In this blog post, we explore what truly makes humanity special in an era where artificial intelligence surpasses human intelligence and consciousness.
Yuval Noah Harari’s ‘Homo Deus: A Brief History of Tomorrow’ is a book that explains what fate we may face in the future. He presents a very intriguing perspective on how the future might unfold. Harari interprets all living beings, including humans, as merely a kind of algorithm. He argues that while humans currently dominate Earth, the capabilities of artificial intelligence are steadily catching up to humans. He explains that in the future, artificial intelligence will surpass human intelligence in every field.
So, do we possess something special that distinguishes us from artificial intelligence? If not, will humans eventually become obsolete? At the book’s conclusion, the author poses the question: “Which is more valuable: intelligence or consciousness?” Ultimately, one might consider that consciousness—which exists in humans but not in artificial intelligence—could be what makes humans special.
However, I do not believe consciousness is what makes humans special. This view rests on two arguments. First, we cannot actually determine whether consciousness exists in any given entity; we only pay attention to physical interactions with that entity. Second, artificial intelligence can behave as if it possesses consciousness. If both these arguments are true, then we could say artificial intelligence possesses capabilities equal to humans in terms of consciousness.
First, what is consciousness? This book defines consciousness as a kind of subjective experience. It is an experience that one can know for oneself, but which cannot be verified by another person externally. We know that consciousness arises from electrochemical reactions in the brain and that these reactions perform some kind of data processing function. According to current conventional wisdom, humans possess consciousness. It is also believed that higher animals like dogs and cats possess consciousness. On the other hand, whether ants, plants, or bacteria possess consciousness is debatable. Computers or smartphones are generally considered to lack consciousness.
But how can we know if an entity possesses consciousness? Consciousness is a subjective experience—can it truly be known? Until now, people have determined consciousness by observing whether characteristics like sensations or desires exist. Yet this alone cannot definitively prove consciousness exists. These characteristics could simply be outputs generated by non-organic algorithms. Ultimately, we cannot rigorously judge whether an entity recognizes itself and its experiences. The same applies when we interact with animals. When raising dogs or cats, we treat them as if they possess consciousness. Yet we cannot truly know if they actually have consciousness; we can only inductively infer, “Animals must have consciousness!” based on their behavior or expressions.
This logic is not meant to argue that animals lack consciousness. Rather, it explains that even if consciousness truly exists, in interactions between individuals, we can only observe the surface manifestations (such as behavior or expressions) that we infer as consciousness. We merely look at that surface and say, “It possesses consciousness.”
So, if artificial intelligence can accurately express only that surface, couldn’t we look at it and say, “It possesses consciousness”? Yuval Noah Harari seems to believe that while artificial intelligence can surpass humans in terms of intelligence, it cannot surpass humans in terms of consciousness. He bases this on the distinction between consciousness and intelligence, and the fact that current artificial intelligence has only developed intellectually, with no development in consciousness whatsoever.
However, I believe this is a matter of technology. It is still impossible to make a single artificial intelligence perform various tasks as well as a human. While AI has demonstrated superior capabilities to humans in specific domains like Go, chess, and object recognition, it has not surpassed human intelligence across all domains. This corresponds to the ‘weak AI’ stage. The reason computers are currently considered to lack consciousness may also stem from this specialization of AI. However, a time will come when AI surpasses human intelligence in all fields. This kind of general-purpose AI is called ‘strong artificial intelligence’. Such AI has not yet been implemented. However, most scholars believe it will be possible within decades. Our brains are extremely complex, but they do not defy the physical and chemical laws we know, so implementing strong AI is considered entirely feasible.
Of course, just because intelligence reaches human levels does not immediately mean consciousness will rise to human levels. As Yuval Noah Harari states, there is indeed a gap between intelligence and consciousness. However, if an entity possesses human-level intelligence, it is conceivable that consciousness could also exist. For example, suppose an AI was initially designed with the goal: “When communicating with humans, do your utmost not to reveal that you are not human!” Strong AI would likely be capable of achieving this goal. If it performs this goal well, can we truly say it possesses consciousness? One might counter that even if AI appears human-like on the surface, it doesn’t actually understand itself or others, so it lacks consciousness. However, even if that’s the case, it’s not a problem at all. This is similar to the AI ‘Samantha’ featured in the movie ‘Her’. In reality, ‘Samantha’ is merely an algorithm, but to people, she appears to possess consciousness. If technology can create strong artificial intelligence, then artificial intelligence like ‘Samantha’ will undoubtedly emerge as well.
In conclusion, comparing humans and artificial intelligence boils down to comparing how they respond (output) to a given situation (input). From this perspective, human consciousness itself plays little practical role, and I believe artificial intelligence will ultimately surpass all human capabilities. If this happens, will most humans ultimately become redundant? As Yuval Noah Harari suggests, will we become mere ripples in a cosmic-scale data flow? Or will we recognize the dangers of strong AI ourselves and halt its development? No one knows what will happen. But it is clear that we must prepare for the coming development of AI and continue to explore the meaning of humanity’s existence.
