There’s no getting round it: AI language models are everywhere and getting better all the time. These tools are also a hot topic in education. After all, how do you use them wisely? And how can you put them to good use? Joppe van der Schoot and Milo Kleijnjan from the educational AI platform Aidyn are making great strides in this area by using AI language models to optimally support students’ learning processes.
When AI language models such as ChatGPT and Bard (now called Gemini) made their debut in the digital world, this led not only to enthusiasm and new opportunities, but also to concerns. After all, what will the future of education look like with these language models? Will they pose a threat to what education is truly about? These concerns formed the basis for the creation of Aidyn.
Three core competences
According to Aidyn, good education should ensure that students develop three key competences: critical thinking skills, problem-solving skills and communication. AI language models have a direct impact on the development of these competences. Good education is about self-learning and building, and not blindly adopting all input from an AI language model. With this in mind, the founders of Aidyn set out to investigate: how can we use AI in such a way to support the learning process?
The Netherlands is only a small country and has a lot to gain in the field of Generative AI, according to Milo. "The Netherlands is surprisingly good at data privacy. But with the rise of big language models from the US and China, we are getting even smaller as a small country. While EU regulations are important, we should not wait to take action. If we want to preserve our educational quality and values, we need to get started on AI applications now. We must not lag behind."
A smart platform
By developing a smart AI platform for higher education, Aidyn wants both students and teachers to work in a positive way in an AI environment. The platform allows educational institutions to create a safe and transparent course-level generative AI environment based on their own educational vision and learning objectives. From this safe environment, Aidyn can take on a number of roles for the student, such as a personal tutor. And with input from the student, tutors are then supported again with insight from those same interactions.
Milo and Joppe: "Simply explained, as a student you get a personal AI tutor who knows everything about the course you are currently taking. This AI tutor communicates with you in such a way that you learn the most at your level. For example, Aidyn asks critical questions back: 'what do you yourself think about the applicability of topic X to situation Y?' This way, we continue to stimulate students' critical thinking skills. Moreover, it also gives teachers insight. Take the subject of economics, for example. If students often ask the AI 'what on earth is inflation?', then a lecturer knows that this might need more explanation in the next lecture."
Aidyn also tests other functions that contribute to its core competences. For example, Aidyn is currently testing a feedback function for essays. "Teachers are often busy, so they have less time to give targeted feedback," Joppe explains. "We are therefore investigating how it goes when our AI platform provides feedback. This way, a student can learn more from a paper and it saves teachers a lot of time. We are also working on an exam module, so Aidyn can take practice exams."
Partner
The idea is that the AI platform really becomes a conversation partner for students while they are at work. And in doing so, Aidyn ensures that every interaction supports student learning and the development of the three educational core skills. Therefore, the AI does not just give good answers, but continues to encourage students to think for themselves. "You steer the student with such a Socratic conversation towards the reflective process, where they start thinking for themselves," she says.
Ready for the next phase
Aidyn is in the early stages with their platform and is busy laying a strong foundation for the future. The company is doing this partly through pilots and talks with teachers and students. Some parts of the platform have already been successfully developed.
"We have now had discussions with about 40 to 50 teachers," says Milo. Joppe adds: "We are also in talks with investors. The world of AI developments is moving fast now. If we manage to bring in our intended investment from investors with valuable networks, we will be able to realise our plans faster and move to a market-ready product sooner. We are very excited about the future, where we can support education significantly. We want to be a calming factor in today's AI hectic and really help students and teachers."
"Briskr and events like the Business Angel Network Nijmegen (BANN) also help us with advice, network access and events," Joppe adds. "Through Mercator Launch, for example, we can meet investors. The BANN was also a positive experience for us. The event was well organised and a nice crowd was present. And the investors we met there, we still talk to!"
They are working on these bright future plans so far with a strong founding team of five. "In addition, from the very beginning we are already putting together an advisory board with people who have a lot of useful experience and knowledge," Joppe and Milo explain. "We are looking within universities and business for people who have an affinity with certain areas of expertise: for example, the ethical implications of innovation on society or sales. Anyway: our idea is going to get there. Because we are sure it is going to help a lot of people!"
This article was previously published at Briskr. Click here to read.

