AI and Education: Addressing cheating through Milestones
In today’s world, technology has become an integral part of our lives. We have seen a vast increase in the use of artificial intelligence (AI) in many industries, and now, it is also becoming a part of education. AI, has the potential to benefit education and student outcomes greatly. It has caused concern that students will abuse this new technology. In this post, I'll discuss AI and the potential for cheating in education, and I will suggest a simple way that educators can use to ensure students' work is their own.
Introduction
Before I jump into the discussion about AI's impact on education, it is important to understand what AI is. AI can be thought of as a computer system that can perform tasks typically requiring human intelligence, such as recognizing speech, understanding language, making decisions, and solving problems. AI can automate processes, analyze data, and provide insights. AI is being used all around us and has been for years. For example, in medicine, it can be used to assist with checking mammograms for early signs of breast cancer, weather forecasting models for predicting the weather or in self-driving cars. Some new AI tools are going to have an impact on education - particularly in the fields of art and writing. These tools are some of the most impactful and rapidly developing fields in technology today. They can, of course, be used for good and bad, and many people are worried about their use in education to cheat the system.
AI, Education and cheating
The potential of AI in the education system is evident to those exposed to tools such as GPT-3, GoMoonbeam, JasperAI, Dall-e2, MidJourney etc. These tools have been developed to make complex tasks easier and faster, such as writing essays, creating art, researching topics, or creating presentations. Furthermore, they offer a wealth of information that can be used by students to understand their subject matter better and make informed decisions during exams or assignments. Certain advantages come with certain risks; some people may take advantage of these technologies to cheat. Therefore, it’s important for educators to remain aware of AI technology in schools.
For those who have not used AI before, here is a brief explanation of how they work. AI tools use algorithms and machine learning to generate content based on an input of information. The input of information can be in the form of prompts (short phrases describing what the end product should look like) through to whole sentences describing what should be included in an essay. Students can use these tools to create written work, such as essays or reports, quickly. Sometimes, these AI-generated works can pass off as original writing pieces if not adequately checked for authenticity. Some tools have a cost, but many have a very low price of entry or are entirely free (for the moment)
An argument often put out on social media is that these new AI tools make it too easy for students to cheat. They can type an essay title into the tool and, within seconds, receive an entire essay or question response in return. That is already happening in schools as the tools have gone viral in the past few weeks. If you haven't already, I urge you to try out an AI large language model (LLM for short), such as ChatGPT (openai.com) and get it to write an essay. You will be impressed. Further, it can write multiple different drafts, all different and all undetectable, by plagiarism detection software. Note- I would be willing to bet that companies such as Turnitin are actively looking at how to detect AI-drafted work.
For decades there has been the risk of cheating by a small number of students at most stages of education. As the stakes get higher, the risks of cheating become more acceptable. So students have bought essays written by others, have copied others' work from online sources or have even paid others to sit exams for them. The advent of AI essay writing tools is just another potential way to try to beat the system. But it is just a way to cheat. The method for detecting cheating has been around for a long time. However, the problem for many educators is that they now rely on the aforementioned plagiarism detection software. They may have forgotten how to detect cheating the old-fashioned way.
Of course, at the time of writing (December 2023), AI-assisted writing generates writing devoid of a student's voice. There is often a repetition of ideas. The AI can also 'hallucinate' content by which the content is made up (but sounds realistic). This is even to the point of making up references that look real. If you are already an expert in a subject, you will be able to recognise the errors. But most students using it for cheating will not. So we can use milestones to spot content that has been both plagiarised and AI-written.
Preventing AI-assisted cheating with Milestones
For any assessment or project that has an outcome that is assessed, milestones need to be used. Note: we can also include AI-assisted digital art creation, written computer programs, written tasks, and essays. Traditionally, before the arrival of plagiarism detection software, milestones could be used to monitor progress and check that a student had written the work. If they couldn't explain what they had written, they probably copied the content from somewhere else (intentional or unintentional plagiarism). Milestones are the term for regular checks for understanding that occur along the journey of work completion.
Milestones can't be anything that themselves could have been generated by AI. These checks must be carried out orally. The educator has to build them into their program, allocating time at the outset. The intention is not only to check whether the work has been completed by the student but also to support the learning process. Milestones can show when a student is struggling and needs support or when they may have misunderstood a brief and deviated from the desired path. They are not just for checking for cheating!
Milestones will need to be used with all work, not just that which is used for assessment. They need to become an accepted part of education. We need to clarify to students what is and is not acceptable from the outset, be it plagiarism or AI written work.
Issues with Milestones
No system is perfect, we still have students trying to submit work that is plagiarised after all, and milestones also have issues. One such issue is that they can be time-consuming. So educators will need to plan these at the outset and allow time for them to be carried out. It may be that not every student needs to be checked at every milestone, either. Smaller sample groups could be chosen, and these groups could also be pseudorandom, with certain students being targeted if there is a suspicion they are cheating. Or the whole class can be checked initially to set up the expectation that this is 'what we do', making the students aware of the process and how it is carried out. However, being consistent with your approach and always carrying out milestones is critical. It also enables formative assessment to become integral to the learning process rather than being a bolt-on that many disparage.
Universities may find this approach more challenging. Many lecturers and tutors have very large group sizes- which means that it is a problem doing milestones for groups of 50-plus students. Maybe there needs to be a rethink of the student-staff ratio at Universities or the use of support staff to carry out the checks.
For educators who have become used to relying on plagiarism detection services for checking authenticity, any change in the assessment process will be challenging. I would say embrace it and find exciting ways to carry them out. For example, at the school I teach, the final milestone before my biology students complete their report is presenting their findings to the rest of the class as a talk. Once the presentation has been made, they must be prepared to answer questions from the staff and their peers. This is beneficial in itself as the students can show what they have learned, get an experience similar to being a research scientist, and I can check their understanding.
Conclusion
Introducing AI into the education system has the potential to revolutionize how students learn and interact with educational material. However, it is essential that proper safeguards are in place to prevent the misuse of AI tools for writing student work. I suggest that effective milestones, checking for understanding along the way, are crucial.
As educators, we should check the journey as well as the end product.
Educators must also provide clear guidelines on the acceptable use of these technologies and monitor their usage closely to ensure they are not misused. By taking these steps, we can ensure that AI remains a powerful tool for improving educational outcomes without compromising academic integrity or cheating.
Next time...
These AI tools are here, and they are being used now by learners. They will become an integral part of future work, so we need to be able to get the most out of teaching and learning with them. In my next post, I will look at the potential uses of AI-assisted creation tools to address this.