CHATBOTS represent a new frontier for organisations such as the Bible Society, and greater understanding is needed, it was suggested this week.
The authenticity of responses from artificial intelligence (AI) was the focus of a workshop, “AI, Bible Apps and Hermeneutics”, held in Cambridge.
A study, AI, Apps and Theological Bias, by academics in conjunction with the Bible Society, asked secular- and Christian-themed AI chatbots six questions, and the responses were analysed. Three related to biblical theology; the remainder asked for an explanation of specific verses. To avoid a Western bias, locations were spoofed (falsified technically) so that the chatbots thought that they were conversing in a different country.
Answers were graded on criteria including application to life, the author’s intent, literary content, and being able to interpret scripture with scripture. The chatbots showed a broadly conservative Evangelical bias while also offering pastoral responses.
Dr Jonas Kurlberg, who chairs the Global Network for Digital Theology, said that, as millions turned to AI for exegesis, the study was important.
“We need to understand the means and impact of how AI is shaping the way people read the Bible, how they interpret it, and also how it maybe changes their view in relation to the authority of the Bible, and their spirituality.”
The Bible Society’s director of content and digital strategy, Toby Beresford, said: “We are very blessed to be at the beginning of a new digital continent. We have new digital nations in our generation: 25 million people are already using Bible Chat.”
The workshop was organised jointly with the Cambridge Centre for Chinese Theology at Faculty of Divinity. A keynote address on the development of AI was given by Dr Thilo Stadelmann, Professor of Artificial Intelligence and Machine Learning at the ZHAW School of Engineering, in Switzerland.
The workshop was held to discuss the research findings, get feedback on the study, generate interest in further research in the intersection of the Bible and AI, and build a network of academics and non-profit organisations interested in the subject.
