Affiliation:
1. Michigan Technological University, Houghton, MI, USA
Abstract
In the past decade, consumer adoption of commercial semi-autonomous vehicles has increased, and along with it user concerns about the shortcomings of these systems, especially regarding safety. Users often turn to social media forums to discuss these shortcomings, find workarounds, and confirm their experience is common. We suggest that these forums may provide some of the best training for users to understand the limitations of AI, as they are not controlled by the vendor who has a vested interest in hiding the limitations of their systems. In two laboratory experiments, we examined how information from Tesla FSD forums impact participants' ability to detect and predict hazardous driving situations in simulated scenarios. Drivers who received the training were better at anticipating and recognizing dangerous driving conditions, suggesting that exposure to user-generated explanations of the shortcomings of the system may in fact improve safety and acceptance of the systems.
Subject
General Medicine,General Chemistry