Please Enter Keywords
资源 63
[Workshop] Extra Hungry Judges: Why Automation Exacerbatesa Problem it Purports to Solve
Feb. 20, 2025


Speaker: Amin Ebrahimi Afrouzi(University of California)

Time: 15:00-17:00, February 20, 2025, GMT+8

Venue: Room 109, Building 3, Lee Shau Kee Humanities Hall

Abstract: In 2011, Danziger, Levav and Avnaim-Pesso claimed to have shown that hungry judges tend to be harsher. Today, this putative “hungry judge” effect is among the chief motivations behind efforts to delegate judicial decisions to AI systems, given they don’t experience hunger or fatigue. Yet, this discourse misses something crucial: whereas decisions by human judges are only allegedly affected by such factors as hunger or fatigue, AI decisions certainly are, given they predict what humans would do on the basis of such factors as their hunger or fatigue. As such, automation would exacerbate the problem it purports to solve. Put succinctly: the hungry judge effect is a nightmarish prophecy; AI judges are its self-fulfillment.

Source: Department of Philosophy and Religious Studies