Artificial intelligence at the University of Tartu

4.3. What about writing theses?

Text-based AI model can be used to write theses, but a proper thesis cannot be completed solely by AI. In the age of AI, it is essential to support students throughout the entire writing process and raise awareness about the significance of its various stages.

Has the advent and spread of AI made you think that writing theses (i.e., bachelor’s and master’s theses) and other extensive texts is pointless because a Text-based AI model could easily do the work for the student? In this chapter, we will debunk the fears and misconceptions associated with writing theses and discuss why theses are still crucial in the age of AI.

The ability of text-based AI models to quickly produce seemingly coherent new texts has caused a lot of fear and debate about students potentially using these applications to write their theses, thus reducing the amount of independent work. This raises the question of whether writing theses loses its value if the text-based AI model can produce a sufficiently good job for the student. There seems to be a public perception that students no longer need to do anything except input a couple of lines of text into an AI application, after which the robot quickly generates the text.

With such a perception, it is easy to question the authorship of written assignments, especially since there is no practical way to determine whether a human or AI created the text (see also the University of Tartu’s stance on detecting text-based AI model usage (in Estonian)). The inability to detect this may lead to academic dishonesty if students submit AI-generated work as their own. On the other hand, constant doubts about authorship may create situations where students are unjustly accused of using AI.

However, it is helpful to remember that as long as independent critical thinking and argumentation skills are valued in academic higher education, the thesis remains relevant. Indeed, using text-based AI models can be beneficial during certain stages of writing a thesis (for more on this, see the section on thesis writing in the University of Tartu’s guidelines), but this does not mean that entire theses can be easily written using text-based AI models.

Here are some reasons why theses are still crucial in the age of AI:

  • Writing theses is at the core of academic education, developing thinking and argumentation skills. Just as bread is an essential food for humanity, a thesis is the fundamental output of academic research, helping to maintain the quality and depth of higher education.
  • The unconsidered use of text-based AI models for writing texts may diminish students’ ability to conduct independent research and think critically. If students rely too much on AI, they will lose the opportunity to develop the skills necessary to engage in argued discussions, both academically and professionally. Educators and supervisors’ task is to support and motivate students throughout the writing process so they are not tempted to over-rely on AI-generated texts.
  • A thesis should not be seen as just a text. To complete a proper thesis, students must undergo several stages, including identifying the research problem, setting objectives, formulating hypotheses, reviewing sources, choosing a methodology, and collecting and analysing data. All of these stages must be explained and justified in a coherent text. Writing a thesis is like project management. Therefore, the process of writing a thesis develops skills and knowledge that the use of AI cannot fully replace.
  • Writing a thesis requires independent thinking, problem-solving, and analysis. AI can be used as a tool, but the quality of the thesis depends on the student’s knowledge and skills, which can only be developed through independent work and discussions with supervisors.
  • Maintaining academic integrity and ethics is essential to avoid the inappropriate use of AI when writing theses. Universities must support students in understanding the actual value and purpose of thesis writing and ensure that AI is not used for fraud.

Longer written assignments could include components that require students to reflect on their writing process and engage in critical analysis while encouraging original author contributions and developing a personal style. The writing process can also be combined with collaborative learning, in which text creation occurs with peers (see more on Collaborative Writing and Writing Camps (in Estonian)).

Reflecting on the writing process, collaborative learning, and a well-designed assessment process help mitigate the risks associated with AI in thesis writing.

In the age of AI, careful thought must also be given to the principles of thesis assessment. It is certainly not enough for the defence committee to read through the text and provide their evaluation. The assessment of theses should more significantly take into account the following:

  • The student’s active involvement throughout the entire writing process,
  • Their participation in supervision meetings, and
  • Continuous work on improving the text.

Moreover, more weight should be given to the defence process. A thesis “created” in one or two weeks should not qualify for defence.

Even though the possibility of using AI may change the thesis writing process, the availability of AI does not diminish the importance of theses in higher education. Theses remain significant in the AI age because they develop students’ critical thinking, analytical skills, and scientific reasoning competence. AI can be a helpful tool but cannot replace the thesis writing process. Universities must continue to emphasise the importance of these and support students in their writing.

Read more: 

Accept Cookies