264 CATALYZING INQUIRY
this is for each detector to have a randomly generated permutation rule, according to which all data
path triples are permuted before being matched against the detector. This effectively changes the struc-
ture of the self set for each detector, with the result that different detectors will be subject to different
holes. Consequently, where one detector fails to detect a nonself triple, another may succeed.
Multirepresentation was particularly effective at reducing the number of holes when the nonself pat-
terns were similar to self patterns. To deal with this problem, the bits in a given triplet of connection
triplets were randomly permuted before presentation to detectors, just as the specific MHC molecules
that are operating to bring pathogens to the surface are probabilistically determined (with respect to an
averaging over the population).
8.2.5.6 Some Possible Difficulties with an Immunological Approach
Although these analogies have appeal, it remains to be seen how far they can be pushed. Given that
the immune system is a very complex entity whose operation is not fully understood, a bottom-up
development of a computer security system based on the immune system is not possible today. The
human immune system has evolved to its present state due to many evolutionary accidents as well as the
constraints imposed by biology and chemistry—much of which is likely to be artifactual and mostly
irrelevant to the underlying principles that the system embodies and also to the design of a computer
security system. Further, the immune system is oriented toward problems of survival. By contrast, com-
puter security is traditionally concerned with confidentiality, accountability, and trustworthiness—and
the relevance of immunological processes to confidentiality and accountability is entirely unclear today.
8.2.6 Amorphous Computing
An area of research known as amorphous computing seeks to understand how to obtain “coherent
behavior from the cooperation of large numbers of unreliable parts that are interconnected in unknown,
irregular, and time-varying ways.”^53 This work, inspired by observations of cooperative and self-
organizing biological phenomena, seeks to identify the engineering principles that can be used to
observe, control, organize, and exploit the behavior of cooperating multitudes for human purposes such
as the design of engineered artifacts.
An individual entity in a collection of cooperating multitudes has the following characteristics:
- It is inexpensive, in the sense that it is easy to create large numbers of them. For all practical
purposes, each entity is identical to every other one. - It is locally guided or programmed. That is, the guidance or programming is carried by the entity
“on-board” rather than being resident elsewhere in the overall system. As a consequence of fabrication,
the guidance or programming aboard any given entity is identical to that aboard every other entity. - It communicates with nearby entities, but in a stochastic manner without the need for precise
interconnections and testing. Note also that the ability to function in a stochastically connective environ-
ment implies that the overall macrosystem is robust in the face of damaged or nonoperational compo-
nents. Furthermore, by eliminating the need for precision interconnections, these entities can reduce the
enormous costs usually associated with interconnection in traditional forms of assembly, costs that are
generally higher than those associated with individual elements. - It interacts with its environment locally, so that the entity is directly knowledgeable about some
aspect of its immediate environment but not about anything more global. To the extent that an indi-
vidual entity gains global knowledge about the environment, it is as the result of a self-organizing
process that develops such information and transmits it to all entities in the system. Similarly, any on-
board effectors affect only the immediate environment.