Let me start by saying that I am not an expert, but rather «curious,» and despite my age, I have an interest in learning. I also have the privilege of being in contact with those interested in the subject. In this regard, I observe that AI is being implemented in very defined phases. Initially, we became familiar with assistants and chatbots. The evidence shows that it is developing in phases or stages.
Predictive AI Phase: Analyzes based on previous data. Historical data, to anticipate possible outcomes. It also relies on repeated patterns. It is used to support decisions.
Generative AI Phase: Adds value, generates new value. It interprets information and creates responses, new content, and linguistic interactions. It does not only work with texts; it contributes documents and knowledge. It accelerates tasks.
Agentic AI Phase, or «AI Agents»: A phase currently in development. In my opinion, this is a further step of advancement with significant impact. It is not limited to analysis; it is already executing commands and performing tasks.
Although it may seem that AI is more associated with employment and companies, the reality is that it is already present at the popular and domestic level. It has become the primary source of information. Among others, ChatGPT, Gemini, Copilot, Grok are already part of our lives, along with Perplexiti and Claude as AI agents.
Related to new technologies and other changes being generated, there is a concept that I find fascinating: «unlearning.» Undoubtedly, many of our knowledge and skills are already obsolete; we need to forget them and think about new ones. We must relearn. Unlearn to learn again. And in this field, it is essential to «know how to ask questions.» The prompts are the door, the key to entry.
One aspect that has caught my attention is that we train AI and AI agents ourselves. Are we training the AI agent that could replace us? It seems that AI agents, as new work companions, learn from us, from our skills, learning from our histories and files, from our applications. Should we mainly create scenarios where agents do the work and our role is to direct, review, and help them improve?
Regarding employment, I personally believe that AI will not eliminate many jobs, but it will transform the labor landscape. I think many professionals believe that AI and agents are not in a position to occupy their jobs, but undoubtedly a different work dynamic is emerging, with the displacement of some tasks and changes in approaches and processes. I understand that we are really no longer competing against each other; the goal is to improve your «competencies» in using AI.
However, I am also struck by the fact that currently, it is technology companies that plan to lay off the most employees. Specifically, Meta has announced a 10% workforce reduction, which will affect about 8,000 employees, but also states that it will increase its investment in AI by $135,000. According to Mark Zuckerberg, «2026 is going to be the year when AI will drastically change the way we work.»
This is a warning for navigators. It will undoubtedly shape company strategies. In the same vein, I believe that AI is not a threat, it is not an enemy, it will be our work companion and must be our ally. How do I integrate it into my activities? Undoubtedly, there are experts, training actions, but initially, it is essential to lose our fear, be proactive, practice, and engage in a lot of trial/error.
Miguel Ángel Otin Lloro
Secretary General, Huesca Excellent Business Forum
