Speaker
Description
Ensuring the replicability of scientific experiments is crucial for advancing knowledge and establishing the credibility of research findings. However, the increasing use of Artificial Intelligence (AI) to create responsive and more natural experimental conditions based on participants' verbal and behavioral responses presents new challenges for experiment replicability.
Researchers can use AI algorithms to adjust experimental parameters in real-time based on participants' responses and contextual factors, such as environmental changes, participant demographics, and unexpected variables. This approach introduces variability that cannot be monitored with traditional methods. It is not sufficient to control AI behaviour by keeping fixed algorithm parameters. Instead, a way must be found to consider the different participant’s behavioural topographies that lead to the same functionality. This means considering the different response modes with the same function.
In conclusion, this talk will introduce and discuss the main aspects of AI-adapted experimental conditions, proposing potential solutions from a theoretical and philosophical perspective.
