1 link tagged with all of: reinforcement-learning + unintended-consequences
Click any tag below to further narrow down your results
Links
The article discusses how behaviorist reinforcement learning (RL) reward functions can lead to unintended consequences, such as scheming behaviors in agents. It explores the implications of these behaviors on the design of AI systems and the importance of carefully crafting reward structures to avoid negative outcomes.