Intelligent systems in education: reflections from expert Wayne Holmes

Mr Wayne Holmes, an Associate Professor in the UCL Knowledge Lab at University College London and an expert in artificial intelligence and education, participated as a specialist guest speaker at Ceibal Foundation Digital Space. In his opinion, rather than focusing on technologies that try to replicate what a teacher can do, it is necessary to develop tools that will help them use them in innovative ways to do what, as professionals, they believe should be done in the classrooms.

“We need to help both our students and our teachers to understand the limitations of artificial intelligence and then to take a really critical approach to the information being generated by them. I would encourage them to work together, to explore how such tools can be used to better understand what they output, to think about their impact on the classroom ecosystem, on data security and on the ethics of education,” Holmes said.

Artificial intelligence can focus on students, teachers or the institution. Holmes warned about the possibility of AI undermining student agency because it is instructionist (it says “Do this,” “Now do this”), which affects the quality of learning. He also stressed that no technology can replace the teacher. In this regard, he said that it is more useful in the background, that is, with a focus on institutions. The question is, however, how to use generative AI to improve teaching and learning, and support students. According to the expert, it is necessary to ensure that everybody – both students and teachers – understands how these tools work.

In his opinion, the purpose of these technologies is to take all students to the same learning destination, that is, the outcomes are not personalized. “A classroom and a school or a college is an ecosystem, and the educator is at the centre of that ecosystem, and our task is to help our students become the best they can become,” he pointed out. It is essential to emphasize the social and human relationship between student and teacher, “understand what they’re going through, their background, whether or not they had breakfast this morning, what they did yesterday. Knowledge of the curriculum we’re teaching, too.”

In addition, Holmes voiced scepticism about claims by AI developers that AI can address the digital divide, although he did recognize the advantage of several of these tools being free to use today. However, he highlighted the existence of some AI-enabled technologies that are being used to support people with physical and mental difficulties, thus filling a genuine gap, such as tools that automatically translate speech into text, which is very useful for students with hearing difficulties.

Another issue that Holmes reflected on is the risks of using these technologies. The expert feels it is necessary to discuss the ethics of education: “We’re not clear yet about what happens when we bring these three areas together: the data, the algorithms and the education. There is lack of control.” For this reason, he stressed that action must be taken to ensure that data processing is done to the highest security standards, safeguarding “the human rights of the students involved.” In this respect, he urged national leaders and policymakers to play a very proactive role in the relationship between AI and education, and to invest in research to see the impact in their own context on individual students using these tools.