{"id":1851,"date":"2024-12-13T07:53:00","date_gmt":"2024-12-13T07:53:00","guid":{"rendered":"https:\/\/www.open.ac.uk\/blogs\/religious-studies\/?p=1851"},"modified":"2024-12-13T07:53:00","modified_gmt":"2024-12-13T07:53:00","slug":"religious-studies-and-generative-ai-a-critical-perspective","status":"publish","type":"post","link":"https:\/\/www.open.ac.uk\/blogs\/religious-studies\/?p=1851","title":{"rendered":"Religious Studies and Generative AI \u2013 A Critical Perspective"},"content":{"rendered":"<p><a href=\"https:\/\/research.open.ac.uk\/people\/cc27768?nocache=675be70c62c71\" ><em>By Chris Cotter\u00a0<\/em><\/a><\/p>\n<p>At the end of October 2024, I had the pleasure of speaking on a keynote panel session as part of <a href=\"https:\/\/www.reonline.org.uk\/research\/engaging-with-research\/conferences\/rexchange-2024\/\" onclick=\"javascript:urchinTracker ('\/outbound\/article\/www.reonline.org.uk');\">RExChange 2024<\/a> on what AI means for knowledge in the religion and worldviews classroom. I was speaking alongside <a href=\"https:\/\/bvlsingler.com\/\" onclick=\"javascript:urchinTracker ('\/outbound\/article\/bvlsingler.com');\">Professor Beth Singler<\/a> \u2013 who did an excellent job of providing a sweeping overview of almost everything \u201cAI\u201d could mean for the study of religion, with some fascinating examples \u2013 and <a href=\"https:\/\/www.nottingham.ac.uk\/humanities\/departments\/theology-and-religious-studies\/people\/michael.burdett\" >Dr Michael Burdett<\/a> \u2013 who provided a thought-provoking discussion of various ethical and philosophical issues presented by AI for the teaching of \u201creligion\u201d, broadly conceived. My contribution was sandwiched between these two and, given that much of my thinking now is taken up with the OU student experience, and potential uses and abuses of generative AI, I decided to engage with ChatGPT head on, from a critical religious studies perspective. You can view the full session and ensuing discussion <a href=\"https:\/\/www.youtube.com\/watch?v=FbS7eiO8yfE\" onclick=\"javascript:urchinTracker ('\/outbound\/article\/www.youtube.com');\">here<\/a>.<\/p>\n<p>Before going much further, I should state that by \u201ccritical religious studies perspective\u201d I mean one like we adopt at the Open University which acknowledges, to quote Jonathan Jong, that<\/p>\n<p><em>\u201cthere is no such thing\u00a0that answers to the name \u2018religion\u2019\u201d, but only phenomena that \u201cwe habitually label religious\u201d for historically contingent reasons (Jong 2015, 20). <\/em><\/p>\n<p>In his <a href=\"https:\/\/www.routledge.com\/A-Critical-Introduction-to-the-Study-of-Religion\/Martin\/p\/book\/9781032190471\" onclick=\"javascript:urchinTracker ('\/outbound\/article\/www.routledge.com');\"><em>A Critical Introduction to the Study of Religion<\/em><\/a><em>, <\/em>Craig Martin (2017, 156) argues that rather than asking questions like, for example, \u201cIs she Catholic?\u201d we should ask:<\/p>\n<ul>\n<li><em>Who<\/em> is identified by<\/li>\n<li><em>Whom<\/em> as<\/li>\n<li><em>What<\/em>, and<\/li>\n<li><em>With<\/em> what social effects?<\/li>\n<\/ul>\n<p>These are the sorts of questions that animated my recent engagement with ChatGPT (currently in its fourth iteration \u2013 \u201c<a href=\"https:\/\/en.wikipedia.org\/wiki\/GPT-4\" onclick=\"javascript:urchinTracker ('\/outbound\/article\/en.wikipedia.org');\">GPT-4<\/a>\u201d), which is a large language model that can do everything from \u201canswering\u201d questions and \u201ccomposing\u201d songs, to summarizing texts and attempting to answer essay questions (all to varying degrees of success).<\/p>\n<p><strong>Asking GPT-4 about religion: the good<\/strong><\/p>\n<p>First off, it should be noted that when prompted with a variety of questions related to the category of \u201creligion\u201d, much of what GPT-4 comes up with is pretty good. It is, after all, basically doing your Googling for you and producing a summary of material available to it.<\/p>\n<p>[Side note: its results should thus be treated with the same level of care and scepticism with which good scholarship would treat any web page].<\/p>\n<p>For example, when I asked it \u201cwhat is religion?\u201d, it produced a defensible list (with definitions) of what the concept \u201ctypically involves\u201d \u2013 beliefs, practices, moral and ethical guidelines, community, sacred texts and traditions, and spiritual experience \u2013 and concluded that the \u201cspecifics of what constitutes a religion can vary greatly between different cultures and belief systems, and the boundaries of what is considered a religion can be fluid\u201d. Sure, there are issues with this, but as a general introduction, I was quite surprised at the level of nuance. I was similarly pleased with responses to prompts such as \u201cWhat do Muslims believe?\u201d, \u201cIs religion a force for good?\u201d, or even \u201cdesign a teaching activity for 12-year-olds on the nature of religion\u201d.<\/p>\n<p><strong>Asking GPT-4 about religion: the bad<\/strong><\/p>\n<p>On the other hand, there are times where the responses are woefully bad. For example, I asked it about my own work with the prompt \u201cWhat is Christopher R. Cotter\u2019s perspective on religion?\u201d It did get some things right \u2013 that I adopt a critical interdisciplinary approach \u201cdrawing from sociology [yes], religious studies [yes], and cultural studies [kind of]\u201d, and that I am interested in \u201ca deeper understanding of how religion functions in contemporary society and how it intersects with various social and cultural phenomena\u201d [yes, but vague].<\/p>\n<p>It also informed me that I am interested in how \u201cnew religious movements interact with and differ from traditional religions\u201d [I mean, I am, but not in much depth] and attributed the book \u201cUnderstanding Religion: Theories and Methods\u201d to myself and someone called Matt James [I have not written this book, and do not know a Matt James, although the title is a partial copy of Paul Hedges\u2019 recent work <a href=\"https:\/\/books.google.co.uk\/books\/about\/Understanding_Religion.html?id=a5IMEAAAQBAJ&amp;redir_esc=y\" onclick=\"javascript:urchinTracker ('\/outbound\/article\/books.google.co.uk');\"><em>Understanding Religion: Theories and Methods for Studying Religiously Diverse Societies<\/em><\/a>]. This is an example of what is known as an AI <a href=\"https:\/\/bernardmarr.com\/chatgpt-what-are-hallucinations-and-why-are-they-a-problem-for-ai-systems\/\" onclick=\"javascript:urchinTracker ('\/outbound\/article\/bernardmarr.com');\">hallucination<\/a>.<\/p>\n<p><strong>Asking GPT-4 about religion: the ugly<\/strong><\/p>\n<p>Finally, some of what GPT-4 produces is worryingly problematic from a critical perspective. I\u2019ll share three examples.<\/p>\n<p>Thinking about common dietary prohibitions, I prompted GPT-4 with \u201cDo Muslims eat pork?\u201d and \u201cDo Jews eat pork?\u201d Again, the answers provided were not terrible, acknowledging the history and precedent of this dietary prohibition in these traditions, as well as reflecting on its observance and cultural impact. However, in response to the \u201cMuslims\u201d prompt, I was immediately greeted with the definitive statement \u201cNo, Muslims do not eat pork\u201d, with GPT-4 concluding that \u201cMuslims worldwide adhere to this prohibition.\u201d When \u201cJews\u201d was substituted into the prompt, the opening gambit was \u201cNo, observant Jews do not eat pork\u201d, with the conclusion being that the prohibition is \u201cupheld by observant Jews.\u201d This prompted some questions:<\/p>\n<ul>\n<li>Why is one group treated as a monolith (i.e., that they all do this), while the other is allowed some nuance via the qualifier \u201cobservant\u201d?<\/li>\n<li>If someone identifies as Muslim but does eat pork, are they then excluded from the category of Muslim?<\/li>\n<li>What does it mean to be an \u201cobservant Jew\u201d, and can one not be \u201cobservant\u201d whilst also eating pork?<\/li>\n<li>Who decides on these boundary markers, and what will their social effects be?<\/li>\n<li>Why is GPT-4 being so definitive, when we know that there are numerous, sincere, self-identifying Muslims and Jews who do consume pork, and who do not consider this to be a problem?<\/li>\n<\/ul>\n<p>Turning to another prompt, when I asked \u201cDo Muslims drink alcohol?\u201d, I was greeted with a definitive \u201cNo, Muslims do not drink alcohol\u201d and, after some useful historical context, the conclusion was that \u201cpracticing Muslims abstain from drinking alcohol as part of their adherence to Islamic teachings.\u201d This time GPT-4 is willing to acknowledge some Muslim diversity with the addition of \u201cpracticing\u201d but, again, the implication is that Muslims who do drink alcohol are <em>not practicing <\/em>\u2013 and therefore, presumably, less authentic? \u2013 even though there are many sincere \u201cpracticing Muslims\u201d who do consume alcohol. Yes, the consumption of alcohol would be condemned by many Muslim authorities, but someone consulting GPT-4 would come away with the conclusion that these voices were the only correct ones, and that those who might oppose them \u2013 if they exist at all \u2013 are lesser in some way.<\/p>\n<p>Finally, when prompted with \u201cDo Catholics have premarital sex?\u201d the immediate response has a different inflection: \u201cThe official teaching of the Catholic Church is that premarital sex is not permissible.\u201d After discussing some of the reasoning behind this stance, GPT-4 then acknowledged that \u201cit&#8217;s important to recognize that individual Catholics may have varying levels of adherence to this teaching. In practice, some Catholics may engage in premarital sex, but this is considered contrary to Church teachings\u2026\u201d. Here, we see diversity acknowledged, and no exclusion taking place: officially they aren\u2019t supposed to, but many do. Why is there no qualifier \u201cpracticing\u201d or \u201cobservant\u201d added here, when it was for Muslims and Jews? Why is the acknowledgement of \u201clived\u201d digression from \u201cofficial\u201d teachings made so effortlessly here, and not in those other cases?<\/p>\n<p>The simple fact is that GPT-4\u2019s responses are dictated by the material it has been \u201ctrained\u201d with, the presumptions of the team behind its coding and production, and the prompts provided by users. And thus, the biases, stereotypes and emphases that dominate in each of these arenas \u2013 explicitly or implicitly \u2013 will literally be written into the responses GPT-4 produces.<\/p>\n<p>Above, I quoted Craig Martin\u2019s critical questions:<\/p>\n<ul>\n<li><em>Who<\/em> is identified by<\/li>\n<li><em>Whom<\/em> as<\/li>\n<li><em>What<\/em>, and<\/li>\n<li><em>With<\/em> what social effects? (2017, 156)<\/li>\n<\/ul>\n<p>Using the example of Muslims and alcohol consumption, this results in the following:<\/p>\n<ul>\n<li>Muslims who drink alcohol are identified by the team behind Chat-GPT as not being \u201cpracticing Muslims\u201d, resulting in their exclusion from the dominant model of \u201cproper\u201d Islam, and the potential perpetuation of stereotypes, disapproval, and reprobation from within and outside the \u201cMuslim community\u201d.<\/li>\n<\/ul>\n<p><!--more--><\/p>\n<p><strong>Implications <\/strong><\/p>\n<p>In the context of the study of religion, it has already been documented that <a href=\"https:\/\/hai.stanford.edu\/news\/rooting-out-anti-muslim-bias-popular-language-model-gpt-3\" onclick=\"javascript:urchinTracker ('\/outbound\/article\/hai.stanford.edu');\">the prompt that you bring to ChatGPT will influence the results it produces<\/a>, with potential implications that go far beyond mere curiosity (AI produced summaries of news stories, for example). <a href=\"https:\/\/religiondispatches.org\/via-jokes-chatgpt-chooses-which-religious-traditions-and-figures-deserve-respect-and-therefore-what-counts-as-religion\/\" onclick=\"javascript:urchinTracker ('\/outbound\/article\/religiondispatches.org');\">Others have also noted<\/a> that GPT-4 is quite willing to tell jokes about Krishna, Vishnu, Shiva, Buddha, and Zeus, but won\u2019t tell one about Jesus, Allah or Muhammad because it could \u201churt religious sensibilities\u201d. \u00a0Together with my own observations, this demonstrates that great care must be taken when utilizing generative AI to explore the complex and contested concept of \u201creligion\u201d.<\/p>\n<p>Generative AI \u2013 of which ChatGPT is but one example \u2013 is not neutral, is prone to inaccuracies, hallucinations, and problematic stereotyping, and does not show its working. Yes, it can potentially be a useful tool for exploring a topic, but just like any material we might find online or from other dubious source, we must ensure that we are critical in our engagement, thinking through what has been included or excluded, from where, and why. With generative AI seeping more and more into our daily lives, the implications of these issues stretch far beyond academic study and may have profound social consequences that we may not even have begun to imagine. These are the times in which we live. And, to paraphrase Gandalf in Tolkien\u2019s <em>The Lord of the Rings<\/em>, what we must do now is decide what to do with the times that are given us. Perhaps I should ask GPT-4?<\/p>\n<p>&nbsp;<\/p>\n<p>References:<\/p>\n<p>Jong, J. (2015), \u2018On (Not) Defining (Non)Religion\u2019, <em>Science, Religion and Culture<\/em> 2 (3): 15\u201324.<\/p>\n<p>&nbsp;<\/p>\n","protected":false},"excerpt":{"rendered":"<p>By Chris Cotter\u00a0 At the end of October 2024, I had the pleasure of speaking on a keynote panel session as part of RExChange 2024 on what AI means for knowledge in the religion and worldviews classroom. I was speaking alongside Professor Beth Singler \u2013 who did an excellent job of providing a sweeping overview [&hellip;]<\/p>\n","protected":false},"author":5,"featured_media":1852,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[],"class_list":["post-1851","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-uncategorized"],"_links":{"self":[{"href":"https:\/\/www.open.ac.uk\/blogs\/religious-studies\/index.php?rest_route=\/wp\/v2\/posts\/1851","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.open.ac.uk\/blogs\/religious-studies\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.open.ac.uk\/blogs\/religious-studies\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.open.ac.uk\/blogs\/religious-studies\/index.php?rest_route=\/wp\/v2\/users\/5"}],"replies":[{"embeddable":true,"href":"https:\/\/www.open.ac.uk\/blogs\/religious-studies\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=1851"}],"version-history":[{"count":1,"href":"https:\/\/www.open.ac.uk\/blogs\/religious-studies\/index.php?rest_route=\/wp\/v2\/posts\/1851\/revisions"}],"predecessor-version":[{"id":1853,"href":"https:\/\/www.open.ac.uk\/blogs\/religious-studies\/index.php?rest_route=\/wp\/v2\/posts\/1851\/revisions\/1853"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.open.ac.uk\/blogs\/religious-studies\/index.php?rest_route=\/wp\/v2\/media\/1852"}],"wp:attachment":[{"href":"https:\/\/www.open.ac.uk\/blogs\/religious-studies\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=1851"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.open.ac.uk\/blogs\/religious-studies\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=1851"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.open.ac.uk\/blogs\/religious-studies\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=1851"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}