Comment by rgambee

I'm curious what it would mean to align the interests of an AI agent with a natural ecosystem. Alignment to human values is still an unsolved problem. Alignment to an ecosystem has a lot of the same challenges as alignment to human values, like balancing diverse and often conflicting interests (what's good for a rabbit might be bad for a fox). But it has its own challenges, like lack of data.