English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Trade-offs between accuracy and efficiency pervade law, public health, andother non-computing domains, which have developed policies to guide how tobalance the two in conditions of uncertainty. While computer science alsocommonly studies accuracy-efficiency trade-offs, their policy implicationsremain poorly examined. Drawing on risk assessment practices in the US, weargue that, since examining these trade-offs has been useful for guidinggovernance in other domains, we need to similarly reckon with these trade-offsin governing computer systems. We focus our analysis on distributed machinelearning systems. Understanding the policy implications in this area isparticularly urgent because such systems, which include autonomous vehicles,tend to be high-stakes and safety-critical. We 1) describe how the trade-offtakes shape for these systems, 2) highlight gaps between existing US riskassessment standards and what these systems require to be properly assessed,and 3) make specific calls to action to facilitate accountability whenhypothetical risks concerning the accuracy-efficiency trade-off become realizedas accidents in the real world. We close by discussing how such accountabilitymechanisms encourage more just, transparent governance aligned with publicvalues.

Accuracy-Efficiency Trade-Offs and Accountability in Distributed ML Systems