0.1 Intelligence Meets Reality

Perceive Environment

Agents, computational foundations of AI (1)

Choose Actions

Agents, computational foundations of AI (2)

Performance Measures

Agents optimize, computational foundations of AI (3)

Rational Agent

defined as an agent that selects actions expected to maximize its performance measure based on its percept history and available knowledge

Objective, Morally Desirable

it does not state that the _ itself is _ _

Actions, Socially Acceptable

it does not guarantee that the chosen _ are _ _

Considers, Fairness, Privacy, human welfare

it does not ensure that the agent _ _ _ or _

Rational Agent

simply optimizes the objective that human specify

Unreasonable, Harmful

if the objective function is incomplete or poorly designed, the agent may exhibit behavior that appears _ or _ despite behaving exactly as instructed

AI Alignment

how AI systems can be designed so that their goals, decisions, and behaviors remain consistent with human intentions and values

AI Alignment

attempts to ensure that AI systems remain beneficial even as they become increasingly capable and autonomous

Human Compatible Objectives

one of the greatest AI challenges is not making machines more intelligent, but making them pursue _ _ _