Surviving Moore'S Law: Security, Ai, And Last Mover Advantage presented at 15th USENIX Security Symposium 2006

by Paul Kocher,

Tags: Security Others

Summary : Most computer security research focuses on the pursuit of a "binary" ideal of security, such as proofs of correctness or cryptographic strength. Meanwhile, security for actual systems and networks increasingly relies on patches, rather than demonstrably strong designs. The cause: advances in complexity are causing greater harm to computer security than benefits.
Coping with this problem requires strategies designed specifically for the needs of complex systems. This talk will explore approaches that have worked and others that have failed spectacularly, while considering the long-term prospects for security.