Some thoughts on (the relation of) AI and religion and on religion as such

If there is one thing everyone is talking about, if it is about cutting edge research funded with cash by numbers exceeding human comprehension, it is AI – so to speak the elephant in nearly every political, social, medical and academic room.

Although there are already some researchers around specialized in the relation of AI and religion like, among others, Beth Singler, Randall Reed and Robert Geraci, fundamental questions on the subject and frame haven’t yet been sufficiently discussed.

In the final session of the panel series “Artificial Intelligence and Religion” (AIR) 2020/21 organized by the “Fundazione Bruno Kessler” German professor of religious studies Inken Prohl (University of Heidelberg) made an extremely important notion on the question of means and ends of AI after raising some not less important questions:

The discussion about the ends is very interesting, but from the point of view of religious studies we have to ask: “Who is defining the ends and who is deciding about Artificial Intelligence and the development of Artificial Intelligence and who has the power in Artificial Intelligence and who is benefiting?” So I think particularly when I hear the word “outsmarting”, I think it’s a gender problem from the very beginning until the end. I mean, what gender has the idea to create a robot? Not women! Women are able to get children! So, I think we have to look into this whole discussion.[1]

Truly, gender is another elephant in most rooms today and frequently discussed in political, social and academic discourses – though not enough for some advocates of gender issues.

There is necessarily the gender dimension in religions to be understood (at least in the world religions as well as in the Eurasian ones), not just in the sphere of religious morals, in order to understand the development of religions in their varieties (including their similarities as well as differences, but which is not needed to be explored here any further – maybe on another occasion).

But by introducing gender to the actual topic (AI and religion) professor Prohl did not drop a smoke granted to deviate from the very topic to an allegedly “simple” social issue – instead she raised one of the most fundamental and mostly overseen questions related to the actual topic. I would broaden and translate her point in the following way: By asking (from the point of view of religious studies) “who is defining the ends” of Artificial Intelligence, it is clear that neither machines nor AI are defining the ends of AI, but humans do. So we basically have to deal with human beings to begin with or to be more precise, with the human morale. We often tend to deal with AI as it is unavoidable, nature given, necessary like the laws of nature. But that is obviously wrong. AI is man-made. So there is in fact already agency before there is a single functioning AI, no matter if we talk about weak- or strong-AIs. That means on the other hand that no matter, if they are aware of it or not, the responsible persons (developers of AI) already made a moral decision, even if they avoided the issue altogether.

But professor Prohl goes even further than this by at least implicitly claiming that there is in fact a dominant masculine-narrative preexistent, i.e. before AI is even developed, and this very dominant masculine- or male-narrative defines the means, and even more important, the ends in developing AI. The very idea that AI should aim for “outsmarting” people is a product of a narrative or part of an ideology which aims for domination over someone or something (humans, animals, plants, space etc.).

Surely, there are women in robotics and AI, I am sure professor Prohl wouldn’t deny this fact, but the basic ideology she sees here at work derives – at least from a gender perspective – from a dominantly masculine one, with all its negative implications.

That brings me finally to another talk of one of Germany’s most famous professors of astro-physics and philosophy of nature, Harald Lesch, who gave a talk in March this year at the adult education center of Munich with the title “Diktatur der Digitalisierung?” (Dictatorship of digitalization?).[2] During his talk, Lesch claims that AI should solely serve (!) human beings. From this moral point of view, AI could never be the ends to (human) agency, but only and always the means. AI as the servant of human beings. What makes this notion striking is the fact that the concept of service / servitude is from the point of view of cultural history a feminin concept. Lesch’s purpose for AI, which is obviously at odds to the common sense about AI in the Silicon Valley, proves indirectly Prohl’s thesis cited in the beginning, namely that if AI is understood as something e.g. to outsmart people, it is in fact a concept and if developed, a product of a certain dominant masculine narrative, based on severely questionable moral intentions. If a generalization of this observation should be legitimate we might have found the basics for a kind of morality check for AI development. While service / servitude is the means to life-giving and life-sustaining, outsmarting as the ends is to dominate and suppress, to rule over and exploit.

Interestingly, if this binary structure, to which Inken Prohl at least implicitly points, could be applied to religions as well, it would in fact indicate that some religions (like the western monotheistic ones) are not as dominantly masculine as they are often labeled, since service / servitude is an integral aspect to them. Hence, femininity seems to be part of their essence.

To be continued!

[1] https://www.youtube.com/watch?v=bxm9uZv2XcQ&list=PLY3KsH4gpiNrahXaGtvcFogKSIUG2ZAWY&index=14
[2] https://www.youtube.com/watch?v=Hs-JjgeOv8c