The Knowledge Commons in the age of AI: opportunities and risks for urban smart learning

This paper reflects on the state of the knowledge commons in the age of artificial intelligence (AI), broadly revisiting areas discussed in prior work by the author concerning human and non-human interactions with place-based knowledge in contexts of informal urban smart learning [1, 2]. The knowledge commons is loosely regarded as all the open, free-access knowledge content available to anyone via the Internet, both amateur and expert knowledge. Smarter urban citizen learning is framed here in terms of accessing knowledge information efficiently at point of need in an urbanised built environment, where knowledge might be readily provided through suitable technical infrastructures. Knowledge mapping, seeking and interactions, push/pull knowledge delivery, recommender system criteria, significance of serendipity, personalised knowledge management and other related issues are considered in the context of artificial intelligence technology capabilities and potential risks. Technical infrastructure is conceptualised as a reasonable interpretive imaginary of what smart ad-hoc urban learning could become, in a civically owned, interoperable, app agnostic platform designed for any kind of citizen learning and collaboration (for example described in [3, 4]). The position of the paper seeks a balanced criticality for perspectives on both AI data analysis and generative AI content, additionally acknowledging that AI technologies as they relate to the knowledge commons for urban smart learning could potentially act in a context of custodianship for governance of the digital knowledge commons as a public good.