Using AI for personalised learning in adult education


------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Text by Sara Rydin
Sara is a writer and the editor at the Nordic Network for Lifelong Learning (NLL) that is tasked with spreading information about adult education in the Nordic countries
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Towards the end of the 1990s when I was at university AI was just getting started. We knew it as ‘machine learning’ and the generative AI we have today was something we could only imagine. We did not have the computing power, knowledge of deep learning and necessary methods for generating information.
Generative AI may very well overhaul adult education in Sweden, as it opens up for personalised learning. The technology can adapt learning by offering increased support or challenges, based entirely on a students’ needs and circumstances.
Non-AI-based personalised electronic programs have been around for as long as computer technology itself. Whereas traditional programs have pre-installed ‘learning paths’, AI can dynamically adjust difficulty, tempo and content based on a student’s actions or answers. Additionally, AI can perform a more in-depth analysis of patterns and parameters encountered in learning situations, things that we humans have more difficulty identifying.
Teachers can use AI to adapt student materials and syllabuses based on elements such as established targets and existing knowledge. This in turn provides a personalised learning pace and just the right amount of repetition. The pace can be further customised based on demonstrated success or lack thereof.
Those working in adult education see first-hand how AI is a valuable tool during encounters with students completely new to Swedish, as it enables them to interpret conversations and translate texts to and from a student’s first language. This technology is so fast, it enables on the spot interpreting. Despite the errors in these translations, the results can still be used as they make communication possible.
Modern generative AI also provides a completely new type of support to students with disabilities. It can clarify, summarise complex texts and translate materials into better formats. It can also help students that struggle getting started with the writing process by offering them keywords, assisting with structuring text, or writing down what the student dictates.
Many education providers lack an AI action plan. How will it be used, and why? It seldom stops teachers from using AI to provide personalised learning . The technology they use is rarely fully adapted to a specific teaching scenario.
The problem with generic AI services is that the learning does not give the education provider control over the service’s knowledge base. This results in the school (and often the teachers) lacking control over the knowledge students receive. A good quality adult education AI service should build on reliable information from familiar course literature (at the school) and be aware of Swedish legislation and regulations.
There is the additional problem of how the data the teachers and students feed into the general AI service is processed and stored. Does this comply with Swedish law?
The education provider’s initial needs analysis will determine how they use AI. The simplest solution is to use one of the existing AI services for education.
The most important thing is that we give teachers the knowledge they need to be able to personalise their students’ education. Schools and public authorities also need to
- establish a plan for using AI
- train teachers and create structures for professional development
- develop personalised teaching
- create new arenas for exchanging experiences
- ensure that teaching remains the focus
- evaluate the technology and learning method.
Generative AI can support writing. It can also interpret and summarise a text, or read it aloud. This support is particularly helpful for those whose reading, writing or counting skills are slower or lower than those of others.
However, for this technology to be helpful, the less tech savvy must be given additional support. Teachers must approach their students based on their level in terms of technology and take their varying levels of digital literacy into account. Naturally, we do not want to create an even greater digital divide.
I also see a great need for special tools that cognitively adapt technology for people with intellectual disabilities or considerable reading and writing difficulties. Otherwise, we will end up ‘digitally excluding’ people, as AI technology builds upon text-based instructions (prompts).
We need to look beyond the technology and focus on the pedagogical benefits of AI for it to become a successful tool in adult education. Personalised learning has been a vision of the education sector for a long time, although it is only now that we have the tools needed to implement this vision on a wider scale.
Ethical considerations are central to developments. Education providers must use AI to boost – and not replace – encounters between teachers and students. Teachers must continue to play a decisive role and the technology needs to be a tool at a teacher’s disposal – and not a replacement.
Exciting times lie ahead, where students from different backgrounds and circumstances, and with various learning styles will have a fairer chance of developing based on their own needs.
Focusing on a well thought out pedagogical implementation will be the deciding success factor. With the right approach, AI will help adult education in Sweden provide individuals with personalised learning.