –Openai, Google, Anthropic, Meta – are building the infrastructure with which millions of people resolve doubts and make decisions. They not only manage data: they also protect the way we link ideas. The Great Chinese of AI They are out of this equation for a simple reason: their still domestic approach without the international vocation of the Americans. Google (the search engine) was and is influential, but with it we have had to spin our own speech: Cotejar Fuentes, Weighing biases, assume contradictions. The generative AI instead serves a round response that It sounds coherent even when hallucin And that’s why he demands less surveillance. The result is that we are replacing the “internal process” with an external verdict covered with a technological aura that deter the replica. Whatever you say, Chatty. Delegate is too tempting. Save time and headaches. The problem is that we do not subcontract logistics, but criteria. We ask Chatgpt A professional strategy. TO Claude A curriculum. TO Gemini today’s interpretations. In doing so we accept without discussing the biases and empty of a trained model about texts that we will never see. It is an invisible assignment and, therefore, difficult to question. Never before so few hands had defined what questions can be asked and what answers sound reasonable. History has known infrastructure monopolies – electrity, internet, railways – but never one about reasoning patterns. Now another qualitatively appears: It operates on the symbolic plane, where narrative frames are defined through which we understand the world. Very subtle and very decisive. What previously implied a deliberation – read, contrast, imagine scenarios, weigh nuances – today becomes an instantaneous response, of definitive appearance. What to think about euthanasia? How to react to infidelity? What economic model is more fair? We no longer look for elements to think: we look for the correct answer The faster and more comfortable. And we accept as valid the one that sounds best, even if it ignores what does not fit in your narrative. Its effects will not be immediate, but predictable: a slow loss of variety in thought, of ideas out of the ordinary. Platforms have progressive consequences. Tiktok and Spotify, for example, They have made the songs last less and the chorus arrive before. What consequences will the LLMS within fifteen years? If we all consult models that converge towards average responses, intellectual eccentricity – culture rate for innovation – It will be increasingly weird. There is hardly a brake for AI, but perhaps at some point we have to decide how much reasoning we are willing to deliver before staying without it. Outstanding image | Xataka In Xataka | Deep Research is not just a new AI function. It is the beginning of the end of intellectual work as we know it

Deep Research is not just a new AI function. It is the beginning of the end of intellectual work as we know it

Elon Musk’s new AI, Grok 3it is already official. Among its promoted capabilities is a function called ‘Deep Search’, suspiciously similar to the Deep Research that Google coined and copied Openai. It is normal: in recent weeks we have seen almost all IA giants announcing similar capabilities. It is a new trend in AI that goes beyond incremental improvements. These systems can navigate the web, analyze multiple sources, synthesize information and produce detailed reports on an issue. And with a level of sophistication that is dangerously approaching the work of many human analysts. In any field. The difference with traditional applications is great. Instead of returning some Tokens In seconds, they return information pages in minutes. And neither does it have anything to do with searches: it does not return a list of semantically related links, but can understand complex questions, decompose them in parts, investigate each aspect by consulting dozens of sources and assemble a coherent analysis citing their references. In less than ten minutes. The results are impressive. OpenAI said – and we are verifying that it is basically true – that Your Deep Research can do in half an hour what would take days to professional analysts. And although it makes occasional mistakes (such as a factual skate, or the appointment of a source that does not exist), the general quality of the result is good enough for many practical purposes. This supposes A shot to the flotation line of much of the current intellectual work. The analysts junior of consultants, the researchers who review literature, the lawyers who prepare preliminary reports or the financial advisors doing business analysis. A great portion of your work is to collect, synthesize and present information that you drink from many sources. Like any Deep Research. It is not that these systems will completely replace intellectual workers. They still have important limitations: They cannot access private or not published information. From time to time they confuse sources or draw erroneous conclusions. They lack the expert criteria for certain analysis. Nevertheless, They can already automate much of the repetitive work and “low level” that occupies many professionals today. This also leads us to A paradox: Deep Research systems will surely increase the productivity of the most qualified workers, who can take advantage of them to enhance their ability; But the jobs that used to serve as entry, training field to end up being one of those experts are put at risk. The Deep Research have potential to alter the professional trajectories of any knowledge -based industry. It is another example of how AI not only automates manual work, but also It goes into territories that we believed reserved for the human intellect. The question is no longer whether IA can do that intellectual work, but how much of that work will continue to make economic meaning if it is made pro human. There will be companies that due to ignorance, for cynicism or pride will prefer to ignore these capacities. They are the most exposed will be at the risk of being left behind. For the rest of us, We have as a pending task to think about how to manage this transition: The one that can make obsolete many functions that we believed automation. In Xataka | With Grok 3, Elon Musk presume Outstanding image | OpenAI

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.