What if the famous P vs NP problem isn’t just about algorithms but about the observers trying to solve them? Research suggests computational difficulty depends on observer limits.