In contemporary times, the human and social sciences have inherited a task that has existed for thousands of years dating back to antiquity: to think about what it means to be human. Philosophy started from the efforts of many great and ancient traditions – such as those in Greece, India, and China – to think about humanity and the universe and is considered the mother of many different academic fields that have developed leading up to the present. History, Literature, and Aesthetics have also all contributed to forming our understanding of what it means to be human and have helped create a rich foundation for us to think about this question. Furthermore, outgoing from this background, we find in the modern era the cultivation of human sciences (such as psychology, sociology, linguistics, and cultural anthropology) and social sciences (such as political science, economics, legal studies, and educational studies).
However, ever since the rapid development of brain imaging techniques (such as fMRI) took off around the 1980s, it has become possible to visualize the activity that takes place within a living human brain. This enables us to know in detail what has taken place in the human brain at the moment of “thinking” or “feeling” something. Using this technology, many neuroscientists have become able to take on research topics that have long been addressed by the humanities and social sciences (for example, philosophy and psychology). These research topics include consciousness, selfhood, society, and rationality. We can take “consciousness” as an example. Although in the middle of the 20th century, it was often said that scientists ought to avoid taking up consciousness as a research topic, during the 1990’s two large-scale academic associations for consciousness studies (the TSC and the ASSC) were founded and have been attracting many scientists ever since.
Looking in another direction, we can see that, after the continuous ups and downs of research on the topic in the 20th century, Artificial Intelligence (AI) was able to reclaim the spotlight after the development of “deep-learning” methods in the 2000’s. For instance, although it had been said that computers would not be able to beat professional players at the game Go for the foreseeable future, AlphaGo developed by Deep Mind was able to defeat some of the top pros in Korea and China (2016, 2017). While it is true that the AI boom is starting to somewhat calm down and we have entered a stage in which we need to calmly consider the possibilities it can offer us, it is still beyond doubt that the field of AI studies will have a grand impact on technology and society in the near future.
We ought to take note of the fact that this new AI technology is based on the neural networks that are using the functions of the human brain as a hint to create mathematical models. Research on AI (machine learning) is continuously being influenced by neuroscience and, conversely, there is a flourishing field of research called computational neuroscience, which uses computational methods to explore the functions of the brain. Thus, in contemporary times, the development of neuroscience and AI have been deeply intertwined with one another.
Considering this state of affairs, we must now once again ask: what does it mean to be human? The startling development of Neuroscience and AI gives to us new possible ways of approaching this question. Such new developments are capable of providing the human and social sciences with a substantial stimulus. For instance, many neuroscientists have started to use empirical experimentation as a means to enter into debates on “free-will.” In this way, neuroscientists will be brought face to face with debates in philosophy on this topic that have raged on since the classical period. Here, those who specialize in philosophy will certainly have plenty to contribute to such contemporary discussions in science. At this point, philosophers and scientists can make use of the knowledge that they can provide for one another and work together to create stimulating collaborative research about “humans.” To take another example, while many may be inclined to think that making an AI that is “driven by curiosity” would be rather odd, research in improving AI performance has already established this idea as a tried and true method (*). There has also been research that suggests agents that are capable of “getting bored” are actually superior to agents who cannot (**). Such research can in turn provide us with an interesting pathway to consider precisely why humans started to experience curiosity and boredom in the first place. Moreover, thinking deeply about the “human” can almost certainly provide a breakthrough to AI studies as well.
(*) Burda, Y., Edwards, H., Storkey, A., and Klimov, O. (2018). Exploration by random network distillation. ICLR 2019.
(**) Yu, Y., Chang, A.Y.C., Kanai, R. (2019). Boredom-Driven Curious Learning by Homeo-Heterostatic Value Gradients. Frontiers in Neurorobotics.