The Time-Bending Revolution: Artificial Intelligence in Academia
Written by Berkay EKİNCİ, Molecular Biology and Genetics, 4th Year Undergraduate Student, İzmir Institute of Technology
In this article, I will share my experiences as an undergraduate student about the use and importance of artificial intelligence (AI) in academic writing.
About the Author: Berkay Ekinci
I am currently a fourth-year undergraduate student in the Molecular Biology and Genetics department at Izmir Institute of Technology (IZTECH). I have also been a member of the Doğanlar&Frary Plant Molecular Genetics Laboratory for three years.
My Academic Writing Background
My first article on gene-editing technologies was published in the science journal GENOMLINE during my first year. In my second year, co-authoring “Molecular Genetics in Agriculture” with Prof. Dr. Sami Doğanlar and Prof. Dr. Anne Frary for KatılımFinans journal solidified my interest in academic writing. In my third year, driven by my passion for plant science, I wrote a review article titled “Structural Defense of Plants and Pathogenesis”, which was published in the IZTECH GCRIS database. I drew all the figures myself and cited 74 references; this process took 11 months in total.
In my fourth year, I wrote a nationally funded TÜBİTAK 2209-A undergraduate project on plant circadian rhythms, which would enable collaboration between my university’s Plant Molecular Genetics Laboratory and the Chronobiology Laboratory under the supervision of Dr. Altuğ Didikoğlu. Because I wrote an 18-page draft in just three weeks, Prof. Sami Doğanlar, head of the Plant Molecular Genetics Laboratory, asked me if I could divide the project into two parts: a 2209-A undergraduate project, and a TÜBİTAK 1002 academic project. So, I created a 10-page 2209-A project with nine references, and then I focused on writing the 1002 project. During the intense five-month writing period from October to March, the feedback from neuroscientist Dr. Altuğ Didikoğlu taught me the academic writing process and the discipline of scientific thought. By using Google Scholar, I eventually completed a 20-page 1002 project with 48 references.
Challenges in Academic Writing
When drafting the introduction section of a project or review article, you are expected to explain your topic of study in detail. To do this, you must scan relevant sources in the literature and cite appropriate articles. This process both solidifies your mastery of the subject and increases the credibility of your writing. This challenging phase, including both research and training stages, requires time and patience to find suitable articles. If you do not cite the right sources, the likelihood of your work being accepted by the scientific community is reduced.
Popular article platforms like “Google Scholar” rely on keyword-based algorithms and only recommend publications that have the word or words you type into the search engine. This causes the suggestion of an extremely broad pool of articles, often causing them to diverge significantly from your research topic. To increase the recommendations matching the researcher’s expectations, a time filter is commonly used in Google Scholar. However, that unfortunately prevents the suggestion of high-quality articles written in the past.
Due to human and time constraints, I often found myself forced to use the “since 2021” filter in Google Scholar. Even with this filter applied, it recommended over 1600 publications, most of which were irrelevant to my research topic. During my 1002 writing process, only 48 out of thousands of recommended articles were suitable for use as references. After 5 months, I thought I had completed the draft until I had a major opportunity that would become a game-changer for the project’s fate.
How Artificial Intelligence (AI) Entered My Life
In March, IZTECH Library Director, Gültekin Gürdal, noticed my efforts in academic writing and offered me a special opportunity: Access to the AI-powered research platform “SciSpace”. Unlike keyword-based tools, these platforms classify and present publications based on their content. All I needed to do was to train the AI system by writing a detailed paragraph describing my project and the information I needed. Based on the information I provided, this innovative approach analyzed a vast academic archive dating back to 1960 and delivered the relevant articles in just 3 minutes. Even more astonishingly, while I struggled to find sources on plant circadian rhythms via Google Scholar’s algorithm, the AI-powered platform directly found the articles about the circadian rhythm of Phaseolus vulgaris (common bean) which is the species used in my project.
All 40 articles recommended by the AI algorithm were directly relevant to my research topic. Although I spent 5 months on Google Scholar for a literature review, to realize that I had overlooked so many publications critical to my project was both astonishing and spine-tingling. Consequently, I decided to rewrite every paragraph of the entire project. Within a single month leading up to April, I reviewed all suggested articles. The platform’s ability to summarize papers through AI-generated audio blogs (podcasts) was the biggest accelerator for my work. Increasing my references from 48 to 78 in just one month demonstrated how revolutionary AI truly is for academic writing. As the project’s scientific scope strengthened, we decided to submit the draft to TÜBİTAK’s 1001 program instead of TÜBİTAK 1002. In other words, integrating AI into my work multiplied the financial value of my draft proposal by 24 times. Moreover, it enabled me, as an undergraduate student, to write a high-level national academic project.
Why AI is Essential in Science?
Time is the most precious and irreversible continuity we have. AI-powered research systems not only scan vast academic archives that would take us centuries to read but also recommend relevant articles in just 3 minutes by bending time and making the impossible possible. By accessing even subscription-based papers, these platforms scan all research since the 1960s, guide researchers, and thus play a critical role in preventing redundancy in academic writing. These content-based algorithms overcome the limitations of keyword searches, ensuring publications crucial to our work are no longer overlooked. Moreover, they enable researchers to use their valuable time with maximum efficiency by summarizing recommended articles through audio and video formats.
In academia, AI is a game-changing power that redefines the rules. To achieve contemporary community standards, follow current global scientific advancements, avoid repetition, and support national scientific development, widespread access to AI tools in academia is strategically vital.
Note: The visuals in this text were created using an AI tool.