Here you can explore the raw data we collected from our surveys on the delegability of different tasks to AI. The raw data is also available on github.
Each participant was presented with one of 100 tasks, and evaluated the task under components of our framework. Finally, they selected their preferred level of machine assistance for the task: the delegability label. Amazon Mechanical Turk was used to collect the survey responses.
We collected 5 responses per task. The data from all responses is available below. You may expand to view individual responses, sort tasks by responses per components, or highlight different levels of task delegability.
Note: Click on a table heading to sort by that component; click again to reverse sort.