Anna Nickelson

Oregon State University

As humans, almost every single decision we make balances multiple constraints and goals, which are often in direct opposition with one another. How do we enable autonomous agents to make decisions when there are competing goals and no singular optimal option? How does this type of nebulous decision-making impact the agent’s behavior and the broader ethical questions that surround AI?

Anna’s research tries to tackle these questions; she focuses on Multi-Objective Optimization for autonomous decision making. Essentially, how can autonomous agents balance multiple competing goals and what are the broader implications of this type of decision making? Currently, her work focuses on multi-objective multiagent evolutionary learning frameworks. This research studies how multiple agents can collectively achieve multiple high-level objectives without minimal communication and no explicit coordination. In parallel, she is also working on projects relating to the ethical implications of artificial intelligence and multi-objective autonomous decision making.