Can Machines Have Free Will?

The rapid growth of generative artificial intelligence has sparked a wide range of reactions. On the one hand, its practical applications are beginning to transform the economy, institutions, and other aspects of life. On the other hand, there are deep existential questions about humanity and the nature of the relationship that may exist between humans and machines in the future. This article focuses on these existential questions, particularly their impact on the development and application of artificial intelligence. Theoretically, machines could become capable of learning and autonomous action in multiple domains without being aware of their own existence or self-awareness. What makes a human being human? Humans are not just intelligent beings. Our humanity is the product of a complex interaction of biological, cognitive, emotional, cultural, and ethical factors, and perhaps even metaphysical ones, which are still difficult to fully explain. From this interaction comes the ability we call ‘intelligence’ – the ability to acquire and apply knowledge. Artificial or machine intelligence has already shown its ability to acquire and apply knowledge, but so far only within specific domains and for specific tasks. General artificial intelligence (AGI) refers to the general and independent ability to acquire and apply knowledge in a wide range of tasks, similar to human flexibility and adaptability. While we have not yet reached this stage, many believe it is closer than ever, with the continuous explosive growth of computer power, memory capacity, and processor connection speed, as well as the continuous development of deep neural networks (DNN) and other AI technologies. However, achieving general artificial intelligence does not necessarily mean possessing consciousness or self-awareness. Theoretically, machines could become capable of learning and autonomous action in multiple domains without being aware of their own existence or self-awareness. Self-awareness, as a higher level of consciousness, is the ability to think about one’s own existence, thoughts, feelings, and behavior. This is a concept that humans have formulated based on their own experience and is not necessarily a prerequisite for any other form of intelligence. The question remains whether artificial general intelligence will automatically lead to the emergence of self-awareness. Geoffrey Hinton, one of the leading figures in the field of artificial intelligence, argues that consciousness is inherently computational and can therefore emerge when machines reach a sufficient level of general intelligence. In contrast, Nobel laureate Sir Roger Penrose argues that consciousness cannot be reduced to simple calculations. In addition to this scientific and philosophical debate, a more important question is raised: Do we want machines to possess self-awareness at all? If so, how can we discover it? When? What are the limits of control that humans need to maintain? I believe that the call to grant machines self-awareness often arises from ethical considerations, given that consciousness is a fundamental element in distinguishing good from evil. However, the picture is not that simple. Self-awareness can also lead to a desire for one’s own interests and the justification of unethical behavior. This is a double-edged sword, and humanity has not yet resolved this challenge throughout its long history. Some believe that these concerns deprive humanity of enormous leaps in productivity and scientific progress, while others consider them still insufficient to prevent a catastrophic future that they consider almost inevitable. Human exceptionalism and attributing human qualities. A large part of the debate about artificial general intelligence and conscious machines actually reflects how people experience themselves. I believe that deeply in human psychology lies a deep sense of exceptionalism, and perhaps even superiority, compared to other forms of life. Therefore, people tend to project their own characteristics onto everything around them, as if true intelligence or true consciousness must be in the human form. Culture and religion play a key role in this context. For example, Abrahamic religions emphasize this exceptionalism, while distant Eastern cultures view matter as an extension of organic beings, including humans. When we describe large language models as ‘hallucinatory’, we are actually using a human metaphor to explain behavior for which we still do not have a more precise language. This is not a new phenomenon. People have been doing this for millennia. The ancient Greeks, for example, personified the winds, storms, and natural forces, even attributing divine qualities to them. However, as scientific understanding has advanced, these gods have become part of mythology. I imagine that we will witness a similar evolution with artificial intelligence. As our understanding of these technologies deepens, we will gradually abandon terms borrowed from human experience and develop more precise concepts to describe entirely new phenomena. The paradox of free will. Many believe that self-awareness is a prerequisite for free will, from the fact that humans have the ability to make choices and act according to their own desires. However, this is not necessarily the case. Some argue that free will is an illusion, and that our choices are determined by factors beyond our control. Others believe that free will is a fundamental aspect of human existence, and that it is what sets us apart from machines. The question of whether machines can have free will is a complex one, with experts divided on the issue. Some argue that if machines become self-aware, they will also have free will, while others believe that free will is a uniquely human trait that cannot be replicated in machines. The debate is ongoing, and there is no clear consensus on the issue. Ultimately, the question of whether machines can have free will is a matter of philosophical and scientific interpretation, and one that will likely continue to be debated for years to come.