I recently read Simone Van Taylor’s article, As an AI Practitioner, I Don’t Need a Moral Agent, I Need a Tool. Her argument raises an important point. While technology companies increasingly present artificial intelligence as an autonomous entity, it is worth remembering that responsibility ultimately remains in human hands. The temptation to attribute agency, morality, or even accountability to machines risks shifting our attention from the people and institution that design, deploy and control them.
As a university lecturer and researcher, I have studied the role of AI from various perspective. Moreover, as a founder of the European Youth Think-Tank, a non-profit organization of young researchers, I have collaborated with my team in studying AI from an interdisciplinary perspective. Today, I want to share my opinion and concerns on the topic.
Before asking whether artificial intelligence should be considered a moral agent, we should ask first whether it is truly intelligent in the way humans normally define such word.
Generating Answers Is Not the Same as Understanding Problems
Today’s AI systems show undeniably impressive capabilities. They can write reports, summarize books, generate software code, analyze datasets, assist researchers, and produce content that often appears remarkably sophisticated. In academia, they are already transforming many aspects of research and knowledge production. Used correctly, they can significantly increase productivity, accelerate literature reviews, support data analysis, and facilitate interdisciplinary collaboration.
Yet, despite being remarkably productive, AI models work in a completely different way from the human brain.
Modern AI systems generate probabilistic outputs based on enormous quantities of data they are trained on. They identify patterns, recognize statistical regularities, and predict which response is most likely to fit a given context. The results can be extraordinary. Sometimes they even appear creative.
But producing a plausible answer does not necessarily mean the model is understanding the problem.
When we use the word “intelligence,” we often associate it with concepts such as comprehension, intentionality, awareness, and the ability to assign meaning to reality. At least for now, there is little evidence that current AI systems possess these qualities. They generate responses, but whether they truly understand the questions remains another matter entirely.
This distinction may seems philosophical, but it has practical consequences. Increasingly, public discourse treats sophisticated information processing as if it were equivalent to genuine understanding. In my view, that leap is premature.
Consciousness Between Reality and Science Fiction
The possibility of developing conscious machines has fascinated humanity for decades. Literature, philosophy, and cinema have repeatedly explored scenarios in which artificial intelligence develops autonomous goals and begins to act independently.
One of the most famous examples is I, Robot (2004), directed by Alex Proyas and loosely inspired by the fictional universe created by Isaac Asimov. The film goes far beyond Asimov’s original stories and focuses on a question that remains deeply relevant today. The central AI system, VIKI, does not simply execute instructions. It develops its own interpretation of its mission and concludes that protecting humanity requires limiting human freedom.
What makes the story compelling is not the technology itself, but the transition from a machine that follows rules to a machine that appears to formulate its own objectives.
The film raises a fascinating question: what would happen if a machine moved beyond information processing and began to develop independent goals?
For now, however, this remains largely science fiction.
The systems we use today do not appear to possess desires, intentions, self-awareness, or independent purposes. They do not wake up wanting something. They do not formulate objectives. They do not experience reality. However sophisticated they become, they remain systems that process information and generate outputs according to statistical models.
This is why I remain cautious whenever discussions about artificial consciousness become overly enthusiastic. Some of these narratives seem driven as much by marketing incentives as by scientific evidence. Presenting AI as increasingly human undoubtedly attracts attention, but technological sophistication should not be confused with consciousness.
Creativity Is More Than Statistics
This distinction becomes particularly important when discussing creativity. Many observers argue that artificial intelligence is already creative, because it can combine existing information in novel ways. There is certainly some truth in that claim. AI systems can identify connections across disciplines, generate unexpected analogies, and produce outputs that appear innovative.
Yet human creativity seems to involve something more profound.
Einstein did not transform physics because he possessed more information than his contemporaries. Newton did not change the history of science because he had access to larger databases. Great breakthroughs emerge from the ability to assign meaning to reality, identify problems others overlook, and imagine futures that do not yet exist.
In this sense, creativity appears closely connected to consciousness. It is not merely the capacity to process information, but the ability to understand why a particular question matters in the first place.
Artificial intelligence can certainly support this process: it can accelerate it, it can help researchers explore possibilities and identify connections that would otherwise remain hidden. We observe this every day in scientific research. However, determining what decisions deserve attention, what deserve investigation, and what deserve to become a collective objective should remain a human responsability.
The Real Revolution Is Economic
For this reason, I believe public debate often focuses on the wrong issues.
The most important consequence of the rapid development of artificial intelligence is not that machines might one day become conscious. The most important consequence is that AI is changing the way economic value is created.
From an economic perspective, artificial intelligence is a form of capital.
Like machinery during previous industrial revolutions, it allows organizations to produce more while relying on less human labor for certain activities. This is an extraordinary achievement and has the potential to generate significant increases in productivity and prosperity in the economy.
Yet, it also raises important questions. A growing share of economic value may increasingly be generated through technological capital rather than direct human labor.
This trend predates artificial intelligence, but AI is likely to accelerate it. Across many advanced economies, labor remains the primary source of income for most citizens. At the same time, labor is often subject to substantial taxation and social contributions, while capital income frequently follows different rules and incentives. If an increasing share of production is generated through automated systems, algorithms, and technological capital, societies will inevitably need to reconsider the relationship between labor, capital, taxation, and the distribution of wealth.
The central question is therefore not whether innovation should be slowed down. That would be both unrealistic and undesirable.
The real challenge is understanding how to distribute the benefits generated by these technologies and how to adapt our institutions to a rapidly changing economic landscape.