Peter Vamplew

Message

I am an Associate Professor in IT at Federation University (based in Ballarat, Australia). My primary research focus is on multiobjective reinforcement learning, and in recent years I have been examining how methods from that field can contribute to addressing the issue of AI alignment.

Scalar reward is not enough for aligned AGI

This post was authored by Peter Vamplew and Cameron Foale (Federation University), and Richard Dazeley (Deakin University) Introduction Recently some of the most well-known researchers in reinforcement learning Silver, Singh, Precup and Sutton published a paper entitled Reward is Enough, which proposes the reward-is-enough hypothesis: “Intelligence, and its associated abilities,...

Jan 17, 202223

This post was authored by Peter Vamplew and Cameron Foale (Federation University), and Richard Dazeley (Deakin University)

Introduction

Recently some of the most well-known researchers in reinforcement learning Silver, Singh, Precup and Sutton published a paper entitled Reward is Enough, which proposes the reward-is-enough hypothesis: “Intelligence, and its associated abilities, can be understood as subserving the maximisation of reward by an agent acting in its environment". Essentially, they argue that the overarching goal of maximising reward is sufficient to explain all aspects of natural and artificial intelligences.

Of specific interest to this forum is the contention that suitably powerful methods based on maximisation of a scalar reward (as in conventional reinforcement learning) provide a suitable... (read 3001 more words →)

LESSWRONG
LW

LESSWRONG
LW

Peter Vamplew

Peter Vamplew

Scalar reward is not enough for aligned AGI

Peter Vamplew

Peter Vamplew

Scalar reward is not enough for aligned AGI