Human Learning about AI (with Bnaya Dreyfuss)
Abstract: We study how humans form expectations about the performance of artificial intelligence (AI) and consequences for AI adoption. Our main hypothesis is that people project human-relevant task features onto AI. People then over-infer from AI failures on human-easy tasks, and from AI successes on human-difficult tasks. Lab experiments provide strong evidence for projection of human difficulty onto AI, predictably distorting subjects’ expectations. Resulting adoption can be sub-optimal, as failing human-easy tasks need not imply poor overall performance in the case of AI. A field experiment with an AI giving parenting advice shows evidence for projection of human textual similarity. Users strongly infer from answers that are equally uninformative but less humanly-similar to expected answers, significantly reducing trust and engagement. Results suggest AI “anthropomorphism” can backfire by increasing projection and de-aligning human expectations and AI performance.
Media Coverage:
Harvard Horizons Symposium talk:
Signaling Universalism (reject and resubmit at Journal of the European Economic Association)
Abstract: Recent research has studied heterogeneity in universalism vs. in-group favoritism to explain various economic and political behaviors. This paper documents that displayed universalism is significantly affected by social signaling concerns resulting from the anticipation of future economic interactions. In an online experiment, a decision maker divides money between an in-group and an out-group member, and I vary both the existence and identity - in-group or out-group - of a third-party audience that will subsequently play a cooperative game with the decision maker. Consistent with a simple model of social signaling, I find that people act substantially more universalist in presence of an out-group audience, as they try to match what they believe to be the audience’s preference. Publicly revealed universalism might therefore be significantly distorted, which garbles the true correlation between universalism and political or moral views.
GenAI and Social Media Content (with Michael Challis, Mateusz Stalinski, and Adrian Segura)
Gradoz, J., & Raux, R. (2021). Trolling in the Deep : Managing Transgressive Content on Online Platforms as a Commons. In Erwin Dekker and Pavel Kuchar (eds), Governing Markets as Knowledge Commons. Cambridge : Cambridge University Press, 217-237.