Matti Tedre and Henriikka Vartiainen
K-12 Computing Education for the AI Era: From Data Literacy to Data Agency
The question of how to teach classical, rule-based programming has been driving much of the computing education research since the 1950s. In the K–12 (school) context, a consensus has emerged over time on the paradigmatic elements of computing education, which implicitly assumes a von Neumann computer executing instruction sequences guided by imperative programs. Within this framework, many researchers have focused on how to facilitate learners to develop an accurate mental model of what the computer does when it executes a piece of code.
However, the traditional programming approach in computing education is inadequate for understanding and developing machine learning (ML) driven technology. ML has already facilitated significant advancements in automation, ranging from speech and image recognition, autonomous cars, and deepfake videos to super-human
performance in board and computer games, and more. Many data-driven approaches that power today’s cutting edge services and apps significantly diverge from the central paradigmatic assumptions of traditional programming.
Consequently, traditional views on computing education are increasingly being challenged to account for the changes that AI/ML brings. This keynote talk presents early results from a study on how to teach fundamental AI insights and techniques to 200 4–9 graders in 14 primary schools in Eastern Finland. It describes the learning environments, tools, and pedagogical approaches involved, and explores the paradigmatic and conceptual changes required in transitioning from teaching classical programming to teaching ML in K–12 computing education. It outlines the mindset shifts required for this transition and discusses the challenges posed to the development of curricula, educational technology, and learning environments. It further provides examples of how AI ethics concepts, such as algorithmic bias, privacy, misinformation, diversity, and accountability, can be integrated into ML education.
The talk discusses the relationship between different literacies in computing and presents an active concept, data agency, that refers to people’s volition and capacity for informed actions that make a difference in their digital world. It emphasizes not only the understanding of data (i.e., data literacy) but also the active control and manipulation of information flows and the ethical and wise use of them.
Matti Tedre is a professor of computer science with University of Eastern Finland. His research in the field of computing education research focuses on how to empower learners with insight into the mechanisms, opportunities, and dynamics of data-driven (AI) systems, but also their weaknesses, biases, and risks, and how they can be used to discriminate, polarize, create insecurity, and break trust.
Henriikka Vartiainen holds a PhD and title of Docent in Education. She works as a university lecturer and senior researcher at the University of Eastern Finland, School of Applied Educational Science and Teacher Education. Her research interests include data agency, co-design and design-oriented pedagogy. In a project funded by the Strategic Research Council established within the Academy of Finland (2022–2028) she leads a work package on data-driven design that addresses the question of what kind of learning processes and best practices emerge when learners invent and create their own machine learning applications.
Paul Denny, Brett A. Becker, Juho Leinonen and James Prather
Chat Overflow: Artificially Intelligent Models for Computing Education – renAIssance or apocAlypse?
Recent breakthroughs in deep learning have led to the emergence of generative AI models that exhibit extraordinary performance at producing human-like outputs. Using only simple input prompts, it is possible to generate novel text, images, video, music, and source code, as well as tackle tasks such as answering questions and translating and summarising text.
However, the potential for these models to impact computing education practice is only just beginning to be explored. For example, novices learning to code can now use free tools that automatically suggest solutions to programming exercises and assignments; yet these tools were not designed with novices in mind and little to nothing is known about how they will impact learning. Furthermore, much attention has focused on the immediate challenges these models present, such as academic integrity concerns. It seems that even in the AI-era a pending apocalypse sells better than a
Generative AI will likely play an increasing role in people’s lives in the reasonably foreseeable future. Model performance seems set to continue accelerating while novel uses and new possibilities multiply. Given this, we should devote just as much effort to identifying and exploiting new opportunities as we do to identifying and mitigating challenges.
In this talk, we begin by discussing several concrete and research-backed opportunities for computing educators. Many of these have already shown great promise in positively impacting current practice. We then discuss more short- to medium-term possibilities in areas such as student recruitment, and curricular changes. Finally – against our better judgement – we speculate over the longer-term, including rethinking the very fundamentals of the practice of
teaching introductory and advanced computing courses. In these discussions we suggest potential research questions and directions. Although making remotely accurate predictions in such a fast-changing landscape is foolhardy, we believe that now is the time to explore and embrace opportunities to help make positive change in as many computing classrooms as possible.
Paul Denny, University of Auckland
Paul enjoys exploring how computing students engage with online learning tools, and is particularly interested in how their experience can be impacted through user interface design and tool feedback. His fascination with large language models began in August 2021, after seeing the extraordinary performance of Codex following its release in private beta.
Brett A. Becker, University College Dublin
Brett is interested in how humans learn to program and how they perceive this process. He is fascinated by the interaction between humans and computers, exemplified by his obsession with programming error messages and what AI has to do with them. He is far from alone in his belief that that generative AI will dramatically change the way programming is taught and learned and is keen to try to keep up with the seemingly non-stop acceleration
of the capabilities of AI. He is not sure if he is surprised or not that LLMs have offered yet another parallel between programming and natural languages, in that LLMs have demonstrated similar capabilities in both domains through very similar mechanisms.
Juho Leinonen, University of Auckland
Juho explores how to best support and engage diverse learner populations with educational technology and artificial intelligence. Recently, he has researched the potential opportunities that large language models could provide for introductory programming instructors such as automatically creating personalised exercises, enhancing programming error messages with LLMs, and creating code explanations for students using LLMs.
James Prather, Abilene Christian University
James is very interested in how novices learn to code. His research has examined novice programmer interaction with compiler error messages and novice programmer metacognition and self regulation. Recently he has worked on multiple papers on the impact of LLMs on introductory computing education.