Is Generative Artificial Intelligence (AI) completely Innocent?

Prof. Dr. Ali Gunes

Ibn Haldun University

Comparative Literature Department

Email: ali.gunes@ihu.edu.tr

 

 

Last weekend, my daughter visited me. She is a lecturer who teaches English and is also a doctoral (PhD) candidate. She has just started her courses. I watched her struggle with her assignments and the articles she had to submit to her professors. Suddenly, I said to her: “Why don't you use AI platforms to write your articles quickly rather than boiling your mind to get ideas from many articles lying in front of you? Everybody is using them, so why do you waste your time and drain your energy?”

She raised her head and stared straight into my eyes at that moment. “Father! Father!” she said. “Is it you who tell me such things?” she continued. 

I insisted on saying: “What is wrong with it? Everybody is using it - professors, undergraduate students, graduate students, writers, journalists, content writers and editors, advertisers, and video creators. Simply everyone. Why is it wrong when you use?”  

Again, she gave me a stern look. Then you can imagine: a heated argument between a young doctoral student and a veteran professor started.

First, we both agreed that AI is part of education nowadays, and we both often use it. It is useful for performing complex reading and writing tasks within a short time, analysing literary texts and statistical data, interpreting health reports, creating PowerPoint presentations and videos for educational purposes, preparing language exam questions tailored to students' levels, and writing novels, short stories, and even poems. AI is used in journalism to create stories and edit news content; advertisers use AI to write advertising content for business purposes to convince customers to buy their products. The list is endless…

Since we are in education, since we both teach, my daughter and I have also agreed: Why should we deprive ourselves of this opportunity used in education by everyone? We have concurred that it is useful and necessary. 

Later, our argument took a different turn, and we asked ourselves a question: Is AI completely innocent, even though it is widely used in almost every field of life? Our answer was “no.”

 

First, the frequent use of AI and disproportionate, excessive dependence on generative AI production visibly atrophies critical thinking and creativity. 

 

But how?

 

Since I am not an expert on generative AI technology, their design, or knowledge production systems, I asked ChatGPT how it produces information or knowledge when asked questions. I avoided questions about its technical aspects because I am not technically trained; to be honest, I am absolutely illiterate when it comes to technology. 

Let me share information ChatGPT provided me with:

a) Learning: During training, artificial intelligence learns the structure of language, the relationships between pieces of information and patterns from millions of books, articles and other texts.

b) Understanding the question: When you ask a question, it strives to work out what you mean by analysing the words and the relationships between them.

c) Generating an answer: Using what they have learnt, they guess the most appropriate words for the answer in order and form sentences.

4. Presenting the information: It presents the answer it has generated to you clearly and understandably. Where necessary, it can also search for up-to-date information online, provided it has internet access.

ChatGPT clarified its point with a simple example

You ask: “Why is the Sun hot?”

The artificial intelligence uses the scientific knowledge it has learned to generate the following answer: “Nuclear fusion taking place at the Sun’s core releases a huge amount of energy. That is why the Sun is very hot.” 

The most important point: Artificial intelligence does not generate answers by thinking like a human, but by using the patterns it has learned and mathematical calculations. For this reason, whilst it can provide very accurate answers, it may occasionally produce incorrect information. In short, artificial intelligence is a system that learns from a vast number of examples and uses what it has learned to generate new answers to your questions.

As seen above, human-created information and knowledge are fundamental to this process because AI analyses millions of existing data points to generate meanings and answers in response to questions. AI generation, therefore, remains dependent on human-produced knowledge and data. 

How can AI atrophy critical thinking and creativity?

Relying too heavily on generative AI could gradually reduce researchers’ motivation to engage deeply with academic literature and develop original ideas that contribute to their fields. When researchers come to rely entirely on AI-generated summaries and answers rather than reading, interpreting, reinterpreting, comparing, contrasting, and reflecting on scholarly work themselves, they risk becoming passive consumers of information and data rather than active contributors to knowledge in a particular field of study.

Reading, as agreed worldwide, is essential to intellectual growth, critical thinking, creativity, and meaningful research. It obviously introduces researchers to various arguments, theories, research methods, and evidence, helping them see issues from multiple perspectives. It also allows them to recognise research gaps in existing research, question established ideas and identify problems that still need exploration in a certain academic field. By engaging closely with the literature and reading, researchers strengthen their ability to think critically and creatively, evaluate arguments, develop their own interpretations and finally generate new ideas.

Critical and creative thinking do not develop simply by having quick access to ready-made answers. Instead, they gradually grow through the strenuous efforts of reading extensively, questioning what others have written and discussed, comparing and contrasting different viewpoints, connecting ideas, establishing relationships among various concepts, and building arguments based on evidence. Generative AI can certainly help with these tasks, but it should support researchers’ thinking rather than replace it. When researchers allow AI to do too much of this intellectual work for them, they risk weakening the very skills needed to produce original research and advance knowledge in their disciplines. My daughter said to me: “When I have AI write something for me, I feel like a robot and do nothing. I feel embarrassed that I am an ineffective element of the process.” This is not a good condition for a researcher. 

A second possible drawback of relying too heavily on generative AI is that researchers may become intellectually and socially isolated, especially if they work alone and turn to AI as their main source of discussion and feedback. Yes, AI can provide quick answers, useful information and suggestions. Yet, it cannot fully replace the benefits of interacting with colleagues, participating in academic communities, or engaging with real-world situations and experience. Talking with colleagues at universities and engaging in friendly or academic arguments with other researchers in the workplace exposes people to different interpretations, perspectives, criticisms, and practical insights that may not arise in conversations with AI in an isolated room. Conferences, seminars, group projects, peer discussions, and even informal conversations in coffee rooms give researchers opportunities to question their assumptions, improve their arguments, and notice aspects of a problem they might otherwise miss. This matters greatly because knowledge does not develop through individual thought alone but through clashes of ideas, for or against, through discussion, disagreement, collaboration, and the exchange of ideas. When researchers start substituting these interactions with AI, they may become less open to different viewpoints and less aware of the social, cultural, and institutional factors and interactions that influence how knowledge is produced. As a result, their research may become too narrowly focused, less connected to real-world contexts, and less likely to produce original, diverse, and well-supported conclusions.

Finally, relying too heavily on generative AI while working in isolation may gradually weaken researchers’ confidence in their own thinking and their sense of connection to the academic community as they loosen their ties to real-world situations and experiences. Researchers may build confidence not just by gaining knowledge on their own, but also by sharing ideas, responding to criticism and disagreements, explaining their views, and improving their arguments through discussions with other researchers. Regular interaction with colleagues, students, mentors, and other university staff may help researchers question their assumptions, understand the strengths and weaknesses of their work, and see how their research fits into the wider academic field. These relationships may also help them feel that they belong to a community where people share ideas, responsibilities, and common goals. In contrast, researchers who mainly rely on AI for feedback may get quick answers and useful suggestions without having to explain or defend their ideas, deal with differing opinions, or engage in meaningful discussions with others. Over time, this could make them less confident in expressing their own views, taking part in academic debates, or responding to criticism and opposing views. Being a researcher, therefore, is not just about producing information or completing tasks efficiently; it also involves working with others, exchanging ideas, learning from disagreements and differences, and developing professional judgement through shared experiences. Everyone agrees that generative AI is a useful tool for research and communication, yet it cannot replace the human relationships and interactions that help researchers build confidence, gain recognition, and feel part of the vibrant academic and researchers’ community.

In conclusion, my daughter and I agreed that generative AI is not entirely good or bad. Its value in education and research depends mainly on how it is used, what limits or emphases are placed on it, and whether it is used responsibly. These tools can save time, make information easier to access, and help researchers with difficult research and writing tasks. However, relying on them too much can weaken some of the skills and practices that are essential to academic work. When researchers rely on AI-generated content instead of reading carefully, thinking independently, developing their own ideas, and discussing their work with others, they risk becoming passive consumers of information rather than active contributors to knowledge. Their research may also become less original, less connected to real-world contexts, and less carefully developed. Overreliance on AI may also undermine researchers’ confidence in their own abilities and weaken their connection to the academic community, since both confidence and connection can develop through discussion, criticism, collaboration, and shared learning. The answer, therefore, is not to avoid generative AI altogether, but to use it thoughtfully and moderately. AI should help researchers do their work more effectively, not replace their own thinking, judgement, or interactions with other people. Universities and researchers need to develop a good understanding of how AI works and its limitations, while continuing to value critical reading, ethical responsibility, careful research methods, and meaningful academic discussions. In this way, generative AI can make a useful contribution to education and research without undermining the creativity, independence, and collaboration that are essential to producing knowledge.