The assignment of autonomy and culpability (too often conflated) is equally messy. Some might anticipate a future in which rival Cloud polities will be defined as much by the real-world violence of their external conflicts as by their internal strategies. The development of weapons for large-scale infrastructural warfare through complex viruses is well underway — spy versus spy, algorithm versus algorithm, City layer versus City layer, and Stack versus Stack — and attracts some of the very best computer scientists to the problem space. In many cases, the strategy is not only to break an enemy ’ s servos, it is to weaponize that enemy ’ s Users “ against him, ” especially when the User is unaware of having being weaponized. The zombified personal computer, taken over by malware acting like a parasitic fungus, is a basic example, but the capacity of unprogrammed Users to wreak havoc on their own platform is almost limitless. In such a case, is the User or the owner more responsible? He who initiates the column of death or he who owns its Stack? Surely intention is important, but not from the system ’ s perspective on its own security. There are only low-risk and high-risk Interfaces and the Users attached to them. One lesson of this is that just as the diffusion of “ responsibility ” into the system cannot lay the blame at the doorstep of the unconscious User position, it cannot be designed around the effable conscious will of the User either. Colonel Jack Ripper might tap the first domino in order to protect the purity of essence of his precious bodily fluids, but even his maniacal self-sovereignty depends on a vast already existing machinic apparatus that springs into action. The sacralization of individuated privacy through encryption might also seem to guarantee purity of essence and the sovereignty of the atomized User, but it anonymizes some of the key variables in the larger human-technological chain of association, activity, and accountability, also withdrawing them from reasonable adjudication. Instead of containing the impacts of an exchange within the closed circle of those transacting, it instead makes some agents invisible to the trace routing of algorithmic governance altogether, thereby automating the free-rider problem, as well as making everything that transaction externalizes (both benign and pernicious) invisible even to the Users that spawn them. Out of sight, out of mind. This kind of tactical anonymization can both overinflate the end User’ s sense of commanding authority over all interactions (an inflation that is both technological and psychological) and withdraws the User from forums of accountability for the partial costs of his partial agency. It achieves the former by promising the latter.
346
The Stack to Come