1.9 AI and Teaching and Learning in VET (Introduction)
In this introduction to AI and teaching and learning in VET, Graham Attwell provides a short description of the different ways AI can be used for teaching and learning. 🎧
He refers to Nigel Shabolt who, in this interview in the National News, says that people may be aware of AI but that they need to understand its impact. Graham Attwell points out that AI tools are all around us, and many tools that we use in education and take for granted are built on AI, including search engines, grammar checkers, voice transcription services and so on. In this report, we are going to focus more on tools that are specific to education.
The key question, he says, is what role we want the AI system to play in the teaching and learning experience. While yt might seem obvious that the aim is to support the teacher rather than replace them and this is the direction that is most often discussed, earlier attempts to use AI for education were actually trying to replicate or replace the teacher. This is most often seen in adaptive learning or intelligent tutor systems, where AI is determining the learning activities for the student, including delivery and providing feedback – activities that are more often associated with the teacher.
Adaptive learning systems are some of the most mature AI-based education technology systems and have been shown to be very effective in some domains. However, they are not suitable for all types of courses or domain areas. They are most suitable when the domain knowledge can be very clearly defined and can be learned in a step-by-step way.
These systems are usually self-contained online systems, where the learner takes the course at their own pace.
Chatbots are a relatively mature application of AI and they are used in many domains. The AI aspects of chatbots include the ability to understand speech (natural language processing [NLP]) and the use of machine learning models to match the intention of the question to answers or actions. Typically, a chatbot interface will sit on a website and a user will type questions, then the system will either answer or ask further clarification questions.
Leeds Becket University has developed a chatbot called Becky for enrolling students. The development was based on research that showed that potential student applicants were largely uncomfortable talking to universities on the phone. Becky takes the user through the whole application process, including making offers.
This short video is a presentation of Bolton College’s Ada chatbot service. Ada can respond contextually to questions posed by students and teachers. It also describes the emergence and evolution of campus digital assistants. The video was commissioned by Jisc for Digifest19 and it was filmed by Suited and Booted.
AI-assisted marking software aims to help with marking beyond what has previously been possible with multi-choice software. Although the technology can be similar, there is a distinction between automated marking software (which aims to assist with providing a student with a grade or mark) and automated feedback software, which could be used to help the student with writing an assignment
China is experimenting with machine intelligence to mark essays on a large scale. According to the South China Morning Post, technology has been developed to interpret the general logic and meaning of the text in essays and is being implemented in a quarter of all schools, around 60,000 institutions. The platform undertakes a human-like judgment into an essay’s overall quality and assigns a grade. The platform can also provide recommendations for improvement, selecting from categories such as writing style, sentence structure and overall theme.
Although some institutions have turned to proctoring solutions for remote assessment, making the use of tools such as face identification, these remain controversial, with concerns expressed that they discriminate against some student groups such as those with disabilities.
The Watson Tutor is an intelligent tutoring system designed to improve student outcomes and engagement. It uses AI software building on IBM’s Watson platform with learning and content expertise from Pearson Education. The Tutor creates an interactive dialogue in chat that replicates the questions and feedback responses of a good human teacher.
AI could be also be used to aid group formation, moderate, or monitor groups, or participate as an active agent in group discussions.
Other uses of AI for teaching and learning include recommendation engines and Augmented Reality and Virtual Reality.
WildFire claims to be the world’s first content creation service, automating the whole process of creating online learning courses. WildFire has been used to automate the production of 138 modules of learning, delivering this in eight weeks and at just 10% of the cost of more traditional method.
At least in the UK, Most education institutions are at the early stages in a maturity model, looking to understand how AI can be used and what its potential is. There is also a relatively small number of institutions with embedded AI-driven services.