AGL40▲ 0 (0.00%)AIRLINK129.06▼ -0.47 (0.00%)BOP6.75▲ 0.07 (0.01%)CNERGY4.49▼ -0.14 (-0.03%)DCL8.55▼ -0.39 (-0.04%)DFML40.82▼ -0.87 (-0.02%)DGKC80.96▼ -2.81 (-0.03%)FCCL32.77▲ 0 (0.00%)FFBL74.43▼ -1.04 (-0.01%)FFL11.74▲ 0.27 (0.02%)HUBC109.58▼ -0.97 (-0.01%)HUMNL13.75▼ -0.81 (-0.06%)KEL5.31▼ -0.08 (-0.01%)KOSM7.72▼ -0.68 (-0.08%)MLCF38.6▼ -1.19 (-0.03%)NBP63.51▲ 3.22 (0.05%)OGDC194.69▼ -4.97 (-0.02%)PAEL25.71▼ -0.94 (-0.04%)PIBTL7.39▼ -0.27 (-0.04%)PPL155.45▼ -2.47 (-0.02%)PRL25.79▼ -0.94 (-0.04%)PTC17.5▼ -0.96 (-0.05%)SEARL78.65▼ -3.79 (-0.05%)TELE7.86▼ -0.45 (-0.05%)TOMCL33.73▼ -0.78 (-0.02%)TPLP8.4▼ -0.66 (-0.07%)TREET16.27▼ -1.2 (-0.07%)TRG58.22▼ -3.1 (-0.05%)UNITY27.49▲ 0.06 (0.00%)WTL1.39▲ 0.01 (0.01%)

ChatGPT’s influence on academic writing and research ethics

Share
Tweet
WhatsApp
Share on Linkedin
[tta_listen_btn]
Dr. Muhammad Fahim Khan

In today’s world, many artificial intelligence (AI) tools are available now that use AI technology to help automate and make any tedious or repetitive task easier for science and research. For researchers, some AI writing tools, such as ChatGPT, have expanded the choices available in this area. A ChatGPT performance medRxiv preprint that also names ChatGPT as an author said the OpenAI tool achieved a “high degree of consensus and insight in its explanations,” reinforcing large language models’ potential in research writing, education, and even decision-making. As the language tool obtains world attention, where does this leave researchers in academic paper writing? This article discusses ways ChatGPT can help researchers create academic papers, focusing mainly on the limitations and ethical dilemmas surrounding using such AI tools to write research papers.

In academic writing, AI tools like ChatGPT by OpenAI are growing as popular as ever among students and researchers. This tool is especially noteworthy for producing both convincing and human-quality text. This isn’t just because that makes it convenient for non-native English speakers at universities to use but also because it involves a responsibility to transmit the correct information, which requires careful guidance. It has many uses: creating clear structures for essays, providing powerful themes to go in the titles of research papers and Ph.D. theses, and producing abstracts that both plagiarism detectors and the results of AI studies have approved. Furthermore, it has good capabilities in natural language processing, can translate text between languages, and helps rework intractable passages for people with learning difficulties so that they become unblocked and trouble-free. Also, it provides helpful one-paragraph summaries of long articles and complex documents, just what to put into your literature review. It is not just a way for people who suffer from overwork and concentrate on the theft of results to soften their working rhythm by pulling an unpleasant task out of their hands, nor is it only an alternative to saving time with Trado’s official CAT tool. Researchers profit from ChatGPT’s ability to summarize long articles and documents, giving quick but comprehensive surveys essential for literature review. Apart from writing, it offers advice on experimental design modifications and can be adapted to work on particular problems–like language understanding. It is thus an invaluable tool in the academic writing process.

When performing research writing, the availability of ChatGPT technology is certainly a help. Nonetheless, researchers need to be aware of which limitations may exist. In the primary sense, rather than germinating original ideas, ChatGPT establishes text from training data patterns that do not include references or citations–thus posing a risk for plagiarism by those who use AI. At the same time, since ChatGPT operates as a statistical model without content understanding, occasionally logical but completely wrong responses may appear plausible. In addition, ChatGPT may be inaccurate when representing specific research fields; specialty writers could encounter potential errors or unexplained differences. Another concern is the possibility of bias or offensive material from ChatGPT, which will reflect the biases in its training data. Thus, researchers should be on their guard regarding these issues when using the tool for research writing.

As we can see, ChatGPT’s knowledge is confined to what was available before 2022. Therefore, users should scrutinize and edit their outputs closely to pass muster with scholars. Computer scientist Yejin Choi illustrates the current state of large language models, such as ChatGPT, as being “a mouth without a brain. Teaching these models to understand causal reasoning, common sense, or moral judgment is still challenging.

Given that academic writing is an essential element of research activities, it is unknown to anyone with an appreciation of such things what may happen when so important a paper is written by machine. While ChatGPT allows one to write quickly from prompts—a gain in the temporal economy and productivity–any moral implications engendered in using automation at laborious points of one’s pictures must be considered.

While some precautions are in place to prevent such unfortunate incidents, the fact that ChatGPT has been nurtured on present information means it is likely to draw negative information or prejudices already in these databases into what it writes. To maintain the truth and fairness of the text, it is up to researchers to remain alert for any potential prejudices and make the appropriate corrections necessary. In addition, there is the danger of unintentional plagiarism. AI can indeed quicken the composition process, but it also runs a risk — that reminding readers where certain words came from will be forgotten. Furthermore, researchers must ensure they appropriately credit all sources. In addition, we can’t ignore the fact that ChatGPT could generate offending or negative text. If the training data includes such language, it may be subconsciously reflected in what comes out of machinery. We researchers need to keep our work politically correct and inclusive. Afterward, while ChatGPT can help with certain aspects of academic writing and even make citations, it doesn’t guarantee the research is reliable or of high quality. This kind of AI-generated text requires scrupulous editing by academics before being published today in scientific venues of any standing.

In our writing research, weare only responsible for AI tools such as ChatGPT. Researchers can use these tools to strengthen their thinking power rather than replace it. Click: In Pedro Domingos’s “The Master Algorithm,” he writes that human intuition plus data is like a rider’s relationship with a horse- that’s crucially valuable. It means softening AI’s hard edge with human insight rather than abolishing it.

—The author is the Director of ORIC and an Assistant Professor at the Department of International Relations, MY University, Islamabad.

Related Posts

Get Alerts