In this paper we argue that conceptual modeling is well suited to define a convincing profile for the information systems discipline. Different from empirical research, conceptual modeling should not only represent existing domains. To fully exploit the potential offered by information technology, conceptual models should also represent constructions of future reality. While such an approach is promising substantial benefits, it is accompanied by severe epistemological challenges. Rational discourses are about the only chance to deal with these challenges - although they do not offer a convincing solution. However, without facing this problem, the field of conceptual modeling is nothing more than a playground for inventing artifacts.