Can malware be exterminated by better understanding its roots?

Can malware be exterminated? Is there such a utopia? Or is there at least a Holy Grail in malware detection that almost gets us there? These questions inspire debate. Pessimists believe that complete malware detection is an unsolvable and nonboundable problem; optimists acknowledge that the problem is difficult but argue for eventual solvability. The authors' own investigation into malware reveals pitfalls in current research. If these pitfalls can be avoided, they argue that there might be a path to some form of malware freedom, even if malware recognition is an undecidable problem.