The field of AI in education has exploded in the past ten years. Many factors have contributed to this unprecedented growth, such as the ubiquity of digital devices in schools, the rise of online learning, the availability of data and fast growth in related fields such as machine learning and data mining. But with great power comes great responsibility: the flipside of the growth of AIED is that now our technologies can be deployed in large numbers to millions of children. And while there is great potential to transform education, there is also considerable risk to destroy public education as we know it, either direcdy or via unintended consequences. This is not an exaggeration: in recent months, we have indeed learned that the combination of social media, technological ubiquity, AI, lack of privacy, and under-regulated sectors can go the wrong way, and that AI has today a disproportionate power to shape human activity and society. On the other hand, most schools of education around the world are not equipped - or not interested - in this debate. They either ignore this conversation, or simply attack the entire enterprise of AI in education—but these attacks are not stopping wide dissemination of various types of AIED projects in schools, mainly driven by corporations and fueled by incentives that might not work in the benefit of students (i.e., massive cost reduction, deprofessional-ization of teachers, additional standardization of content and instruction). In this scenario, the academic AIED community has a crucial responsibility—it could be the only voice capable to steering the debate, and the technology, towards more productive paths. This talk will be about the hard choices that AIED needs to face in the coming years, reviewing the history of AI in education, its promise, and possible futures. For example, should we focus on technologies that promote student agency and curricular flexibility, or on making sure everyone learns the same? How do we tackle new learning environments such as makerspaces and other inquiry-driven spaces? What is the role of physical science labs versus virtual, Al-driven labs? How can AIED impact— positively and negatively—equity in education? I will review some of these issues, and mention examples of contemporary work on novel fields such as multimodal learning analytics, which is trying to detect patterns in complex learning processes in hands-on activities, and new types of inquiry-driven science environments. The AIED community is strategically placed at a crucial point in the history of education, with potential to (at last) impact millions of children. But the way forward will require more than technical work—it will require some hard choices that we should be prepared to make.
展开▼
