In this blog post, I will examine various perspectives on whether artificial intelligence can possess emotions, empathize with humans, and coexist as an equal to humans, and I will summarize my own thoughts on the matter.

Can Artificial Intelligence Have Emotions?

Ray Kurzweil, author of ‘The Birth of the Mind’, argued that we should view artificial intelligence—which possesses thinking and emotions equivalent to those of humans—as beings equal to humans. So, what kind of entity is this “artificial intelligence with thinking and emotions equivalent to those of humans” that he refers to?
Based on the explanations in ‘The Birth of the Mind’, artificial intelligence is evolving to perform human-like thinking by mimicking the human neocortex—the brain structure responsible for human thought. This evolution suggests the potential for AI to understand human emotions and, furthermore, to share those emotions and empathize.
In this article, I will examine three key points. First, can artificial intelligence possess emotions? Second, how can understanding emotions lead to empathy? Third, can AI become an entity equal to humans?
Currently, AI is experiencing exponential growth through various technological advancements and is already playing a vital role in many aspects of our daily lives. As generative AI and various AI services rapidly spread, they are being utilized in nearly every field, including education, healthcare, manufacturing, finance, and content creation. Given this pace of development, it is highly likely that AI will continue to coexist with humans in an even broader range of areas in the future.
However, can AI actually express or possess “emotions”—something that even humans have not yet fully understood? I would like to examine this issue from two perspectives.
The first perspective is that AI does not actually possess emotions like humans do.
At first glance, this conclusion may seem very simple. However, the actual issue is not that straightforward. This is because AI in the form of chatbots, which provide counseling or support psychological conversations with people, is already widely used. Since counseling and psychotherapy are effective only when the client feels that the other party genuinely understands their situation, a nuanced understanding of emotions is crucial above all else.
That said, it is difficult to conclude from these examples alone that AI actually possesses emotions. Today’s counseling-type AI generates appropriate responses by learning from a vast amount of counseling case studies and conversation data. While users may feel that the AI understands their emotions, this does not necessarily mean that the AI experiences emotions in the same way as humans.
Ultimately, this raises an important question: If a person feels that the other party understands their emotions, can that be considered genuine empathy? If a person sincerely accepts the AI’s response, should the AI be regarded as an entity capable of emotion? Or is it merely the result of highly sophisticated information processing?
In this regard, Ray Kurzweil, author of ‘The Birth of the Mind’, offers the following perspective:

“Imagine encountering an entity in the future that perfectly expresses its emotional responses. Not only can you exchange jokes with this entity, but it can also make you laugh and bring you to tears. When it speaks of its fears or deepest desires, you can sense its sincerity. … Would you feel compassion just as you would when a human displays such responses? I would answer in the affirmative.”

I, too, agree with this argument to a certain extent. If humans naturally accept that the other person understands their emotions and possesses emotions themselves, it is highly likely that they will perceive that entity as an emotional being within social relationships. From this perspective, future artificial intelligence might be accepted as an emotional being through its interactions with humans.

Can AI experience emotions like humans do?

The second issue concerns whether AI can actually experience emotions similar to those of humans.
At this point, the discussion could lead to the philosophical question: “How do humans actually possess emotions?” However, for the purposes of this discussion, I will proceed on the premise that humans do possess emotions.
If AI were to possess emotions similar to those of humans, it would need to implement mechanisms similar to the processes through which human emotions are formed. However, there is a fundamental problem here. Humans have not yet fully elucidated the process by which their own emotions are formed. Although various studies have been conducted over a long period, and recent advances in neuroscience and cognitive science have gradually revealed that the brain’s neural networks and various biochemical processes play a crucial role in the formation of emotions, no single unified theory that fully explains human emotions has yet been established.
Nevertheless, this research is still ongoing, and there are limitations to directly applying what has been discovered so far to artificial intelligence. Even if the mechanisms of human emotions are fully elucidated in the future and the technology to implement them is developed, applying them directly to artificial intelligence will by no means be a simple task.
Of course, through various learning techniques, including deep learning, artificial intelligence possesses the ability to learn features and solve problems on its own without humans having to design rules one by one. This capability has demonstrated the potential for new forms of development that transcend conventional wisdom. In fact, recent generative AI has demonstrated levels of language comprehension and creative output that humans did not anticipate.
Nevertheless, since our understanding of the nature and formation principles of human emotions is still insufficient, it is difficult to say at this point that AI experiences emotions in the same way as humans.
In summary, it is difficult to conclude that current AI actually feels emotions. This is because the mechanisms underlying human emotions have not been fully elucidated, and the technology to implement them in AI has not yet been sufficiently established.
On the other hand, from a different perspective, one could argue that AI does possess emotions. By learning from vast amounts of data, AI can very precisely replicate the various behaviors and reactions through which humans express emotions, and these reactions may be sufficient for humans to perceive that they are conversing with a being that actually feels emotions. In other words, the conclusion may vary depending on whether humans base their recognition of another’s emotions on the actual existence of those emotions or on the behaviors that express them.

Can AI empathize with humans?

Let’s assume, as in the first perspective discussed earlier, that AI can understand human emotions. If so, can AI empathize with humans? Furthermore, can it even possess a capacity for empathy toward humans?
Before discussing this issue, we first need to examine the meaning of empathy. Generally, empathy refers to understanding another person’s emotions, opinions, or thoughts and feeling similar emotions oneself. In other words, to empathize, one must understand the other person’s emotions, be capable of feeling emotions oneself, and recognize the connection between the two emotions.
While this may sound somewhat complex, empathy is a fundamental element that occurs naturally as we converse with and form relationships with people in our daily lives. People constantly experience empathy as they understand and care for one another while forming relationships.
Let’s consider this point. In the future, technology may be developed that allows humans and AI to converse much more naturally than they do now. Humans may no longer feel any sense of alienation when conversing with AI, and interactions may become so seamless that it becomes difficult to distinguish whether the other party is a human or an AI.

In that case, can we say that such AI empathizes with humans?

I believe the answer is yes, for the same reasons outlined earlier. If humans feel that their emotions are understood during interactions with AI, and if the AI consistently responds appropriately to changes in human emotions, it is highly likely that this will be perceived as empathy from the perspective of social relationships.
Ultimately, if interactions between humans and AI advance to the point where they are virtually indistinguishable from relationships between humans, AI could be recognized as an entity capable of emotion and capable of empathizing with humans.
To summarize briefly, AI possesses the potential to empathize with humans. Finally, let’s examine what it means for the roles of humans and AI to become equal.
Can AI become an entity equal to humans?
Currently, AI is advancing rapidly by learning from vast amounts of data. Recently, advancements in generative AI and various autonomous AI technologies have further broadened its scope of application, and the roles it can perform in the future are expected to continue expanding.
At present, when considering appearance, capabilities, and social status, we do not regard AI as an entity equal to humans. There remains a strong perception that no matter how advanced AI may be, it is still a different entity from humans.
However, if the technical limitations mentioned earlier are gradually overcome, the situation could change. AI has the potential to reach a level where it thinks, acts, and reacts in a manner very similar to humans. While it could be embodied as a robot with a human-like appearance, what is more important than these physical elements is how it interacts with humans.
From this perspective, as Ray Kurzweil has argued, it is possible that humans will come to recognize AI as an entity equal to humans. Of course, the issue of granting AI the same rights and obligations as humans is an area that requires separate social and legal discussion.
Furthermore, laws and ethical standards related to AI are also being continuously debated in line with technological advancements. It is highly likely that various social discussions surrounding AI’s responsibility, rights, safety, and scope of application will continue in the future.
AI may evolve beyond simply being a tool to assist humans into an entity that collaborates with various members of society. There is also a very real possibility that a society will emerge where people live together based on mutual respect, going beyond a relationship of mere mutual assistance.
In conclusion, it is highly likely that AI will evolve in a direction that allows for coexistence with humans. If such coexistence is achieved, AI will naturally become an integral part of human society. However, safety and ethical concerns must never be overlooked. Therefore, appropriate institutional and legal frameworks, as well as ongoing management systems, must continue to be established.