Abstract
How does the visual design of digital platforms impact user behavior and the resulting environment? A body of work suggests that introducing social signals to content can increase both the inequality and unpredictability of its success, but has only been shown in the context of music listening. To further examine the effect of social influence on media popularity, we extend this research to the context of algorithmically-generated images by re-adapting Salganik et al's Music Lab experiment. On a digital platform where participants discover and curate AI-generated hybrid animals, we randomly assign both the knowledge of other participants' behavior and the visual presentation of the information. We successfully replicate the Music Lab's findings in the context of images, whereby social influence leads to an unpredictable winner-take-all market. However, we also find that social influence can lead to the emergence of local cultural trends that diverge from the status quo and are ultimately more diverse. We discuss the implications of these results for platform designers and animal conservation efforts.
Highlights
We investigate the role of social inuence and information design on ve outcomes: inequality, unpredictability, diversity, divergence and engagement
For both the list and cloud displays, we nd that worlds with social inuence exhibit more inequality than the worlds where participants made independent decisions (? = 0.002, preregistered), which is a direct replication of the Music Lab experiment
The circle packing algorithm placed most top voted ganimals near the center,. This explains why worlds in the social inuence cloud condition had the highest Gini coecient on average: social inuence induced users to click on top voted ganimals near the center, further increasing inequality. For both inequality and unpredictability, we observed the same pattern observed in the Music Lab, that social inuence can increase the inequality and unpredictability of the success of content
Summary
Theexplosionofinformationcontainedonmodernonlineplatformsrequiresuserstouseheuristics tobothecientlysearchthroughthisinformation,andtomakeinformeddecisions.Onesuch heuristicisthesocialsignalsofhowotherusershaveengagedwiththeplatform.Socialinuence occurswhenthedecisionsofauserareimpactedbythoseofotherusers[10],andhasbeenshown tobeakeydesigndimensionforcontextsasvariedashealthbehavior[9],politicalengagement [5] , collective behavior [41], online book purchasing [8], food ordering [30], and digital news engagement[46,62].Theubiquityofsocialinuencesuggestshowcrucialafactoritisforplatform designersseekingtojointlyoptimizeforthequalityanddiversityofcontentonline[28].Perhapsthemostinuentialstudyonhowsocialinuenceandinformationhierarchyimpact onlineplatformsisthatofSalganiketal.[54],informallydubbedthe“MusicLab”experiment.In thisstudy,theauthorscreatedan“articialculturalmarket,”whereparticipantscouldlistento anddownloadpreviouslyunknownsongs.Critically,someparticipantswereprovidedalayout whichdisplayedinformationaboutpreviousparticipant’schoices,whiletheothershadnosuchProc. Salganik et al found that introducing social inuence increased the inequality of song success, as dened by the number of times they were downloaded. This suggests a cascading “winner take all” phenomenon, whereby social inuence increased the availability of songs that were perceived as successful by past participants. They found that social inuence increased the unpredictability of success, as dened by the variation in a song’s success across the worlds in an experimental condition. From these two ndings, the authors infer that the underlying quality of a song only partly determines its nal success: social inuence causes a snowball eect that results in the emergence of local preferences
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have