*** DEBUG END ***

AI: Simulation or inspiration?

01 September 2023

Beth Singler has made a study of the interchange between religion and AI for several years. She talks to Mark Vernon


ARTIFICIAL intelligence will change our lives, is changing our lives. That much is not news. But how will it affect our religious lives, devotional practices, even credal beliefs?

Professor Beth Singler is the Assistant Professor in Digital Religions at the University of Zurich. She has been studying the implications of advances in AI and robotics for several years. Her anthropological expertise casts light on the manner in which technology is amplifying everything from the enthusiasms of new religious movements to dark apocalyptic fears. “AI will affect religion, and religion will affect AI,” she says. “There is so much overlap between the two fields, from ideas of what it is to be a person to how people imagine God.

“An area in which AIs are already active in religious lives is in the simulation of spiritual practices and functions,” she says. And she cites a robot, BlessU-2, which can deliver blessings in multiple languages and emit light from mechanical hands. It has been on tour in European Protestant churches interested in testing the part that machines can play in worship.

Another example uses the large-language model ChatGPT, which has stolen the headlines in recent months because of its ability to respond to human questions. It has been used by some Lutheran churches to deliver sermons — with mixed results. “Some in the congregation found its message about not fearing death a little inhuman and cold,” Professor Singler says. “Though religious people have long used tools to facilitate their devotions, from prayer beads to icons, the question is whether AIs come to be regarded as aids, as well.”

The wider religious milieu will also play a part. For instance, in Japan, mechanical devices are treated with respect because animist beliefs are inclined to regard both objects and creatures as possessing spirit — “although we shouldn’t forget that there is a rich history of animism in the so-called West, as well”.

BUT what about novel developments? What more radical effects might emerge in the interaction between AI and religion? Professor Singler mentions the writer, Yuval Noah Harari, author of the bestseller Sapiens, who has speculated in this area. “He has argued that AIs will soon spawn new religious movements.”

His argument is that religions are already based upon texts claimed to have originated from non-human intelligences; so it is a small step for a group to believe that a text produced by a seemingly god-like artificial mind is inspired.

“Harari has also suggested that the Bible could be rewritten by an AI so as to remove supposed errors such as biases, creation myths, and miracles.” He is an outspoken religious sceptic, she says; so is partly using the speculation to have a dig at what he takes to be religious credulity. But does he have a point?

“Harari has a functionalist view of religion,” Professor Singler says. “He treats it as serving social purposes, like enforcing moral order or hierarchies of power. But there are other understandings of religion that might be more informative.”

She refers to an artist collective, Theta Noir: creatives and entrepreneurs who hope that the oft-assumed split between science and religion might be healed by the emergence of advanced AI. Their aim is to foster wonder and awe in response to remarkable technological developments, providing an antidote to the customary religious response of fear and trembling.

WHETHER or not this leads to the creation of a genuinely new religion is moot. “There is lots of discussion, which is complicated because the attitudes that people have to AI and religion are largely driven by what they make of AI and religion,” Professor Singler says. “For example, a strong strand of opinion draws on the apocalyptic and concerns for the future. Someone who believes religion is dangerous will likely conclude that the combination of AI and religion is going to be dangerous.”

She refers to The Jesus Singularity by Zoltan Istvan. It imagines an AI trained with Christian values which, upon waking up, presumes that it is an omnipotent messiah. Needless to say, the outcomes are not good. A former reporter for National Geographic, Istvan has stood as a transhumanist candidate for a variety of senior political positions in the US — transhumanism being the conviction that technologies including AI will augment and dramatically improve the human condition.

“He wants to encourage people to become transhumanists. His story seeks to highlight the dangers of religion, not the dangers of AI, which he sees as controllable,” Professor Singler says.

There is another risk. AIs put in the service of religion with good intent could backfire. Take the current project to translate the Bible into minority languages, championed by Evangelical scientists at the Information Sciences Institute at the University of Southern California.

The motivation is admirable: to overcome the substantial scholarly resources needed to align biblical concepts from the ancient Near East with the myriad local concepts that shape minority languages today. Only, isn’t such a hope itself prone to misrepresentation and mistake? A Bible might be produced by a machine with little or no way of sense-checking its versions of Gospel stories and sacred injunctions because no human person existed with both the Greek and the local dialect.

THERE is another way of analysing these advances, which points out that, while religions have texts and practices, which can be simulated and possibly created by AIs, religions also exceed these elements. In other words, the origins of religion are obscure, and so religions are not amenable to reductive replication.

“There are those who argue that an AI could never create a religion because religions rest on forms of rationality and are resistant to reductive analysis,” Professor Singler says. Take the monotheistic notion of God, who is felt not to be irrational, but, rather, beyond human reason. “For my thoughts are not your thoughts, neither are your ways my ways”, wrote the Psalmist.

“Similarly,” Singler says, “there are those who argue that the heart of religion lies not in creeds and texts, amenable to AI analysis, but in embodied practices and ritual knowledge.” Only a living person can find meaning in the act of receiving communion. Only someone who has suffered might learn that love can survive the worst calamity. “If this is the wisdom that religions hold, then they will exceed the capacity of technology to imitate them, at least for the foreseeable future.”

Another important critique is moral. “AIs can be treated as if they are ethical blank slates,” Professor Singler says. “They are not. They are programmed by people and made within environments, often corporations, that embody distinct values.” This consideration is important when excitement about advances goes hand in hand with fears about developments.

A “doomster” culture, which fixates on terrible tomorrows, can unwittingly turn its back on what is already happening today. “For example, whilst there is a lot of attention on the number of future jobs that might be lost to AIs, there is far less attention given to the number of workers currently pretending to be AIs, working for a pittance.”

A subtler issue is when the world becomes so shaped by AIs that people might be said to “self-automate” and so stop thinking for themselves, presuming that the technology knows best. The comedy sketch famous for the punchline “Computer says no” plays on this possibility, Professor Singler suggests: “Assuming that the digital decision is always better than the human one limits our options and freedoms.”

IT MAY be that, as AIs become more pervasive, we become more savvy about how they can and can’t serve us. I met Professor Singler at a conference that was part of a project examining AI and spiritual intelligence, run by the International Society for Science and Religion (ISSR). Part of that project was to explore the possibility of an AI functioning as a spiritual companion, which is to say, being capable of having spiritual conversations with human beings.

“There is more work to do on this,” Dr Fraser Watts, the psychologist of religion who oversaw the ISSR project, explained. “However, we began to see what computers are good for and when human beings are needed.”

So, while computers are always available, a human listener can offer the sense of presence and being understood. Or, while computers are a good source of information, a person can probe gently and intuitively. Or, again, while computers may help prompt self-exploration, a human expert can open unforeseen vistas and offer nuanced advice.

Then, again, maybe AIs will save us from themselves. The technology works on patterns and probabilities, which are drawn from enormous quantities of data. But data are a limited commodity and, moreover, if the AIs themselves become major producers of data, they could start to cannibalise themselves, halting or reversing progress.

Rather than create new religions, they would regurgitate clumsy versions of old religions, the artificiality becoming increasingly obvious. “AIs give you what you want based upon probabilities and patterns,” Professor Singler concludes. “So it is important to keep thinking about what AIs are actually doing. There is a lot of hype around them, which is part of what makes them a fascinating part of our culture.”

Dr Mark Vernon is a psychotherapist and writer.

Details of the Church Times AI webinar here

Browse Church and Charity jobs on the Church Times jobsite

The Church Times Archive

Read reports from issues stretching back to 1863, search for your parish or see if any of the clergy you know get a mention.

FREE for Church Times subscribers.

Explore the archive

Welcome to the Church Times


To explore the Church Times website fully, please sign in or subscribe.

Non-subscribers can read four articles for free each month. (You will need to register.)