Home › Claude › aifoundations › what_is_ai_intro › The 'right thing' for an AI agent to do is defin…
The 'right thing' for an AI agent to do is defined by
AA universal moral law from physics
BThe agent's internal logical rules
CMajority vote among multiple agents
DThe objective that we provide
Answer & Solution
Correct answer: D. The objective that we provide
AIMA: rationality is defined relative to a performance measure that WE supply. Hence aligning that measure with what we truly want is the central long-run challenge of the field.
Related questions
The rational-agent view subsumes the laws-of-thought view becauseWhy is cognitive science conceptually distinct from AI per AIMA?AIMA's 'standard model' across AI, control, OR, statistics and economics isWhy does AIMA argue we must move beyond the standard model of fixed objectives?The value alignment problem, as AIMA frames it, isLimited rationality, per AIMA, refers toThe 'laws of thought' approach in AI traces back toWhy has the rational-agent view prevailed in AI?