Abstract: With the improvement of machine design and the proliferation of social application scenarios, more and more machines have been endowed with anthropomorphic features, which has led to a series of justifications for anthropomorphism in machines. The mainstream defenses can be summarized into three approaches to justifications: the instrumentalist approach represented by Brian Duffy, the cognitive approach represented by Louisa Damiano, and the relational approach represented by Mark Coeckelbergh. Based on the commonalities and differences of these three evidential approaches, the occurrence logic of machine anthropomorphism and its realistic roots can be analyzed from such two levels as the “design-use” context and the human-machine relationship scenarios. In addition, although the above three justification approaches have responded to and partially alleviated some ethical challenges to anthropomorphism in machines, there still remain some ethical challenges such as deception, acceptability, and new types of human-machine relationship, which need to be further clarified and addressed.

Key Words: Anthropomorphism in machines; Human-machine relationship; Relationalism; Design-use context; Acceptability

This article can be downloaded here

https://jdn.ucas.ac.cn/public/uploads/files/67078b5d6fa64.pdf