Discussion
In the following section, the structure and elements of preaching and
their effectiveness are further examined based on existing literature.
The transferability of this study and possibilities for future research
are discussed and explored.
The sermon structure adopted by ChatGPT was observed to be more
inductive in nature. An inductive sermon begins with an introduction
that only contains the first point of the sermon. Subsequent points
raised build on each other till the complete idea of the sermon is
developed before the preacher states the idea of the
sermon33. When delivering an inductive sermon, the
preacher would work towards a biblical solution to a human problem,
fostering a conversational tone rather than a lecturing style. ChatGPT
was observed to first introduce the issue of sin and conclude the sermon
with the core idea and solution to the issue of sin by acknowledging the
necessity of the transformative power of the gospel and the need to
accept it. This narrative-driven approach has a “special appeal to
inhabitants of a culture dominated by television and motion
pictures”34. This allows the preacher better connect
with modern audiences. Inductive sermons are also effective with
indifferent or hostile audiences who may reject the central idea if
stated at the beginning. This preaching style aids in sharing the gospel
with skeptical unbelievers and making an impact on the lives of the
listeners, as demonstrated in a study by Stoorvogel, van Vuuren, and de
Jong35.
Another observed technique used by ChatGPT is the inclusion of personal
testimonies within the sermon. This helps establish a bond between the
preacher and the listener, as well as lends authority to the preacher as
someone who has personally experienced the transformative power of the
gospel and can lead others to a similar experience, thereby enhancing
the appeal to listeners36. The use of invitations and
calls to respond to the gospel is common in the generated sermons and
proves effective in generating intentions for change among
listeners37.
The inclusion of rhetorical elements and invitation, coupled with the
inductive nature of the sermon, aims to persuade listeners of the core
message and direct them toward the sermon’s intended direction. Overall,
the results of this study demonstrate that ChatGPT is capable of
structuring and generating comprehensive evangelistic sermons that are
specifically targeted toward non-Christian listeners, resembling those
delivered by human preachers.
Regarding the transferability of this study, it is important to note the
limitations associated with the small sample size and ChatGPT’s
sensitivity to input prompts. However, the results provide valuable
insights for future studies on the use of AI in automating the preaching
process. They shed light on ChatGPT’s approach to preaching through the
development of a substantive grounded theory, which could contribute to
the formulation of a formal theory on ChatGPT’s preaching process.
This study has established that ChatGPT is capable of generating
theologically sound and comprehensive evangelistic sermons that closely
resemble those delivered by human preachers. It has also demonstrated
ChatGPT’s ability to fabricate personal testimonies that summarize human
experiences of divine intervention, although specific examples were not
provided. Furthermore, ChatGPT has shown proficiency in crafting and
leading prayers to God, a feature not observed in previous journalistic
studies. These findings highlight the potential for using AI language
models, such as ChatGPT, to automate the sermon preparation process and
potentially explore the concept of digital clergy. However, further
investigation is needed to determine the receptiveness of human
congregants to sermons generated by AI and their effectiveness in
instilling a desire for change among the congregants.
Further studies are needed to explore the implications of these findings
for the nature and practice of preaching, prayer, and the role of humans
within the Christian context. Additionally, it is crucial to delve
deeper into the theological and ethical issues surrounding the use of AI
in the sermon preparation process to enrich the ongoing discussions in
this area38. The impact of AI on the role of church
clergy and their relationship with congregants also warrants
investigation. Understanding the receptivity and sentiments of human
listeners towards AI-generated sermons is essential to gain insight into
the feasibility of automating the preaching process using AI language
models.