Autonomous digital assistants
Unlike other industrial revolutions, at this very moment artificial intelligence is no longer a simple support tool; it has become a transformative force that redefines the foundations of our daily life, work and social interaction. In this context, the rumor of the possible arrival of GPT-5 represents a…

Unlike other industrial revolutions, at this very moment artificial intelligence is no longer a mere support tool; it has become a transformative force that redefines the foundations of our daily lives, our work and our social interaction. In this context, the rumor of the possible arrival of GPT-5 represents a turning point.
This model, more advanced and powerful than its predecessors, promises to bring us closer to a general artificial intelligence, capable of operating with a degree of autonomy that used to be seen only in science fiction.
OpenAI's models and other developments, such as Google's “Project Jarvis” and Microsoft's “Copilot Vision”, are driving a new generation of digital assistants capable of carrying out complex tasks independently.
These assistants no longer just respond to simple commands; they analyze information, interpret context and make decisions. This raises a crucial question: how far do we want, or need, these systems to be truly autonomous?
Autonomy in digital assistants: benefit or risk?
The benefits of a highly autonomous digital assistant are evident: they can simplify our lives, optimize processes and widen access to services that improve general well-being.
Consider a system like Google's Project Jarvis, which anticipates our daily needs —by booking appointments, researching complex topics, managing purchases—, or Copilot Vision, which proactively interprets our digital environment and acts as a true copilot in digital life. However, this autonomy also poses significant challenges.
A truly autonomous assistant is able to act without human intervention. That is useful for routine tasks, but it is a risk in critical areas where precision and human judgment are essential.
Like cyber-horror stories, tools such as OpenAI's Whisper have shown this complexity by generating “hallucinations”, or fictitious fragments of information, in audio transcriptions, which in medical settings can have serious, even fatal, consequences. These errors are not simple misunderstandings; they are complete fabrications that emerge from the models and, in critical cases, can put public trust and safety at risk.
Are we moving toward responsible AI?
AI is advancing faster than regulations can control it, and this is particularly evident in developing countries, where ethical and regulatory frameworks are still weak or under construction.
It is essential that both governments and companies set clear limits on the degree of autonomy we allow these assistants. Autonomy without oversight can lead to situations where an artificial intelligence makes decisions without full human context, replicating biases or acting without considering the ethical implications of its actions.
GPT-5 and its successors challenge us to build rules of the game that allow technology to advance without putting at risk privacy, data security and fundamental social values.
In his latest book, «Nexus», Yuval Noah Harari addresses the dilemma of human information networks. He argues that, unlike new technologies, books, papyri and even stones could not write by themselves. This makes us reflect on how our way of communicating as a species could change if artificial intelligence models continue to condition our main tool of evolution: language.
In critical areas such as health and education, it is imperative that these technologies act as a complement to human oversight, not a replacement for it. This list must be extended to other areas necessary to safeguard our survival. In light of Harari's ideas, it is essential to think in these terms today.
Redefining the role of work
As we delegate more tasks to AI, the demand for technical skills —and humanistic ones— in our workforce becomes more critical. The challenge is to train local talent to adapt to this transformation, preparing them to operate in a highly competitive global market.
To make the most of these tools, we need comprehensive digital literacy. Future generations must learn to work with AI not only from a technical standpoint, but also from an ethical and critical perspective.
In the Dominican Republic, the transition toward Industry 4.0 requires us to train professionals capable of using and managing these technologies, always with the aim of fostering sustainable economic and social development.
Autonomy with responsibility
GPT-5 and the new wave of autonomous digital assistants present an unprecedented opportunity to improve productivity and ease access to critical services. However, the real challenge is determining how far these systems should be autonomous.
AI autonomy must be handled with prudence and oversight. In my view, a digital assistant can be as autonomous as its context allows; in critical areas, human intervention will always be necessary as a safeguard of accuracy and ethics.
Artificial intelligence has the potential to profoundly transform modern life. Although its advance is inevitable and, in many respects, desirable, our responsibility is to ensure that its integration into society respects ethical principles and fundamental human rights.
We have in our hands an extraordinary tool for the common good, but to achieve that, we must be aware of its limitations and control its autonomy with a responsible and ethical approach.
The question we must ask ourselves is this: do we want this technology to be just a tool, or are we prepared to hand over decisions that used to belong only to humans?
The trend these days is to say that AI is a technology that is not only in our hands but that «relates to us». Such expressions can create a conditioning in the path we choose, which will shape the relationship between humans and machines in the years to come.