Abstract

Anthropomorphism is one of the most important concepts for future AI research. This is because the very term ‘Artificial intelligence (AI)’ implies the creation or reconstruction by human agents as a non-natural object of ‘intelligence,’ which has been considered as an element of the mind that only human being possess. Therefore, it is no coincidence that the academic discussion of ‘Human-Like AI’(HLAI) in the field of artificial intelligence has been developing in earnest in recent years beyond the current cultural imagination. However, if academic discussions on HLAI do not include reflections on humanity, they will be limited to superficial arguments based on researchers’ implicit understanding of humanity. Therefore, a deeper and more sustained dialogue between AI and human studies is required to advance HLAI research. In this paper, we discuss the limitations of a one-directional approach to HLAI and propose the need to move towards interdisciplinary research in terms of multi-dimensional hermeneutics. As an attempt at interdisciplinary research through dialogue with neuroscience, we explored the future of HLAI by suggesting the possibility of HLAI as a multi-dimensional and adaptive information bottleneck.