As the world continues to evolve, technological advancements in computer science have become an essential part of our daily lives. Artificial Intelligence (AI) as a new technology has increasingly become an important tool in various fields, including education. The use of AI in education has significantly changed how students learn and interact with course materials. One of the questions that arise in this context is whether AI tools can help students pass their university courses and what are the ethical implications.
In recent times, universities have been grappling to comprehend the capabilities of AI applications like ChatGPT and provide guidelines on their usage. Additionally, there is now a call to teach students how to utilize these tools effectively.
The faculty at the University of Bath has been exploring the potential challenges and opportunities associated with ChatGPT. James Fern, in particular, posed the question of whether students could utilize the online tool to respond to assessment questions or write cover letters, write poetry, or even short stories. ChatGPT can provide responses in natural language and engaging content, including generating essays and emails.
Regarding multiple-choice questions, ChatGPT has shown impressive performance and has been able to answer them accurately to almost 100%. However, for more complex questions that demand critical thinking, which typically constitute a significant portion of assessments, the tool needs help to provide correct responses.
Chat GPT Essay Example
During a final-year assessment, an example indicated a student did not produce the response given by ChatGPT. According to James, the answer seemed well-written and professional at first glance. However, some of the statements were reminiscent of a GCSE student rather than a university-level writer.
Furthermore, the tool tends to repeat the exact phrasing of the question in its introductions and conclusions but with slight variations. Additionally, when referencing sources of information, a standard requirement in academic writing, ChatGPT generates false sources. Without knowledge of how significant language models function, you can easily be deceived into thinking that the references generated by ChatGPT are legitimate.
Since its public release about six months ago, several students have been trying to decide when using this AI in their academic work is acceptable.
During a speech on Monday, Education Secretary Gillian Keegan acknowledged that AI has already started to impact the education sector and suggested its possibilities that could assist teachers with creating and curriculum lesson plans and grading assignments.
The Quality Assurance Agency, responsible for reviewing standards at UK universities, has released new guidelines recommending equipping students with AI skills to apply in the workforce. It encourages universities to educate new and returning students, upon commencement in September, about the appropriate use of AI and adjust their courses accordingly.
According to Kim Watts, a marketing lecturer, ChatGPT is considered an additional tool in the toolbox. Some students in her department have already utilized the tool in coursework assignments that require them to develop marketing plans. Kim encourages students who need help with how to proceed to experiment with its prompts.
Analytical Reasoning
Kim uses ChatGPT to create its own marketing plan, which consists of numbered points ranging from brand identity to social media usage. However, Kim indicates that such a plan would not meet the necessary standards. She explains that such a simplistic approach lacks detail and fails to demonstrate learning or critical thinking.
Kim suggests that neurodivergent students and non-native English speakers benefit most from using ChatGPT. However, she requires students who use the tool to attach their ChatGPT prompts and responses as an appendix, emphasizing their critical thinking progress beyond the chatbot-generated answers.
Exams in the Summer
Like many other universities, Bath University is currently developing its policy on AI tools such as ChatGPT, which is expected to be in place in September. The university will have a dedicated team to ensure that the policy remains up to date with the rapidly changing technology.
In the meantime, most staff are conducting in-person, invigilated summer exams. Dr Chris Bonfield, who leads the assessment design team, stated that the default assumption is that students should not be using ChatGPT this year. If staff permit the use of ChatGPT, they should provide clear plagiarism guidelines for students.
Keeping pace with the ever-evolving technology is a challenge for universities, but Bath has decided against considering a ban on its use.
Dangerous Times Ahead
The godfather of AI, Geoffrey Hinton, resigned from Google last week, regretting his work, as chatbots could soon surpass human intelligence. While computer scientist Prof Verena Rieser at Heriot-Watt University notes that her students are using chatbots creatively, she warns that they can generate misinformation on a large scale, causing concern in education.
The developer of ChatGPT, OpenAI, acknowledges the risks and incorporates safety measures into the system. Other companies, like Google, have also focused on AI, launching Bard for adults only. Prof Rieser expects to see different types of ChatGPT and safer models from various companies, hoping to mitigate possible dangers. However, she highlights that the models’ outputting of inaccurate, toxic, or hateful information still needs to be solved with no known solution.
