Task Description
The modern development of technology has yielded excellent results, creating innovative systems that simplify people’s lives by making them more convenient and accessible. Artificial intelligence is one of the systems that can optimize many workplace processes when correctly set up. In some cases, AI systems can also prevent the problem by generating content that is not always strictly true. Writing an essay on any topic requires adherence to precise rules and norms of argumentation, as well as a clear division into sections. Thus, the AI was asked to complete an initial draft that could be further refined and used for more in-depth work.
Problem of Sources
A lot of the work involved training AI to write sentences that were based on accurate data and citations. The use of articles is the main obstacle to writing an essay in its entirety. It is worth noting that to complete this task, the experiment used the AI model ChatGPT 3.5, which has significant limitations due to its free version. Citing and relying on academic works and journal articles is an essential part of writing any essay.
Thus, it was studied that in 7 out of 10 cases, AI invents sources by substituting the names of authors and indicating the titles of works that cannot be found on the Internet(Grossman et al. 75). This problem does not allow the AI to write an essay completely from start to finish since it will not be evidence-based. This problem can be solved by sequentially loading academic articles into the AI chat to compose an essay.
At this time, one may encounter another problem: the limited search capabilities of AI and the limited capacity of texts. It is impossible to simultaneously upload large volumes into the chat that exceed several pages (Abioye et al. 2). Thus, to write an essay that may contain several sources, it will be necessary to make sufficient efforts to periodically and gradually download different articles.
At the same time, AI can also incorrectly incorporate information from them when complex structures cannot be fully perceived. Thus, the first and most critical problem in writing essays with AI is the use of literature. However, this also applies to external resources that Chat cannot fully reference, and accordingly, its citing capabilities are limited.
Problems with Originality of Thought
Writing an essay that is intended to analyze a profound and significant topic in the context of the fiction under consideration. To do this, the author needs to create their own unique opinion, which can be based on certain previously studied materials, including the primary source. Writing essays using AI in this context is associated with specific difficulties that are reflected in the algorithms’ processes. They can browse the Internet to see whether a given title matches certain words in the search parameters. Thus, the theoretical knowledge of AI is, in any case, based on already existing data (Jordan 4). This information cannot be original and must therefore be analyzed and finalized by a person.
The power of AI can be used to provide a comprehensive search of existing knowledge on the topic being sought. Given these limitations, most of the analysis in essay writing should be done without the help of AI. A significant part of the work in this case will be setting parameters for ChatGPT that can move the research in the right direction.
For this purpose, the most suitable goal may be to write general headings that can advance the implementation of the writing process. Such a basis will differ significantly from the initial prompt, as the results may not fully reflect the specified parameters. One weakness of the language model is its inability to construct sufficiently long texts that cover all necessary subtopics (Jordan 4). Thus, to preserve the material’s originality and the presence of new ideas, a necessary step was to compile a large portion of the information from the data received.
Phased Content Change and Reasons for It
A significant part of teaching AI to write involved adjusting the veracity of the information. To correctly observe all the structural features of the work about which the essay is being written, it was necessary to double-check all aspects of the data provided by the AI. One of the significant weaknesses in its work was the numerous incorrect cuttings. In some cases, entirely original quotations have been discovered that have no relation to the original work (Grossman et al. 75). This significantly limited the possibilities of writing the work since, to rely on the text of the work, ChatGPT had to be provided with the full text of the book.
However, the existing character limits in the free chat model do not allow the entire text to fit in a single message. Because of this, it was necessary to divide the process into several sections in accordance with the outline compiled by the chat. Such painstaking work significantly increased the overall time required for essay writing. Thus, this aspect is another reason why the initial prompt could not be applied without changes and constant additions.
Positive Aspects of Writing
The strength of ChatGPT’s essay-writing process was the speed of query processing and the further opportunities for personal interpretation of the results. Collaboration with a text editor enabled comprehensive evaluation of all submitted results. By loading analytical articles one by one, the AI could produce relatively high-quality results that were always structured, including writing an introduction and conclusion (Zhang et al. 10). This helps to navigate what information the AI assistant provides.
Another positive aspect is the presentation of possible directions for a person to take when composing an essay. This can be presented either in thesis form or in complete paragraphs, depending on the changed prompt. Thus, with sufficient regulation, AI can provide entry-level information that can become a suitable basis for further processing and analysis.
Utilization of Personal Experience
One of the main elements of essay writing is processing the work or topic under review through the prism of personal experience, which can effectively complement the material. In this context, the AI cannot provide this aspect under any prompts. This data can be falsified in a specific way by assigning ChatGPT an initial role from which the final response will be constructed (Paschen et al. 406). However, this method has limitations, as it cannot fully capture the thoughts and concepts a person might have when writing an essay on the same topic.
Adaptation of various complex elements, together with the questions raised in the original work on which the essay is based, cannot be fully covered by a single prompt, which creates additional obstacles. Taking these data into account, a distortion of possible emotions or the transfer of experience may be allowed, which would be impossible when composing a text by a person who personally lived certain important moments and draws conclusions based on them.
Essay Writing with and without AI
Teaching AI to write essays involves several imperative processes that must be continuously regulated and monitored to achieve the highest quality results. As described above, the initial prompt may not be suitable for the final version of the text, as it requires multiple attempts to adapt to the downloaded data. This results in gaps in sent and received messages, which affects integrity. A significant advantage is ChatGPT’s fast responses and the provision of general information that can be used after verification. However, creating a complete work is not possible as it will require section-by-section writing for better control.
In general, I find writing an essay without AI easier because it does not require spending much time entering all the critical parameters and data for the AI to compose the entire text. In addition, independent work will be of higher quality as it may contain original interpretations of human experience in relation to the topic of essay writing. This ensures completeness and completeness in both academic practice and artistic approach.
Works Cited
Abioye, Sofiat O., et al. “Artificial intelligence in the construction industry: A review of present status, opportunities and future challenges.” Journal of Building Engineering, vol. 44, 2021, pp. 1-13.
Grossman, Maura R., Paul W. Grimm, and Daniel G. Brown. “Is Disclosure and Certification of the Use of Generative AI Really Necessary?” Judicature, vol. 107, no. 2, 2023, pp. 69-77.
Jordan, Michael I. “Artificial intelligence—the revolution hasn’t happened yet.” Harvard Data Science Review, vol. 1, no. 1, 2019, pp. 1-9.
Paschen, Jeannette, Matthew Wilson, and João J. Ferreira. “Collaborative intelligence: How human and artificial intelligence create value along the B2B sales funnel.” Business Horizons, vol. 63, no. 3, 2020, pp. 403-414.
Zhang, Liang, et al. “A review of machine learning in building load prediction.” Applied Energy, vol. 285, 2021, pp. 1-43.