Artificial Intelligence in Education as a Rawlsian Massively Multiplayer Game: A Thought Experiment on AI Ethics

Benjamin Ultan Cowley, Darryl Charles, Gerit Pfuhl, Anna-Mari Rusanen

Research output: Chapter in Book/Report/Conference proceedingChapterScientificpeer-review

Abstract

In this chapter, we reflect on the deployment of AI as a pedagogical and educational instrument. When AI enters into classrooms, it becomes as a project with diverse members who have differing stakes, and it produces various socio-cognitive-technological questions that must be discussed. Furthermore, AI is developing fast and renders obsolete old paradigms for, e.g. data access, privacy, and transparency. AI may bring many positive consequences in schools — not only for individuals, or teachers, but for the educational system as a whole. On the other hand, there are also serious risks. Thus, the analysis of the educational uses of AI in future schools pushes us to compare the possible benefits (for example, using AI-based tools for supporting different learners) with the possible risks (for example, the danger of algorithmic manipulation, or a danger of hidden algorithmic discrimination). Practical solutions are many, for example the Solid protocol by Tim Berners-Lee, but are often conceived as solutions to single problems, with limited application. We describe a thought experiment: "education as a massively multiplayer social online game". Here, all actors (humans, institutions, AI agents and algorithms) are required to conform to the definition of a player: which is a role designed to maximise protection and benefit for human players. AI models that understand the game space provide an API for typical algorithms, e.g. deep learning neural nets or reinforcement learning agents, to interact with the game space. Our thought experiment clarifies the steep challenges, and also the opportunity, of AI in education.

Original languageEnglish
Title of host publicationAI in Learning : Designing the Future
EditorsHannele Niemi, Roy D. Pea, Yu Lu
Number of pages20
Place of PublicationCham
PublisherSpringer
Publication date6 Nov 2022
Pages297-316
ISBN (Print)978-3-031-09686-0, 978-3-031-09689-1
ISBN (Electronic)978-3-031-09687-7
DOIs
Publication statusPublished - 6 Nov 2022
MoE publication typeA3 Book chapter

Fields of Science

  • 516 Educational sciences
  • 113 Computer and information sciences
  • artificial Intelligence
  • learning assistant
  • learning analytics
  • massively-multiplayer game
  • thought experiment
  • ethics

Cite this