A symposium titled “Generative Artificial Intelligence in Scientific Studies and Education” was organized by the Dean's Office of the Faculty of Health Sciences at Üsküdar University, coordinated by the Head of the Department of Artificial Intelligence Engineering. At the symposium, the development process of artificial intelligence and its contributions to academic careers were thoroughly discussed. During the program, Prof. Dr. Arif Aktuğ Ertekin, Dean of the Faculty of Health Sciences, drew attention to the widespread use of artificial intelligence and emphasized ethical and conscious use. Dr. Öğr. Üyesi Nuri Bingöl, Head of the Department of Artificial Intelligence Engineering, explained the working mechanism and historical development of generative artificial intelligence, addressing risks such as hallucinations, data bias, and security. Applications such as NotebookLM, SciSpace, Turnitin, Google AI Studio, Canva, Consensus, and Jenni AI were also introduced during the program.
Academicians and students, including Prof. Dr. Arif Aktuğ Ertekin, Dean of the Faculty of Health Sciences, Dr. Öğr. Nuri Bingöl, Deputy Dean and Head of the Department of Artificial Intelligence Engineering, and Master's students Kerem Tekin and Furkan Karataş, attended the event held at the İbni Sina Auditorium of Üsküdar University NP Health Campus.
Prof. Dr. Arif Aktuğ Ertekin: “It’s not right to say don't use it…”
The opening speech of the event was delivered by Prof. Dr. Arif Aktuğ Ertekin, Dean of the Faculty of Health Sciences. Ertekin drew attention to the development of artificial intelligence and its presence in academic life, stating; “There have been significant changes in the fields of science and daily life. One of these was the advent of computers in our lives during our time. Another is the use of artificial intelligence today. It seems it will be widely used and will develop greatly. It is impossible to foresee its limits; everyone has started using it. That’s why we are all probably using it, but we organized this training to provide at least some basic knowledge and awareness. With the more efficient, faster, and smarter use of artificial intelligence, tasks that take a long time for us will be completed in a much shorter period. In thesis studies and articles related to artificial intelligence, the use of AI is now being questioned. When you look at the plagiarism rate, it’s 5 percent; when you look at artificial intelligence support, it’s 95 percent. There is such a big difference. It is not right to say ‘do not use it’; rather, its rational, proper, and ethical use should be considered. This is one of the main objectives of this meeting.”
Following the opening speech, Dr. Öğr. Nuri Bingöl, Head of the Department of Artificial Intelligence Engineering, delivered an informative presentation.
Dr. Öğr. Üyesi Nuri Bingöl: “A technology that increases our productivity”
Bingöl emphasized that artificial intelligence extraordinarily increases production speed; “Generative AI, or productive artificial intelligence, is a technology that increases our productivity. It can present a task that a person might do in 3 to 6 months in minutes, sometimes hours. It provides tremendous speed and can be used in research, scientific studies, and articles. However, there is a certain limit. If the use of artificial intelligence rises to 95 percent, problems will arise. It can be used as an assistant and for help with literature reviews. YÖK (Council of Higher Education) also permits this. However, publishing a text written by artificial intelligence as if it were your own constitutes plagiarism.” he stated.
“Artificial intelligence is not a new concept”
Bingöl stressed that, contrary to popular belief, artificial intelligence is not a new concept; “Artificial intelligence has been progressing very rapidly, especially since 2022. However, artificial intelligence is not a new concept. It emerged in the 1950s. Alan Turing raised the question, ‘Can machines think?’ and is considered the father of artificial intelligence. In 1956, the term ‘Artificial Intelligence’ was first used at a conference held in Dartmouth, spearheaded by MIT. In 1957, Lisp emerged as an AI language, while today languages like Python, R, and Julia are used. Some of the algorithms, however, date back to the 1700s and 1800s.” he explained.
Training, fine-tuning, and production…
Bingöl explained the working mechanism of artificial intelligence in three fundamental stages; “There are three fundamental stages here: training, fine-tuning, and production. First, the model is trained, and the majority of the data is used in this stage. Since significant machine power is required, models are usually offered as ready-made models trained by organizations like OpenAI and Google. Then, in the ‘Fine-Tuning’ stage, the model focuses on specific tasks. Human evaluation is performed using the ‘Reinforcement Learning with Human Feedback’ method, and the model is adjusted. The final stage is ‘Deploy,’ which means production and continuous improvement. Information provided through the RAG (Retrieval-Augmented Generation) method can be made more customized.” he said.
“One of the biggest problems is hallucination”
Bingöl stated that alongside the opportunities offered by artificial intelligence, it also carries serious risks; “One of the biggest problems is hallucination. The model can provide false information as if it were real, cite fabricated sources, or make historical errors. Therefore, it must be checked. Data bias is also a significant problem; the model is dependent on the data it was trained on. There is a limit to originality, and outputs can be repetitive. Due to the black box problem, the internal workings of the system are not fully known. There are also serious concerns regarding security, deepfake, data privacy, copyright, and intellectual property. For this reason, experts particularly emphasize that personal data should not be shared on systems like ChatGPT.” he explained.
Artificial intelligence applications introduced
Following the presentation, Bingöl, introduced the NotebookLM application to the participants and provided information on its areas of use.
The operation and role of SciSpace, AI Detector, and Turnitin applications, used for artificial intelligence detection, in academic studies were also shared in detail.
Master’s student Kerem Tekin introduced Google AI Studio and Canva applications, providing practical usage examples.
Master’s student Furkan Karataş also explained the Consensus application, used in the literature review process, and the Jenni AI platform, which contributes to the post-doctoral reference generation process, to the participants.
The program concluded with a group photo.