This last in our trilogy explores data as the foundation of AI systems. We learn how this enables mapping individual learners' progress and benchmarking in a teaching context, but also how that data exchange raises ethical issues.
We explore how artificial intelligence builds functionalities on different data streams and consider our options to select and influence such 'training data'. Investigating this from a position understanding teaching as enabling a learner’s response, we discover how intimate conversations with Romeo & Juliet arise from what manifests as the AI’s ‘agency’. Yet we have to check in how far this also enables interactions that we wouldn't want to encourage or support. Prompting listeners to engage in their own observations and interactions with machine learning, we advocate curiosity outside academic’s traditional comfort zones and building your own critical attitude alongside symbiotic relationships with relevant partners, agreeing work packages which relate to differential skill sets. Setting out a space for serendipity, and claiming a license to fail emerge as key catalysts in the process of applying artificial intelligence in the arts and humanities.
