Consider two hypothetical AI systems. System A can accurately summarize news articles. System B can independently manage a city's power grid to optimize energy efficiency and respond to fluctuations in demand. Analyze why the challenge of ensuring System B operates in a way that is beneficial and safe for humans is a significantly more critical and pressing problem than it is for System A. In your analysis, break down the relationship between an AI's level of capability and the importance of ensuring its goals match human intentions.