No CrossRef data available.
Published online by Cambridge University Press: 13 May 2024
Identifying the conditions of proxy treadmilling is crucial for determining whether reliable signals can persist over time. I present a framework that maps evolutionary models of reliable signals according to their assumptions regarding the effects of Goodhart's law. This framework can explain the contrasting outcomes of different modelling approaches, and identify in which models proxy treadmilling is expected to occur.
Target article
Dead rats, dopamine, performance metrics, and peacock tails: Proxy failure is an inherent risk in goal-oriented systems
Related commentaries (20)
An updated perspective on teleonomy
Animal welfare science, performance metrics, and proxy failure
Behavioral proxies compete by the time courses of their rewards, including endogenous rewards
Changing the incentive structure of social media may reduce online proxy failure and proliferation of negativity
Dynamic diversity is the answer to proxy failure
Genies, lawyers, and smart-asses: Extending proxy failures to intentional misunderstandings
It's the biology, stupid! Proxy failures in economic decision making
Navigating proxy failures in education: Learning from human and animal play
On abstract goals’ perverse effects on proxies: The dynamics of unattainability
Proxies, heuristics, and goal alignment
Proxy failure and poor measurement practices in psychological science
Proxy failure as a feature of adaptive control systems
Proxy failure in academia: More than just another example
Proxy failure in social policies as one of the main causes of persistent sexism and racism
Proxy failures in practice: Examples from the sociology of science
Reductionism and proxy failure: From neuroscience to target-based drug discovery
Regulator and agent sophistication as an explanation-generating engine for proxy failure dynamics
Subjective and objective corruption of intuition and rational choice
The cost of success or failure for proxy signals in ecological problems
The determinants of proxy treadmilling in evolutionary models of reliable signals
Author response
Teleonomy, legibility, and diversity: Do we need more “proxynomics”?