A lot has been written about the topic of empathy in the metaverse in the past time. Primarily it was about the question whether empathy can be experienced or felt in the metaverse. I am firmly convinced that this is possible and that it happens consciously or unconsciously when working in the metaverse.
However, this question takes on a new quality with the increasing capabilities of artificial intelligence (AI). Can an AI be empathic, and what impact does this have on virtual encounters in the metaverse. Specifically, the issue is whether in a situation where one avatar is a natural person and the opposite is an avatar controlled by an AI. One can approach this issue on two levels. One is a purely neurological approach. The other approach is more an ethical one. This question was already raised by John Wheeler (1) in his consideration. The prerequisite for the existence of empathy is not only the biochemical process, but also depends very much on our "I" understanding as a human being.
The question now arises whether the use of AI-controlled avatars in the metaverse creates a completely new situation? Basically, one has to say that superficially nothing changes in the basic statement. However, in the metaverse and the use of photorealistic avatars, further components are added. Through the immersion, i.e. the mental "immersion" in the virtual world, and the possibly objectively natural behavior of an AI-controlled avatar, something like a "mock empathy" can be conveyed. This is also the conclusion of Andrew McStay (2) in his article published in October 22 ("It from Bit") on the moral problem of an AI-controlled avatar. His conclusion is that while AI is able to provide large parts of empathy, it is incomplete in significant parts. Aspects such as responsibility, solidarity, community, etc. are missing.
In my opinion, these aspects must be taken into account when we think about ChatGPT and similar systems and their use in the metaverse. Basically, this development offers huge opportunities and the potential to create free space for areas where direct human-to-human interaction is necessary. But in the ethical evaluation of the development, we are just at the beginning, and we should conduct this discussion at least as forcefully as we think about new business models with AI.
(1) Wheeler, J.: Information, Physics, Quantum: The Search for Links. Proceedings of the 3rd international symposium on the founda- tions of quantum mechanics, Tokyo. https://philpapers.org/archi ve/WHEIPQ.pdf. Accessed 3 Oct 2022, (1989)
(2) McStay, A. Replica in the Metaverse: the moral problem with empathy in 'It from Bit'. AI Ethics (2022). https://doi.org/10.1007/s43681-022-00252-7
