Recently, I posed my first ever question to ChatGPT. After hearing so much about it all over the media and campus, curiosity got the better of me. To measure the AI’s abilities and intellectual prowess, I decided to torture ChatGPT with the same prompt as the final essay of a history class I took last semester: “Making of the Muslim Middle East”, on the decline paradigm of pre-modern Middle East.

I typed: Is there a decline in the Muslim world?

The essay prompt in question was about the once standard narrative of the history of the Muslim world. This curiosity stemmed from the argument that following the destruction of Baghdad in 1258 by the Mongols, the Islamic civilization slipped into a slow decline, characterizing later centuries as a period of political and economic decline, and intellectual stagnation.

ChatGPT's response, however, was related to the challenges faced by the modern Middle Eastern region, commenting on the "political instability and conflict in many Muslim-majority countries; widespread poverty and inequality; lack of access to education and job opportunities."

I changed my question: Is there a decline in the Muslim world after the destruction of Baghdad?

This time, with a more specific question, ChatGPT's response was: "Yes, the fall of the Abbasid caliphate and the destruction of Baghdad in 1258 by the Mongol invaders marked a significant decline in the power and influence of the Muslim world. This event had far-reaching consequences and resulted in the fragmentation of the Muslim world into smaller states and empires."

If my final essay were to argue similarly, the professor would have been very disappointed as he provided significant evidence from the works of recent scholars to challenge this decline paradigm, which views the Muslim world between the 15th and 18th-Centuries as a period of intellectual stagnation. However, in recent decades, dissatisfaction with such a narrative has grown as the decline paradigm fails to account for the historical reality of the two reform movements of the 17th-Century and 18th-Century.

I typed: What do you think about the intellectual reformation during the 17th-Century in the Muslim world in light of the decline narrative?

After the usual claim of not having any personal opinions or thoughts due to it being an AI language model, ChatGPT continued, saying, "This movement, known as the Mu'tazilite revival, was a response to the decline of the Muslim world." According to ChatGPT, the intellectual reform movement was a result of the Muslim elites realizing that they are heading towards a decline, further supporting the decline paradigm instead of proving it wrong or giving a more nuanced argument.

According to what I have learned, newer understandings suggest that the reform movements responded to emerging internal social issues in the rapidly changing Muslim world rather than external threats of European countries. 18th-Century thinkers were aware of the infringements on the Muslim world. However, they did not perceive the situation as an urgent threat to the prosperity or existence of the Muslim community, and Europe was absent from the thought of all the major 18th-Century thinkers. Therefore, the reform movements show the flexibility of the Muslim world to adapt to social changes, instead of reflecting on the decline of the Muslim world and the urgency to learn from Europe. This flexibility, at least, proves the narrative of intellectual stagnation to be wrong.

I asked ChatGPT: If there are intellectual reform movements, how can you call that a decline?

ChatGPT responded, "You raise a good point. The term 'decline' often refers to a decline in power, influence, or prosperity and can be misleading when used to describe complex cultural and intellectual developments. In this sense, the term 'decline' may not accurately capture the full complexity of the situation. It is more appropriate to view the 17th-Century as a period of transformation and renewal, in which the Muslim world was able to overcome some of its challenges and make important contributions to global knowledge and culture."

Despite being somewhat repetitive in its expression, ChatGPT now agreed with what the readings my professor assigned suggest; instead of a decline, this period should be viewed as transformative. However, the response seems to suggest that "decline", though not accurate in describing intellectual developments of the Muslim world, can still be used to describe its level of prosperity, which contradicts what recent scholars have theorized. Despite the weakening of central power, the autonomous local powers with vibrant and revived economies emerged in provinces such as Egypt, Iraq, Palestine, Syria and major Arab cities of the Ottoman Empire, such as Aleppo, Cairo, and Damascus, grew substantially, both economically and demographically.

I followed up, typing: Despite decentralization during the 18th-Century, the autonomous local governments still managed to develop a stronger economy with continuing urbanization. How could you call that a decline when the economy kept developing?

ChatGPT responded, "You raise a valid point… the economy continued to develop and urbanization proceeded. This suggests that there were other factors at play that contributed to the vitality and resilience of the region, and that the concept of decline may not fully capture the reality of the situation." At this point, ChatGPT had changed its "opinion," agreeing that, at least intellectually and economically, “decline” is not a good summary of the Muslim world after the destruction of Baghdad by the Mongols in 1258.

Now, ChatGPT had been "convinced." And as I typed my previous question again, asking, "Is there a decline of the Muslim world after the destruction of Baghdad," here is ChatGPT's response:

“The destruction of Baghdad by the Mongols in 1258 was a significant event in Islamic history, but it does not mark the decline of the Muslim world as a whole… While there may have been a decline in military power, this was accompanied by other developments such as continued economic growth and urbanization, as well as an intellectual reformation characterized by a new emphasis on reason and scientific inquiry. Scholars and thinkers continued to make important contributions to fields such as science, medicine, mathematics, and literature, and Islamic societies remained influential and powerful in many parts of the world… Therefore, while the destruction of Baghdad was a significant event in Islamic history, it did not mark the decline of the Muslim world. The Islamic civilization and its cultural and intellectual achievements continued to develop.”

Similarly, when using ChatGPT's previous logic, asking if the intellectual reform movements demonstrate decline, ChatGPT responded, "Revival aimed to revitalize intellectual and cultural life, and to address the challenges that the Muslim world faced through a new emphasis on reason and scientific inquiry… it is important to recognize that reform movements can also be motivated by a desire for change, innovation, or progress, regardless of whether there is a perceived decline or not… In conclusion, reform movements can emerge for a variety of reasons, and do not necessarily indicate a decline."

The dramatic shift of ChatGPT's stance on the decline paradigm indicates that its response depends on how the questions are formulated. It searches for the existing explanation and data online. In this example, traditionally, the decline narrative dominates the historiography of the Middle East, so ChatGPT answered with the same narrative of decline. However, as more information regarding the intellectual reform movements and continued development was given, ChatGPT seemed to start formulating its response based on other sources, providing dramatically different answers. The quality of ChatGPT's responses depend highly on how the individuals curate their engagement with it. Vague questions lead to vague answers while high-quality questions with specific details often lead to higher-quality answers.

That is because, ChatGPT, as a generative AI, does not learn about the world; instead, it learns about the available texts and how people use words in relation to other words. When formulating its answers, ChatGPT, based on the texts it has read, decides to output words that make the most sense given the words in the prompt. ChatGPT is a double-edged sword, the performance of which depends highly on the ways in which one formulates their prompts.