Malaysian university seeks approval for AI-taught degree

A Malaysian university expects to be the first to launch a degree programme taught and assessed entirely by artificial intelligence, following research into technology that scans students’ brain signals. 

Sasa Arsovski, dean of the Faculty of AI and Robotics at the privately owned Raffles University, has developed a platform called MyAi Teaching Assistant that allows academics to upload their lecture presentations, which are then turned into text notes by AI. The technology also creates an avatar based on an image of the lecturer and can clone their voice, allowing the avatar to deliver a virtual class from the notes on behalf of the academic.

“We take the experience and the expertise of the lecturers, which is all summarised in their PowerPoint presentations, and then we have a properly designed lesson,” Professor Arsovski said. 

Students are also able to ask the avatar questions, and vice versa. Then, “according to their conversation and the lesson topic, the system will generate personalised homework for the students”, he said. “The system will automatically grade [their] homework.”

Professor Arsovski said he was already using the technology, including to cover his classes while he was away. Now, the university plans to expand its use of virtual teachers, launching a bachelor’s degree in data science “completely delivered by AI”. 

The course, which is set to start in October, will be available to students living anywhere in the world and will be significantly cheaper than a traditional degree programme. Professor Arsovski was hopeful that this will help “democratise education”. 

One hurdle still to overcome, however, is accreditation: the course has been submitted to the Malaysian Qualifications Agency for review, but it is unclear how accreditation bodies will react to programmes of this nature.  

Balancing the opportunities for innovation that AI brings with quality assurance is “difficult”, said Helen Crompton, professor of instructional technology at Virginia’s Old Dominion University. “AI is changing so quickly, with so many systems, that it is hard,” she said. “Accreditors have to be agile and open to new ideas and ways of providing new opportunities.”

Commenting on the proposed degree programme, Professor Crompton warned against replacing teachers, “as the quality of the teaching cannot be as good”. 

“While AI can now mimic emotional intelligence as well as cognitive intelligence, it is still only mimicking and does not really care if the student learns or not,” she said. 

However, she added, where access to education is limited, AI “can do a great job”.  

“In cases where students are remote and cannot be in person…AI tutors are wonderful,” she said. 

Dr Arsovski has also conducted research into decoding students’ brain signals, using technology to determine what they think are the correct answers to multiple-choice exams. Using similar technology, the professor has also been able to control the direction drones fly using his mind.

This research is in the same vein as developments happening in Silicon Valley, including wearable technology patented by Apple that monitors brainwaves. 

“All technology currently is going in this direction to make synchronisation with the human mind,” Dr Arsovski said. “AI will see through our eyes and understand where we are and what we are doing…[and] eventually it will give us the advice and suggestions [of] what to do in a specific situation.”

Dr Arsovski believes students should be the ones to decide whether they’re prepared to sacrifice human teachers. “Yes, we will lose the face-to-face connection with the students, but I think the market will give us the answer,” he said. 

helen.packer@timeshighereducation.com