{"time":1759738338010,"blocks":[{"id":"09Zoj1v5s5","type":"paragraph","data":{"text":" Autonomous<\/span> agents are systems capable of performing complex tasks on the computer without permanent human intervention. Some believe that these tools are a decisive step in the direction of General Artificial Intelligence (AAO), but more importantly, if you have something to understand that allows the interaction between artificial intelligence, your own computer and the programs contained in it dute.Ez, this would be an example: imagine that you want to edit an <\/span> <\/span> "}},{"id":"WFiA_7c2uf","type":"paragraph","data":{"text":" audio and instruct a digital assistant to do so. When he receives the voice recording, he will take care of everything else himself: he will remove the noise, adjust the volume, add background music, balance the sounds and export the final result. It will do everything automatically, taking control of your computer and using the programs you normally use; you will not need to master any programs or press any buttons. <\/span> <\/span> "}},{"id":"VAtOixaJU2","type":"paragraph","data":{"text":" Many times, of course, you will know these tools, but even in these cases the assistant will be able to perform laborious tasks more quickly. Now imagine the same applied to presentations, texts, images or programming. You’ll see the agent do exactly what you’ve told him to do, that is, open the screens, press the buttons and execute the commands for you in front of your <\/span> <\/span> "}},{"id":"i34Bje2lIk","type":"paragraph","data":{"text":" eyes.On October 22, 2024, the company Anthropic presented the Claude 3.5 Sonnet model with the “Computer use” function. This allows the system to open applications from our computers, manipulate files, copy and paste texts, and automate various tasks without human presence. Some believe that on the path of General Artificial Intelligence, we have reached level 3 with it. At this level, in addition to providing textual answers, NEW agents horiek.AUKERA are able to use tools and interact with resources from <\/span> <\/span> "}},{"id":"mBkbySACqh","type":"paragraph","data":{"text":" the outside world, NEW DOUBTS<\/span> <\/span> "}},{"id":"JyZ0QDe0VX","type":"paragraph","data":{"text":" THAT THESE TOOLS CAN HELP IN SOME REPETITIVE TASKS. The same series of manual commands can be used to make them easier and faster. Even in environments for specific functions, they can also be very useful for optimizing the processes themselves. Unlike so far, they will have the expected capabilities, that ability to execute programs and processes bidez.Horrek that are present on our <\/span> <\/span> "}},{"id":"Ue17KlXc99","type":"paragraph","data":{"text":" devices, but, it brings clear risks. Some autonomous behaviors can be incomprehensible because the way these algorithms make decisions is not very transparent. There are also security and privacy risks, as these tools will have another access to our private data. Another security focus is the programs that will run on our system. Normally, Artificial Intelligence tools have been executed in the browser or in the controlled area, but now, if allowed, they will cross that area and there the possibility of losing control is much higher <\/span> <\/span> "}},{"id":"pNceBthdD6","type":"paragraph","data":{"text":" da.Ez with enough regulation: new approaches to coexistence are needed:• <\/span> <\/span> "}},{"id":"8KG-h-_i1k","type":"paragraph","data":{"text":" The new trainings: AI, data analysis, digital security and technological ethics.• <\/span> <\/span> "}},{"id":"ZPwYtyhiDn","type":"paragraph","data":{"text":" Collaborative AI Model: Having Helpful Tools, Not Simple Substitutes.• <\/span> <\/span> "}},{"id":"SDoEm1XfC5","type":"paragraph","data":{"text":" Hybrid teams: structures based on the cooperation of the person and the agent. <\/span> <\/span>"}},{"id":"hUEM_w1FgE","type":"paragraph","data":{"text":" In the history of<\/span> <\/span> the "}},{"id":"RV7u-Z-Ddl","type":"paragraph","data":{"text":" TRANSITION FULL OF POSSIBILITIES, great technological changes have led to social transformations. We are entering the <\/span> <\/span> "}},{"id":"ocm7-syAoD","type":"paragraph","data":{"text":" era of agent autonomy, and we can turn this transition into an opportunity if the time is to combine political courage, social competence and ethical vision baditugu.Orain: to put technological power at the service of man and not alderantziz.ADIMEN IN THE DIRECTION OF THE GENERAL<\/span> <\/span> "}},{"id":"DQ_Lc69wqj","type":"paragraph","data":{"text":" ARTIFICIAL [AAO], THESE FIVE LEVELS OF PROGRESS are not official standards, but have been formulated based on technological analysis around Claude, OpenAI and other strategic models.1 Level -<\/span> <\/span> "}},{"id":"4iNCNKbYfx","type":"paragraph","data":{"text":" Chat<\/span> <\/span> "}},{"id":"sylxDzqiaW","type":"paragraph","data":{"text":" Bots: Systems based on LLM (such as GPT-4o or Claude) tend to have coherent dialogues with text and perform basic tasks such as translations, summaries, content creation, etc. 2. <\/span> <\/span> "}},{"id":"OVW3dhx2ow","type":"paragraph","data":{"text":" Level - Reasoners: AI designed to operate with a human or doctoral level of competence. Solve complex problems, perform predictive analysis, generate hypotheses, and use advanced logic.3. Level <\/span> <\/span> "}},{"id":"uDqdLhsfMz","type":"paragraph","data":{"text":" - Autonomous agents: The AA systems are capable of acting on their own. They conduct negotiations, optimize processes and execute strategies without constant supervision. They are able to carry out joint planning and actions, operating 24 <\/span> <\/span> "}},{"id":"iBTrcj95Oj","type":"paragraph","data":{"text":" hours.4. Level - Autonomous innovators: AA are creators and researchers. They generate new ideas, develop patents, design original projects and collaborate with human beings for technological or scientific innovation.5 <\/span> <\/span> "}},{"id":"h-vs7LrH41","type":"paragraph","data":{"text":" Level - Autonomous organizations: They are systems capable of managing entire organizations that operate without human intervention. They automatically coordinate finance, logistics, marketing, and human resources, and by operating exclusively through AI <\/span> <\/span> "}},{"id":"DM5T5e-Mfc","type":"paragraph","data":{"text":" dira.Maila, the ladder goes from ordinary AI to systems that would work completely autonomously. The ladder, of course, does not problematize this achievement; it represents a concrete path and does not even question the achievement of the AAO to which it refers. <\/span> <\/span>"}}],"version":"2.30.7"}

Self-employed agents on the computer