A lot of research is going into considering how to design AI systems aligned with human values and goals, or at least that do not actively harm them. This is not an easy task, as humans themselves often have conflicting or unclear values and goals, and translating them into computer code is fraught with challenges. Moreover, there is no guarantee that an AI system will keep its original values and goals as it evolves and learns from its environment. Adding to the complexity is that no one really knows whether AGI is possible, or whether alignment is a real concern. Forecasting when and if AI
...more
