The Future of AI and Education: Our Final Human+ Tech Talk

The fourth Human+ Tech Talk was led by programme Fellow Dr Qian Xiao, who works on intelligent tutoring systems (ITSs) that can complement human teachers’ judgement. The topic of the talk was the challenges and opportunities of building online education systems. She was joined by both academic supervisors. Her humanities supervisor is Prof Keith Johnston, Assistant Professor of Education in TCD. Her  science supervisor is Prof Vincent Wade, Co-coordinator of the Human+ programme alongside History Professor Jane Ohlmeyer, and Chair of Computer Science in TCD. He has recently concluded his tenure as the founding Director of ADAPT. Chairing their panel was Peter Gillis, Innovation Services Lead at Trinity’s Learnovate Centre.

Mr Gillis began the session by encouraging debate on a topic that is very much in its early stages, inviting questions from both audiences in the room and on Zoom. The new and unique nature of intelligent tutoring systems was clear from the talks which highlighted that education and AI are still very much in their infancy: but thanks to innovation as found in Dr Xiao’s research, this may be a significant turning point. 

Prof Johnston was the first speaker, introducing an educational perspective to AI in classroom settings and the wider educational eco system. Giving a brief history of technology in education over the last 50 years, he noted that, despite two previous waves – the first with the microcomputer and the second with the internet –  technology has mainly been subsumed into wider educational systems. However, lately there has been a greater understanding based on these two waves. Progress, in this way, may not always happen as we expect it to: “We now know technology adoption is complex and takes time.” Prof Johnston suggested that AI may be the third wave of technology. As a result, we should be asking what combination of humans and computers can be the future of teaching. Identifying various possibilities of AI support – for students, teachers, and systems – he concluded by considering some of the problems of adopting AI in education. A shared understanding and educator input is needed, including what pedagogical approaches underpin it. Concerns about privacy and ethics must be addressed.

Taking the discussion in another direction, Prof Wade focused on the opportunities and challenges of AI itself. He began by highlighting its impact on society and how it is changing the way we work. Currently, we are not in a position where we feel in control of AI. This is because, he points out, AI thinks faster than humans. However, “when you realise this you can then learn to collaborate with it.” He made the important observation that by 2023, 47% of learning management systems will have AI in them. With this in mind, we need to decide what we will do with it. Personalisation is Prof Wade’s area, and he suggested that this will be a key way forward. Weaker learners benefit the most from personalisation, getting the support and attention they need. Crucially, it can also diminish the gaps between socio-economic groups. Concluding his talk, he echoed Prof Johnston’s caution, learning to balance surveillance and assessment, and reiterated that AI is there to support the teacher: “The context of learning is key, and AI may need a human tutor to get this context.”

Introducing the third and final speaker, Mr Gillis observed that Dr Qian Xiao’s talk takes off where Prof Wade’s ended, demonstrating how AI can be used practically in the classroom. Dr Xiao shared the results of her Human+ research on state-of-the-art deep learning techniques and how they are bringing new assessment measures to learning, something she has termed “deep assessment.” Discussing her case study in language learning, in which students used a mobile phone app to learn English, they used an ITS to record the actions of the learners. This resulted in a significant amount of data in which patterns of students’ behaviour can be found without any need for human labelling. It also led to fascinating findings regarding academic activity. The question of whether good learners always have good learning habits or poor learners always have poor learning habits was met with a resounding no. Dr Xiao then identified the need for human tutors. “Teachers know the specific types of learners. These can help broaden their potential.” It is important to have AI-human teacher models to find out who are the learners who need help, emotionally and motivationally.

The panel then answered questions from the floor, addressing the financial challenges of AI in the classroom,  and Dr Xiao discussed how her model is generic and can be adapted across subject areas, not just language classes. A pervading theme of the Q&A, as well as the tech talk as a whole, was the importance of empowering teachers as well as students and ensuring any AI systems are user-centric.

For further information on Qian Xiao‘s Human+ Project, continue reading.

Human+ is a five-year international, inter and transdisciplinary fellowship programme conducting ground-breaking research addressing human-centric approaches to technology development. Human+ is led by the Trinity Long Room Hub Arts and Humanities Research Institute and ADAPT, the Science Foundation Ireland Centre for Digital Content Innovation, at Trinity College Dublin. The HUMAN+ project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 945447. The programme is further supported by unique relationships with HUMAN+ Enterprise Partners.

Share with others