it considers the potential hazards, but not the likely benefits of an innovation, shifting the burden of proof to an innovator to prove that its product will not cause harm, but not allowing that innovator to demonstrate that it might cause good, or might displace a technology that already causes harm.