Rowan Haen
@rowanhaen.bsky.social
PhD candidate working on human reasoning at Utrecht University
(3/3)
Moreover, the results suggest that task complexity is a boundary condition for the conflict detection effect: the more complex the tasks, the lower the performance accuracy, the number of conflict detectors, and the conflict detection effect size.
Moreover, the results suggest that task complexity is a boundary condition for the conflict detection effect: the more complex the tasks, the lower the performance accuracy, the number of conflict detectors, and the conflict detection effect size.
January 19, 2026 at 8:59 PM
(3/3)
Moreover, the results suggest that task complexity is a boundary condition for the conflict detection effect: the more complex the tasks, the lower the performance accuracy, the number of conflict detectors, and the conflict detection effect size.
Moreover, the results suggest that task complexity is a boundary condition for the conflict detection effect: the more complex the tasks, the lower the performance accuracy, the number of conflict detectors, and the conflict detection effect size.
(2/3)
We found evidence of conflict detection across three probabilistic reasoning tasks (i.e. base rate, inverse fallacy, and Bayesian inference tasks). Although the tasks rely on an important shared mindware component, we found limited evidence for consistency in conflict detection across tasks.
We found evidence of conflict detection across three probabilistic reasoning tasks (i.e. base rate, inverse fallacy, and Bayesian inference tasks). Although the tasks rely on an important shared mindware component, we found limited evidence for consistency in conflict detection across tasks.
January 19, 2026 at 8:59 PM
(2/3)
We found evidence of conflict detection across three probabilistic reasoning tasks (i.e. base rate, inverse fallacy, and Bayesian inference tasks). Although the tasks rely on an important shared mindware component, we found limited evidence for consistency in conflict detection across tasks.
We found evidence of conflict detection across three probabilistic reasoning tasks (i.e. base rate, inverse fallacy, and Bayesian inference tasks). Although the tasks rely on an important shared mindware component, we found limited evidence for consistency in conflict detection across tasks.
Isn’t this a similar issue with autonomous vehicles? Even if autonomous vehicles reduce overall harm (or are expected to do so), we tend to hold autonomous systems to a much higher standard than humans, which suggests we may also judge chatbots more harshly in this regard.
January 9, 2026 at 1:46 PM
Isn’t this a similar issue with autonomous vehicles? Even if autonomous vehicles reduce overall harm (or are expected to do so), we tend to hold autonomous systems to a much higher standard than humans, which suggests we may also judge chatbots more harshly in this regard.