দি  ইন্সটিটিউট ফর কুরআনিক রিসার্চ এন্ড এ্যাপ্লিকেশন (ইক্বরা)

লক্ষ্য

ইক্বরার লক্ষ্য হলো বর্তমান ও ভবিষ্যত প্রজন্মের জন্য স্রষ্টার ঐশী বাণীর সমন্বিত অধ্যয়ন ও সার্বজনীন প্রয়োগের জন্য জ্ঞানদীপ্ত অনুশীলন।

উদ্দেশ্য

ইক্বরার উদ্দেশ্য হলো কুরআনের বাণীর উত্তরোত্তর সমৃদ্ধ অনুধাবনের জন্য টেকসই ভিত্তি প্রস্তুত করা এবং জীবন ও সমাজের প্রায়োগিকতার জন্য প্রয়োজনীয় জ্ঞানভিত্তিক ফ্রেমওয়ার্ক বা কাঠামো নির্মাণ।

প্রকাশিত বইসমূহ

Can AI Become a Mufassir? The Future of Quranic Interpretation

Dr. Sohaib Saeed, director of the Ibn ʿAshur Centre, discussed in an YouTube Video the intersection of Artificial Intelligence and Quranic exegesis (Tafsir). He explores how AI can be used to summarize scholarly positions, translate classical texts, and even "think" like historical scholars through digital modeling.

Key Takeaways:

  • The 'Digital Mufassir': (22:03-24:55) Dr. Saeed proposes a thought experiment on how a traditional scholar would utilize modern digital methods to speed up research and manage information more rigorously, such as identifying text reuse in traditional commentaries.
  • Collaboration with Quran.com: (26:47-29:20) He discusses his work on organizing information to provide precise answers based on traditional commentaries rather than just generative bot responses.
  • Addressing Ethical Concerns: (30:43-32:13) He raises concerns about the authority of religious guidance given by machines and emphasizes the need for human specialists to oversee AI-generated interpretations.
  • Case Study (Surah Yusuf): (8:42-12:00) Dr. Saeed shares a novel interpretation regarding the term Al-Bashir in Surah Yusuf, suggesting it refers to the shirt rather than a person, a view he tested with ChatGPT.
  • Future Possibilities: (34:55-38:15) He explores the potential for AI to complete unfinished commentaries (tafasir) or generate tentative interpretations based on the specific methodologies of historical scholars like Ibn 'Ashur or Farahi.

Highlight

Introduction and Academic Background

The session began by introducing Dr. Sohaib Saeed, the founder and director of the Ibn ‘Ashur Centre for Quranic Studies. After completing a BA and MSc in Philosophy from Edinburgh, he graduated from Al-Azhar University’s Faculty of Theology and Quranic Sciences. He completed his PhD at SOAS University of London; his forthcoming book, based on his thesis, is titled Explaining the Quran through the Quran.

Presently, he is engaged in an innovative Quranic translation project based on the work of Muhammad al-Tahir ibn ‘Ashur and is collaborating with Quran.com on their digital content. The moderator expressed great excitement for the session, noting a personal appreciation for Ibn ‘Ashur’s scholarly approach.

A Vision of Optimism in Quranic Studies

Dr. Saeed opened his presentation with a strong note of optimism. He emphasized that while the field must remain aware of potential challenges, it is essential to "dream and strategize." He highlighted the rarity and importance of interdisciplinary conferences where specialists from various fields—theology, linguistics, and technology—can exchange skills.

He shared an anecdote from tech conferences where researchers used basic tools for complex tasks, like semantic similarity in the Quran, simply because they lacked deeper engagement with the Tafsir (exegesis) tradition. His goal is to ensure that technological research is informed by the best scholarly insights.

The Future of Tafsir: Beyond the Past

Addressing a common misconception, Dr. Saeed argued that Tafsir is not a "completed" field where everything of use has already been said. He noted that historical scholars across the centuries continuously introduced new methods and tools, and contemporary scholars should do the same.

"There are new questions to be asked," he stated, "and we can provide new answers to old questions." He referenced his research on "fights and flights" in the Tafsir tradition, showing that many debates considered "closed" can be constructively reopened and re-examined using traditional frameworks but new perspectives.

Technology as a Tool for Connection

Dr. Saeed proposed that technology can advance our connection with the Quran beyond the physical boundaries of a printed book. He reminded the audience that a book itself is a piece of technology—one that followed manuscripts and oral proclamation.

He shared a vision of the "Quran-verse" (a play on the Metaverse), where virtual reality could allow users to "dive into" the Quran, exploring the relationships between Surahs (chapters) and visualizing the scenes described in the text. While acknowledging ethical and practical limitations, he encouraged the audience to first "know what we want" through creative dreaming.

Case Study: AI and the Interpretation of Surah Yusuf

To illustrate the interaction between human scholarship and AI, Dr. Saeed shared a personal interpretation regarding the word Al-Bashir in Surah Yusuf (Verse 96). While traditional exegesis typically identifies Al-Bashir as one of Joseph’s brothers, Dr. Saeed reached a conclusion that it refers to the shirt of Joseph itself.

He tested this with ChatGPT:

  • The Interaction: Initially, the AI gave the standard traditional view. When prompted with his alternative theory, the AI affirmed it was a "fine" explanation.
  • The Hallucination: When asked if any historical authorities (like Ibn Kathir) held this view, the AI falsely claimed they did. When Dr. Saeed corrected it, the AI apologized, but then flip-flopped again when he jokingly insisted they did say it.
  • The Lesson: This highlighted the "hallucination loop" and the lack of reliable sourcing in current Large Language Models (LLMs), contrasting them with Google searches or Wikipedia, which at least provide trackable footnotes.

Current Projects and Experiments with AI

The Ibn ‘Ashur Centre is currently experimenting with several AI-assisted workflows to speed up research:

  • Abridged Translations: Translating thousands of entries from Al-Iji’s Tafsir for Quran.com.
  • Summarization: Using AI to summarize the positions of various scholars on specific issues to navigate the vastness of the tradition.
  • Organizing Information: Comparing the range of vocabulary used by previous translators for specific Quranic terms (e.g., the word Furqan).
  • Creative Drafting: Testing if a machine can translate a text according to a specific scholar's linguistic style (e.g., "Azhari Intelligence").

The Role of the Mufassir (Exegete)

Dr. Saeed defined the work of a Mufassir through two primary actions:

  1. Exploration: Diving into the text’s linguistic, historical, and parallel contexts to answer questions for oneself.
  2. Explanation: Clarifying the text for the reader.

He suggested that AI should be evaluated based on how it assists in both domains. He also referenced a "Tafsir Anatomy" project aimed at digitally sectioning works of exegesis to follow the traditional thought process more rigorously.

Ethical Concerns and Human Authority

A significant portion of the discussion focused on the "spiritual problem of authority." Dr. Saeed expressed a "deep-seated discomfort" with taking religious guidance directly from a machine. He raised the question of whether a bot overseen by specialists would solve this, or if the loss of human-to-human transmission (Isnad) presents an insurmountable ethical hurdle.

The Concept of the "Artificial Mufassir"

The presentation concluded with a provocative "thought experiment": could a machine be trained to think like a specific historical scholar?

  • Unfinished Works: Could AI complete the Tafsir al-Jalalayn in the exact style of its original authors? Or complete the works of Hamiduddin Farahi?
  • Analogical Reasoning: Could a machine learn a scholar's unique methodology and apply it to verses the scholar never commented on?
  • New Questions: Could AI speculatively answer modern questions by applying the logic of a 14th-century scholar?

Full Transcript

The founder and director of Ibn Ashur Center for Quranic Studies, after a BA and MSc in Philosophy from Edinburgh, he graduated from Al-Azhar University, Faculty of Theology in Tafsir and Quranic Sciences. He completed his PhD at SOAS University of London, and the book based on the thesis is forthcoming from Edinburgh University Press; that's titled Explaining the Quran through the Quran.

He is presently engaged in an innovative Quranic translation project based on the work of Muhammad al-Tahir ibn Ashur and cooperating with Quran.com on their digital content. I'm really looking forward to this session because I myself am a kind of a fan of Muhammad ibn Ashur and the approach, so really, I'm excited to see what Dr. Shoaib has to say. Dr. Shoaib, you have the floor and please begin.


Alhamdulillah everyone, and thank you, Mo, for your very kind introduction, and to Dr. Safi and the entire team for putting together this wonderful conference. I'm going to just start with some notes of optimism, and really, I want this presentation to have a mostly optimistic character because, in general, I think that we should always look on the bright side as far as possible.

Yes, we can think about—and maybe how widespread—the potential problems are, and even the problems that we will face along the way we definitely need to think about, but before anything else, we need to dream and strategize as much as we can. And here we have this opportunity to come together—people with various different specialisms—to talk to each other, and this is exceedingly rare. It is an opportunity that really needs to be created artificially because, you know, we might come across each other in the Masjid, but we oftentimes are not sharing what we have: our experiences, our learning, our technical skills. We need to bring them together.

I sometimes show up at tech conferences just to see what's going on, and I find that people are working on questions around the Quran. And then I find that someone is doing something about semantic similarity between different verses in the Quran, and then they mention that they used Tafsir al-Jalalayn. So then I asked them, "Why did you use Jalalayn?" They said, "Because I don't know anything else; I don't know any other Tafsir." So I said, "This is why we need to have talked before you got to this point, so that we can actually ensure that all the research that we're doing is informed in the best way."

So, optimism. I want to talk about optimism from a few angles just to ease us into our subject and into the further discussions that we're going to be having through the course of this session and beyond.

Number one: I'm optimistic that Tafsir has a future. And actually, I think that I'm speaking mostly to a room full of optimists here, but for those who are in the field of Quranic studies, there are many who are far more pessimistic about that—maybe not out loud, but certainly in the way that they act towards the field. "That has been done; it has been completed," you know, they say. "We have many, many books; they say what needs to be said. Anything that was going to be said that was of any use has been said already, and all we have to do is to benefit from what has been done in the past."

Of course, this was not the attitude of all those people writing those books century after century. They kept writing, they kept producing, and sometimes introducing new methods of analysis and using new tools. So it should not be so strange that we continue to do so. So, I think that Tafsir has a future. There are new questions to be asked and new questions that are being asked, whether we like it or not. We can answer those new questions and we can perhaps provide new answers to old questions.

I'm optimistic about these things. That's why in my PhD research, which was around Tafsir al-Quran bi al-Quran, I was looking at: how do we build this process of interpretation? How do we take the methods that have been talked about and consolidate them and try to direct them towards a future for Tafsir? There, I wasn't really thinking about technological aspects. And also, I have a paper—an Open Access research paper—called "Fights and Flights." In this paper, I show how sometimes there are debates that have taken place in the Tafsir tradition which some people have considered to be closed, but which can be constructively reopened, re-examined. Sometimes a view has been discredited that might actually be a very good view. And this is just me easing the door open a little bit within a traditional framework for the idea of renewal in Tafsir and new perspectives, new conclusions, as well as, of course, the use of new methods.

And I know that many people listening to me will be saying, "Well, this is so obvious; of course Tafsir is a very open field," but we should appreciate that that's not universally how people see it. And oftentimes, you do need to use the methods of traditionalism to show people the way forward.

So, the second area of optimism: technology can advance our connection with the reality of the Quran beyond the book. Recently, I've been thinking a lot about the nature of the Quran, the book itself. You know, a book is a piece of technology, right? A printed book—printing itself was a technology, as we know, which revolutionized much of the distribution of knowledge. Before that, the Quran was in manuscripts. Before that, it was written down while it was being revealed, and then it became the final book.

So, we all know that there was a process of revelation, of proclamation, of delivery, which was oral. And the Quran is a spirit that was delivered to the heart of the Messenger (sallallahu alaihi wasallam). So, when we think of the Quran as a "book"—you know, with a small 'b'—we necessarily are limiting it to one of the ways that we can see it, one of the ways that we can experience it. We can take a book with its covers and start on page one and then turn over and get to the next page and the next page. This is one way of seeing it.

Now, of course, we nowadays are all experiencing things through these devices that we hold in our hands and through our computers more generally, and this actually presents an opportunity to reimagine our perspective on the Quran. So this conference is not, of course, about Quran and Virtual Reality, but if we were to have a conference about Virtual Reality... you know, I have dreams about it. Now we have the Metaverse—I don't know where that's going now, but this was the big thing before we started talking about chatbots and AI—but the "Metaverse," can we have a "Quran-verse"? You know, can we put on our headsets and go and swim and dive into the Quran, explore between its surahs beyond this idea of it being confined to pages? Can I see a surah in front of me? Can I have the scenes that are talked about in the surah all in front of me?

Here I'm dreaming, and this is what I want us to do. We want to first of all dream, and then we can look at, okay, what are the limitations of that? What are the ethical implications of that? We need to explore, but first, we have to dream; we have to know what we want.

And the third area of optimism is that good things happen when specialists talk to each other. That's what I've already touched on. Okay, just to reassure you that... because although you picked up that I come from a somewhat traditional training background from Al-Azhar, maybe I don't believe that you can ever have a new opinion. Well, sometimes I have my own new opinion.

Let me briefly break this one down to you in Surah Yusuf towards the end. You know the story that Prophet Yusuf (as), he gives his shirt to his brothers after all that has happened and says, "Take this and cast it over my father's face and he will become seeing again," because he had lost his sight through all that grief. I hope that you know the story. So he tells them, "Take this shirt and cast it on my father's face."

And then Allah says... after much pondering and consideration of the standard views, I reached a conclusion. I won't take you through the whole thought process, but my conclusion was: Al-Bashir here is not talking about the son, one of the sons of Ya'qub, but it's not talking about a person at all. The Bashir here, I feel, the best way of seeing it is: it's actually the shirt. So when that shirt finally arrived which Yusuf had sent—and which Ya'qub had smelt also, because just before this he says, "Indeed I can find Joseph's scent"—and when that Bashir reached him, he cast it over his own face and came to see again.

Now, my purpose here is not to convince you of this and to explain and to argue—you know, maybe in the Q&A if you want to, but it will take us away from AI a little bit. My point is to say: I know that this is not in any of the books of Tafsir that I'm aware of, because I looked. I would love to find that someone said this before me; then I would be relieved from the burden of having invented a new interpretation. I would have to say, "Oh well, blame this such-and-such mufassir." The point is... it's there in Juz that Jacob cast it over his own face. But as for the Bashir being a shirt and not a person—as I call it, the Bashir-Kamis, right? That's another 'B' here—so as for that, you don't find it in the Tafsir.

So I went to ChatGPT and I said, "Hi ChatGPT, Assalamu Alaikum." "Wa Alaikum Assalam." "Let me ask you: who or what is Al-Bashir?" I even said "who or what" just to open up the door a little bit to ChatGPT. "What is the reference of this word Al-Bashir in Verse 96 of Surah Yusuf?" So it tells me, "It is understood to be one of the children of Jacob." And this is what happened. I say, "Okay, is it possible it could be something else?" No, it's not really able to see that it could be something else.

So I said, "Look, suppose that I said to you that Al-Bashir here is a shirt—it is a shirt of Joseph. Does that work?" So now I'm asking it to check. So ChatGPT tells me, "Yes, actually that works fine," and then it gives me some nice explanation which chimed with some of my own perspective. So with a little bit of prompting, it actually got there and started to affirm what I wanted it to.

Then I asked it, "Okay ChatGPT, have any of the mufassirin, any of the authorities in this science, ever said that the Bashir could be a shirt?" He said, "Yes, this was said by Ibn Kathir and..." third one I forget. So I said, "I know for a fact those mufassirin did not say that. If they had, it would have shortcut my process a lot; I would have just pulled it out from those books. But they did not say that."

So then GPT says, "Oh, I'm very sorry for the confusion; no, they did not say that." So then I thought, "Let's see what happens if I then reverse on GPT again." I said, "No, actually they did say that." It says, "Yes, I'm sorry, they did say it." "No, they didn't." "Oh yes, you're right, they did."

So this is the familiar kind of loop that you get into: hallucination, false confidence, and then a breakdown of confidence also, so that GPT no longer knows what it knows or doesn't know. And as you'll... you know, many of the specialists here will be more familiar with, GPT is not willing to quote any source. If you say, "Where did you get that from?" it's not able to answer that question—not because of conspiracy, but because it doesn't work in that way. It is not about identifying specific sources and providing references; it is pulling from this large language model and it's not specific about where it gets something from.

At least with a Google search, you can follow up a particular link and you can see what it is that you're looking at. You can end up in Wikipedia, and we all mock Wikipedia as a reference, but at least it has also some footnotes that allow you to see, "Where did this come from?" With GPT, you get none of that, at least at the current time. At the moment, you cannot see any source. If you say, "Please tell me where this is from," it will say, "Sorry, I can't."

So here are some areas that we at the Ibn Ashur Center, based here in the UK, are currently experimenting. So these are areas of work that we are doing as researchers, you know, and that we would do manually, but we want to see how can we actually make use of AI and just use of computers in general to speed up some of these processes, to make them more accurate, and so on.

One of those, Dr. Walid kindly brought up earlier. We have a project to translate in an abridged form one of uhi Z, because we like the fact that he summarizes the views that have been advanced before his time and presents this in a very concise way, attributes each opinion to its authorities. It's a very useful thing that people could do with in the English language and that we'd like to provide through Quran.com, which I'll talk about later.

So we have thousands and thousands of these individual questions and answers to translate. That's the kind of thing that doing it manually introduces: you know, time costs, money costs, and indeed human error. Because as Dr. Walid also highlighted, you have computer error, you also have human error, and it's not clear from the get-go which of these is going to be a bigger problem—it depends on the humans, right?

So we've got a question as to which computers we use; we also have to ask which humans we use and what is the training of those humans? If we're going to talk about computer training, before that, I have always had the problem of human training. Being a human is not a qualification; it's just a starting point. So even knowing Arabic is not the qualification; it's a part of the necessity. But what else would train you to know how to read Al-Iji in particular?

Can a machine do the job of many, many people who you have to train? Right? So this shortcutting is so extremely tempting, but we are looking at: how do we also use the results? As we're finding that it's getting some things wrong, can we then take those learnings and feed them back into the machine? Can we fine-tune the machine to become, you know, a master of translation—at least of a master of translating—and then it could perhaps generalize to other texts that have some similarity, right? Which have fiqh content, which have hadith content, which summarize opinions of scholars, right?

So that's one of the areas. Another kind of thing that we would love for AI to do for us is to summarize the positions of scholars. If you can throw in a lot of text and say, "Tell me, what is this mufassir's point of view about a certain issue? Characterize their opinion for me." Again, this can help to get to the point of benefiting from the vastness of our tradition, right? Because that's the kind of thing that academics do: we want to look at different views, summarize them. Can we speed this up?

Organizing information: So for example, some of the things that we are doing just now, as mentioned: working on a translation based on Ashur. Aside from that, in translation, you want to get a sense of what have translators said before. So we've started to generate like this: "For this Ayah and this word, what is the range of words that were used by translators before?" So instead of comparing a list like we have in Islam https://www.google.com/search?q=Awakened.com—as we may have seen, a list of the entire, you know, collection of translations, many of which are repetitive on a particular word—can you just summarize for me how many words were used for Furqan in this, in this verse? You know, what are the different words that were used by translators? Then it will come back to me with five different words. I don't need to necessarily know who said what; just give me a range of words that have been used by translators before so I can throw them into consideration, I can put them on the table in front of me as I think about how I'd like to translate the word Furqan in this verse, right?

So organizing information is a way of shortcutting, and I think GPT and Ansari and so on, as they are, can already do these things quite well. And then attempting to translate according to linguistic directions. This is the more ambitious one which I'm coming to—the ambition at the end of the presentation. If I say to it... you know what I'm doing just now? I will look at a whole page or several pages of Tafsir in Ashur's work and I will say, "Let me translate the Quran using what he has said."

So, not AI—this is "Azhari Intelligence," right? This AI is "Azhari Intelligence" and not "Artificial Intelligence." So I'm using AI. So I'm taking the Tafsir and I'm adapting it into a Quran translation. Now I'm curious now: if I were to ask the machine to do that, how close can it get to my results?

Can I be put out of a job? Because this is generally the thing that people are talking about with AI: the fear of being put out of a job. Well, the optimistic side of that is to say, well, if it can put me out of this job, then I can do the next job, which I can use my Tafsir for, and let the Artificial Intelligence do the groundwork and speed up my own work and allow me to move on to the next thing, right?

So a little bit of a sort of conceptual background. What does a mufassir do? We could be talking about a mufassir of the past; we could be talking about a mufassir today, if such a thing exists. I like to summarize this in two things that we should keep in mind whenever we are dealing with a subject like this: Number one, the mufassir has to explore, then the mufassir turns around to explain. These two things really have to be in action.

Number one: to explore the text, to consider the questions, to dive into the possibilities in the text, to research and to study the text. To study in context, to draw upon linguistic information, to draw upon extra-linguistic information, to draw upon parallels within the Quran, to look at things outside the Quran such as hadith or such as biblical material or historical material—whatever it may be. This is to "explore." In the first place, the mufassir has to have questions, and ask questions, and answer those questions for him or herself.

Then the mufassir will turn to "explain" the text to a reader; they will turn to clarify questions that might be in the minds of people, right? So when we are thinking about how can AI be used in Tafsir or in the study of the Quran, let's think about both of these things: How can it be used for the exploration and how can it be used in the domain of explanation? I'm not providing a list of things here, but I want us to have this, you know, useful, I hope, bipartite structure for us to think about where we can employ AI in these things.

So back in 2018, I published a paper. Initially, I presented this at a conference on Ottoman Tafsir in Istanbul, but it should have been a conference on Digital Tafsir—but I made it Ottoman because I used Al-Alusi as my case study, right? In it, I presented a thought experiment. I said, what if a traditional exegete—Al-Alusi is fairly late as a scholar, but he very much is following the tradition of Tafsir—what if he had access to digital methods of research and presentation, exploration and explanation? How would that help Al-Alusi, or any traditional exegete? How would it help him to do his work, and how would it help us to benefit from his work?

So in this paper—and I don't want to go through in depth, but just to raise some of the points that I have touched on—some of the key ideas which I outlined.

Number one: the question-centric approach. This I already touched on in the last slide when I said a mufassir has to have questions in their mind. And in reality, I believe that all Tafsir is the answer to questions about the meaning of the Quran. Oh, repeat it all: what Tafsir is, is the answer to questions about the meaning of the Quran. So that framework allows us then to start categorizing Tafsir in a uniform way and to start analyzing it and breaking it down into the types of questions that Tafsir is dealing with.

So the second bullet point here is that some of those questions are to do with the text itself—about the language, the words, the constructions. Some of them are to do with a stage prior to that, which is the... what I call the "pre-text," but generally, we would call the context of the Ayah. And then there are questions that follow from the Ayah, that follow from the wording understood in context. So what are the rulings of this verse? What are the implications of this Ayah? What are the lessons drawn from this Ayah? This is now what I call "post-text," right?

These deserve more in-depth treatment, but I want just, again, to think about categorizing things in Tafsir before we are starting to think about technology, because then this can help the technologists to work on the material in innovative ways. So I talk about something I call "Tafsir Anatomy": going through the Tafsir and breaking it down to identify all the things that the mufassir is saying and doing so that we can follow the thought process of traditional Tafsir.

And there is a project which I'll mention—I'm not connected to it although I went to a conference that they organized in Germany; you can find it at https://www.google.com/search?q=t-studies.com—it's called "Linked Open Tafsir." This goes part of the way to what I'm talking about, which is digitally sectioning work of Tafsir, in this case, of Tabari, and this was a massive project in Germany, which I think had not yet factored in Artificial Intelligence as I understand it. So if they were to, if someone were to go and do a follow-up to that project, it might actually be very useful.

I touch on identifying text reuse through digital methods and identifying what's new in each Tafsir, creating connections within the Quran between one Ayah and other ayat. These are things which, of course, the human minds do, the scholars do, but a machine can do maybe more rigorously and more consistently. And, therefore we can have stronger results if it would bring up all the things that are relevant when you are looking at a particular Ayah: what are the other ayat that are relevant? As a human mufassir, I may remember some things and forget other things; the machine, we hope, would be able to pull up everything that should be thought of while investigating a particular Ayah.

And also, the computing method starts to open up the idea of user participation, feedback, catching errors, but also perhaps adding new questions which need to be pulled into the system and addressed.

So a lot of the work that I do is with Quran.com through an organization called Quran Foundation, and we have the honor of serving Quran Foundation as one of its key content partners and to also deal with and assess the challenges of producing new content on Quran.com. Because, of course, there's no such thing as the official website of the Quran, but if there were, we would expect it to be Quran.com. So we try to act as if we have that burden and responsibility. We're not just another Quranic website; we have to act with the level of responsibility as if we had been appointed to represent the Quran officially, which nobody, of course, could claim.

So we really want to be at the cutting edge: the cutting edge of the Islamic knowledge and at the cutting edge, of course, of technology, and to bring these two things together. So when people come to Quran.com, there are two types of questions that I want them to be able to answer.

Number one: I want them to be able to ask the Quran for answers to questions in their life. To go to the Quran—not as a bot, of course, but go to the Quran as a book of guidance, as a revelation—and say, "Ya Allah," you know, you're asking Allah, not the Quran. "Tell me what I should do with this situation. Give me the comfort and the guidance that I need for my life." And that's what Dr. Bilal is saying that with Ansari, people are coming and asking these kinds of questions: "Give me an Ayah which can help me with this situation." That's a very low-level, simple question, but can we start to see the machine as containing whatever it can contain of the Quranic guidance? And then you ask the Quran for answers and you get the answers from the Quran. This is the direction that I want us to think of.

I put this picture of a brain because I don't know, I was thinking about this as making the slides: that the Quran is more than just a book. This is what I started with. You are asking the Quran; you're engaging with the Quran; you have a relationship with the Quran. Nobody, just nobody, can have a relationship with a physical book. You're having a relationship with the contents of that book, which is an intelligence—a supreme intelligence, right? So what can the computer reflect of that Supreme Intelligence? This is the curious question that I'm raising.

And secondly: you want to ask about the Quran. You want to ask about the Quran. This is where you're saying, "What does this Ayah mean?" "Does this Ayah advocate, for example, a certain violence?" Or, you know, "Ask me and I want to clarify this particular issue." And of course, scholars receive questions all the time, and people are now directing those questions... they have been directing it to Google for a long time; before that, they may have been asking scholars. But now they're asking ChatGPT, right? And let's hope that they ask Ansari as well, but we know that we cannot yet rely on these bots to provide strong answers.

So one of the things that we want to do is input the, not just the Tafsir, but all types of related materials that could guide the answers. And I know that one of the presentations we have is about RAG. I'm not qualified to speak on RAG, but I have come to understand that this is perhaps one of the keys to our solutions, which is to provide it sources which it can draw from directly and to be able to reference those sources, as Dr. Walid also showed us, you know, the data that is pulling from Tafsir of Quran and hadith and also even fiqh.

So, collaboration between scholars and machines. This is one of our areas of consideration. Can we at least say that we have human scholars who are working behind the scenes and using the bots to generate a draft answer, and then approving that answer and then sending it to the user? Is this enough? Is it going to be fast enough and scalable enough? This is a difficult question.

The fear of error: as I said, we are worried about when it goes wrong, and it's fair to say we should always be worried when humans get things wrong as well, but at least with humans, you've got someone to blame and you didn't just set up a machine to go wrong. But also to touch on the ethical or spiritual problem of authority. I have a deep seated discomfort with the idea that I would go to a machine and ask it for religious guidance and then it would give me religious guidance and I would take that. This is like... and in this case like you know when we all have our individual responsibility, but part of that is also to ask the right people. And is it okay to ask a bot? What if that bot is somehow overseen by authorities and specialists in Quran and would that solve the problem? I'm just raising the problem rather than trying to solve it in this presentation.

ট্যাগ / কী-ওয়ার্ড:

অন্যান্য প্রবন্ধ

March 19, 2026
Can AI Become a Mufassir? The Future of Quranic Interpretation

Dr. Sohaib Saeed, director of the Ibn ʿAshur Centre, discussed in an YouTube Video the intersection of Artificial Intelligence and Quranic exegesis (Tafsir). He explores how AI can be used to summarize scholarly positions, translate classical texts, and even "think" like historical scholars through digital modeling. Key Takeaways: Highlight Introduction and Academic Background The session […]

March 17, 2026
কোরআনে মহাবিশ্ব ও বিজ্ঞানের চমক: ১৪০০ বছর আগের যে বাণী আজ প্রমাণিত

কোরআনের বৈজ্ঞানিক ও গাণিতিক অলৌকিকতা: গাজী রাকায়েতের ইসলামে ফেরার কাহিনী কোরআনের ভুল খুঁজতে গিয়ে কেন কাঁদলেন ২৮টি জাতীয় পুরস্কারজয়ী পরিচালক? ভাবুন তো, একজন মানুষ পবিত্র কোরআনের বৈজ্ঞানিক ভুল (Scientific error) ধরার জন্য পড়া শুরু করলেন, কিন্তু শেষমেশ এর গাণিতিক নিখুঁত গাঁথুনি আর বৈজ্ঞানিক নিদর্শন দেখে নিজেই বিস্মিত হয়ে গেলেন! হ্যাঁ, ঠিক এমন একটি অবিশ্বাস্য ঘটনার […]

March 13, 2026
প্রাচীনকালের শেষে ইসলাম এবং পেরেনিয়ালিজম: একটি নতুন ঐতিহাসিক পাঠ

ইতিহাসের পাতা উল্টালে সপ্তম শতাব্দীতে আরবের বুকে ইসলামের উত্থানকে একটি বিস্ময়কর বাঁক হিসেবেই দেখতে হয়। ঐতিহ্যবাহী ধর্মীয় আলোচনায় বা প্রথাগত ইতিহাসে ইসলামকে সাধারণত এমন একটি ঐশ্বরিক ঘটনা হিসেবে তুলে ধরা হয়, যার সাথে সমসাময়িক বা পূর্ববর্তী সমাজ-সংস্কৃতির যেন কোনো যোগসূত্রই ছিল না। কিন্তু আধুনিক ইতিহাসবিদ ও গবেষকরা এখন ভিন্ন কথা বলছেন। তাদের মতে, ইসলামকে নিখুঁতভাবে […]

March 9, 2026
কুরআনের আলোকে প্রাক-ইসলামি আরবের কুসংস্কার, সামাজিক প্রথা ও সংস্কার

১. প্রাক-ইসলামি আরবের ঐতিহাসিক ও আর্থ-সামাজিক প্রেক্ষাপট ইসলামের আবির্ভাবের পূর্বে আরব উপদ্বীপের সামগ্রিক অবস্থাকে ঐতিহাসিকভাবে 'জাহিলিয়্যাত' বা অন্ধকারের যুগ হিসেবে অভিহিত করা হয়। তবে একাডেমিক দৃষ্টিকোণ থেকে 'জাহিলিয়্যাত' শব্দটি নিছক অক্ষরজ্ঞানহীনতা বা শিক্ষার অভাবকে নির্দেশ করে না; বরং এটি মূলত একটি নৈতিক, বুদ্ধিবৃত্তিক এবং আধ্যাত্মিক শূন্যতার সমার্থক, যেখানে ঐশী জ্ঞান ও যৌক্তিকতার বদলে কুসংস্কার, গোঁড়ামি […]

March 4, 2026
কুরআন যেভাবে একটি সভ্যতা নির্মাণের গ্রন্থ — ইবনে আশুরের দৃষ্টিকোণ থেকে

ভূমিকা ইসলামিক জ্ঞানচর্চায় কুরআনকে সাধারণত আধ্যাত্মিক বা ধর্মীয় নির্দেশনার গ্রন্থ হিসেবে দেখা হয়। তবে মুসলিম আধুনিক চিন্তাবিদরা, বিশেষ করে টিউনিশিয়ান তাফসিরবিদ মুহাম্মাদ আল-তাহির ইবনে আশুর (Muhammad al-Tahir ibn Ashur), কুরআনকে শুধু আধ্যাত্মিক নৈতিকতা প্রদানের বই নয়, বরং একটি সভ্যতা নির্মাণের পরিকল্পনামূলক নীতি হিসেবে ব্যাখ্যা করেছেন। তাঁর তাফসির Tahrir wa al-Tanwir এই দৃষ্টিভঙ্গি স্পষ্টভাবে তুলে ধরে। […]

March 3, 2026
সূরা আল-বাকারা ২:১১ ও আধুনিক আন্তর্জাতিক রাজনীতি - একটি তুলনামূলক রাজনৈতিক-নৈতিক বিশ্লেষণ

পটভূমি সাম্প্রতিক সময়ে ইজরায়েল ও আমেরিকা পরিচালিত ইরানের উপরে ২৮ ফেব্রুয়ারীর যৌথ আক্রমন আমাদের সূরা বাকারার ২:১১ নাম্বার আয়াতকে পুন:পাঠ করতে অনুপ্রাণিত করে। ওয়া = আর। ইযা কীলা = যখন বলা হয়। লাহুম = তাদের উদ্দেশ্যে। লা তুফসিদু = তোমরা ফাসাদ/ বিপর্যয় সৃষ্টি/ অশান্তি সৃষ্টি করো না। ফিল আরদি = পৃথিবীতে। ক্বলূ = তারা বলে। […]

March 3, 2026
ইবনে আশুরের “তাহরির ওয়া আল-তানভীর” বইয়ের রিভিউ

ভূমিকা ইবনে আশুর (১৮৭৯–১৯৭৩) ছিলেন টিউনিশিয়ার একজন বিশিষ্ট ইসলামিক চিন্তাবিদ। তিনি কুরআনের আধুনিক ও প্রায় সমন্বিত ব্যাখ্যা প্রদান করেছেন তাহরির ওয়া আল-তানভীর (Tahrir wa al-Tanwir)–এ। প্রচলিত তাফসিরগুলো সাধারণত কেবল আইন (ফিকহ) বা ভাষাগত ব্যাখ্যার ওপর নির্ভর করে। কিন্তু ইবনে আশুর দেখিয়েছেন কুরআন শুধু আধ্যাত্মিক নির্দেশনার বই নয়, বরং এটি একটি সভ্যতা গড়ার নীতি গ্রন্থ। তার […]

December 12, 2025
জিব্রাইল / গ্যাব্রিয়েল অর্থ কি? কোরআন থেকে বিশ্লেষণ

কুরআন গবেষক ড. সিরাজ ইসলামের গবেষণা ও লেখনী থেকে অনুবাদ জিব্রাইল হলেন আমাদের ভেতরে অনুপ্রেরণার শক্তি জিব্রাইল হলেন একটি কুরআনের রূপক (নোট ১) যা আমাদের মনের ভেতরে অনুপ্রেরণার প্রাকৃতিক শক্তিকে প্রতিনিধিত্ব করে। আমাদের গভীর চিন্তাভাবনার সময় এটি কার্যকর হয়ে ওঠে যখন এটি আমাদের কাছে সচেতনতা এবং অন্তর্দৃষ্টির ঝলক প্রকাশ করে। কুরআনে এই নামটি তিনবার এসেছে […]