The Future of AI in Education: Uses, Ethics & Applications

The field of education is quickly becoming revolutionized with the power of AI. With the pandemic forcing many students to learn online and the emergence of AI-powered educational tools, we have to explore how this new phenomenon is unfolding. 

Marc Antoine Dilhac, Scientific Director at Algora Lab and Professor of Philosophy at the University of Montreal, along with Ekaterina Kochmar, co-founder and CSO of Korbit, and Louis Normand, Pedagogical Advisor at Collège de Rosemont, provide us with an insight into the current use of AI in education, and its moral and ethical limitations. 


Current Functions of AI

AI has proven to provide many valuable functions in education. Dilhac explained, for example, how AI can now be used to assess students’ comprehension in real-time, adapt the pace of learning to the students’ abilities, help in the design of courses, and in the evaluation of exams. 

However, what’s important to note is that there are some functions that AI can perform in isolation, and some functions that can only be performed with the help of humans. For example, AI can help provide instructions and exam exercises to aid students in understanding certain concepts. Furthermore, it can stimulate the memorization of content where use of natural language processing techniques can be used to select specific learning content. AI, however, needs the help of humans when it comes to planning the learning stages and objectives for the student, facilitating the empowerment of the learner, and finally, personalizing support throughout the learning journey. 


Korbit as a Modern Example of AI in Education 

One-on-one tutoring is an essential aspect of education. It has proven time and time again its contribution to a learner’s success. However, Ekaterina Kochmar emphasizes the difficulties of one-on-one training, such as attending to the needs of each student, especially with the current pandemic. For this reason, Korbit was created. 


“Our main intention is to provide to a wide variety of students a personalized education of high quality, and to democratize education,” said Kochmar. “This is the mission we pursue at Korbit. This Socratic dialogue-based system uses machine learning, natural language processing, and reinforcement learning to provide personalized interactive online learning.”


With this AI-powered educational system, students are provided with step-by-step guidance when solving academic problems. These active learning and problem-solving techniques are very effective in helping students learn and efficiently retain information. 


Incorporating AI at the College Level 

“To my knowledge, it’s not a widespread practice among teachers in their daily practice to use artificial intelligence to do tasks like automated assessment, or to do planning, or to create instructional scenarios,” said Louis Normand.

However, despite AI not being fully utilized in colleges presently, Normand stresses its usefulness. Educational institutions tend to be diverse in many ways. In our present day, the student population almost always encompasses ethnic diversities, socio-economic diversities, and special needs diversities. This is where the concept of personalized automated learning would thrive, as it would cater to the need of every student without the extra strain imposed on the professor.

“If you want to be able to provide an educational device that is most effective, you have to start thinking about personalized, automated learning,” said Normand.

Not only this but there is hope within the academic community of AI developments that would eventually allow AI-driven exam and essay evaluation, as that would remove a burden off the shoulders of many teachers and give them more time to interact with their students.


Perspectives, Criticisms, and Ethics of the Uses of AI

AI has developed the functions of effectively creating content and training specific skills. Once discovering the skills that can be taught automatically, AI can be adapted to these skills and be used to harness them, improving the learner’s journey. However, one aspect of education that plays a crucial role in the learner’s journey is socialization, an aspect  AI notably leaves out. 

Dilhac stresses the social function attached to education and emphasizes that its presence is crucial to allow the acquisition of skills and the retention of information that AI can present. 


“I started by saying that artificial intelligence can transmit content and promote the acquisition of skills,” said Dilhac, “but what it leaves out, in reality, are functions that make it possible to concretize the skills and the acquisition of content.” 


The value that human relations have in education and contribute to the overall learning outcome cannot be undermined. In light of this, Dilhac emphasized that AI cannot be used in isolation but rather as a tool to be used within the social context of education. 

There is also the issue of determining the student’s learning profile. AI cannot escape the bureaucracy in every institution. With the incorporation of AI, it might even be accentuated, introducing the tendency to lock the learner within a certain learning profile that they might not necessarily be fit for in the future. 

Not only this, but the digital divide imposed by systems of AI and online learning also poses a problem for students. Normand noted that it wasn’t easy getting all students equipped with the right resources to complete their studies online. This was luckily achieved at the College de Rosemont as they had the resources to do it, but in many places in the world, students would not have the means to access the necessary equipment to use AI-powered tools.

These are the dangers AI can impose if not well integrated within the social sphere of education.

So how can artificial intelligence tools be used within the educational, social sphere? With the development of Korbit, the platform aims to offload many of the pressures that professors face with their students to use their time with their students in a more effective way. For example, by delegating more tedious tasks that can be performed by AI-powered tools, teachers can give helpful advice to students on specific difficult questions. 

“Humans, both student and teacher, are at the heart of our application,” said Kochmar.

In light of this, major ethical points were taken into consideration when developing the platform. The only piece of information required from the student or teacher when accessing the Korbit platform is an email address. The use of sensitive information is strictly prohibited, honoring principles of human autonomy, confidentiality, ethical data collection, diversity, and non-discrimination. On top of that, the platform is designed without exclusion criteria, paying much attention to the transparency of the decision algorithms, links, and their potential impact on the well-being of students and teachers. 


Regulating AI Moving Forward

The Montreal Declaration for a Responsible Development of AI serves as the framework for AI in education. Dilhac presents it and points out the multiple principles that need to be adopted as we move forward with the use of AI in education, such as the well-being of both student and teacher, autonomy and solidarity, diversity, and democracy. 

As the ethical issues are taken into account, the benefits of artificial intelligence in the educational domain can be revolutionary.

“In Korbit, we strongly support the principle of welfare because we believe, and we hope, that Korbit adapts to the individuals to improve the conditions of their work and life, and help them better exercise their intellectual abilities,” concluded Kochmar.


Text: [Huda Hafez]


Leave a Comment

Your email address will not be published. Required fields are marked *