AI Paper Atlas

Research map / Large language models / AI and society

Position: AI Should Not Be An Imitation Game: Centaur Evaluations

Andreas Haupt, Erik Brynjolfsson

ICML 2025 · Large language models › AI and society · approach: Social-science view of AI evaluation

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Position paper arguing ML should increasingly use centaur evaluations, where humans and AI jointly solve tasks, to promote human augmentation and capture human-centered desiderata.

Problem

Current ML evaluations test systems in isolation from humans, leading to saturated benchmarks, poor proxies for human-centered desiderata, and a development bias toward human replacement rather than augmentation.

Approach

Defines centaur benchmarks with three components (human selection, interface, scoring) and argues for jointly human-AI evaluations, including a production-function framing in which the marginal value of human time measures human augmentation.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
combine a human with an AI system for codinghuman-AI coding team · functional HTTP server codingspeedImproved
provide machine-learning-powered coding recommendations to coderscoders using Visual Studio Code · coding in Visual Studio CodespeedImproved
equip workers with a machine-learning system for supporthuman workers · work tasks studied by Cui et al. (2024)productivityImproved
combine human decisions with machine-learning predictionshuman-AI systems · several socially relevant decision settingsperformanceHurt
use human judges rather than counterfactual algorithm-only decisionsjudges and algorithms · judicial decisionsperformanceHurt
combine radiologists with screening algorithmsradiologists and screening algorithms · radiology screeningscreening performanceHurt
combine humans with AI rather than use algorithms alonehuman-AI systems · fairness evaluationsfairnessHurt
use human-AI collaboration instead of human-only or AI-only systemshuman-AI systems · studies published in 2020 to 2023human augmentationNo clear effect

Limitations

Experimental setup

Builds on: Turing Test Turing Trap (Brynjolfsson 2022) learning-to-defer

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy