Who’s Laughing NAO? Examining Perceptions of Failure in a Humorous Robot Partner

on YouTube ↗10m

" summary

Study pilots 50 jokes then runs a main experiment with a 32-person dyadic lab sample and 256 bystanders on MTurk (total n=288) using a SoftBank NAO controlled in Choregraphe with speech from Amazon Polly. Robots either succeed or fail on a game and failing robots attempt recovery with affiliative, aggressive, self-enhancing, self-defeating, or no-humor lines. Dyads rate the self-defeating robot higher as a teammate while bystanders prefer affiliative and successful robots; experienced participants find jokes funnier and show less variation in teammate ratings.

AI-written summary · about

" chapters (17)

1. 0:00Collaborative Robotics Lab
2. 0:27Background - Interactive Robots
3. 0:52Background - Failure Recovery Tactics
4. 1:24Background - Humor in Human-Human Interactions
5. 1:47Research Questions
6. 2:28Approach
7. 2:58Method - The Four Humor Types
8. 3:35Method - The Robot
9. 3:50Pilot Study
10. 4:47Study 1: Dyadic Interaction
11. 5:39The iSpy Recordings
12. 6:26Study 2: Bystander Interaction
13. 6:50Humor and Perspective
14. 7:18Humor Type and Perspective
15. 7:55Humor and Experience
16. 8:26Humor Type and Experience
17. 8:59Discussion and Contributions