Control & networks

Architectures of Control in Design, Blog

To some extent, the desire of companies to control what consumers do with their products has parallels with attempts at price discrimination in industries such as freight transportation, and, especially, telecommunications.
Andrew Odlyzko of the University of Minnesota’s Digital Technology Center points out that telecommunications companies are currently able to achieve revenues of $3,000 per megabyte of data sent through SMS, yet the same data sent through cable TV would yield only $0.00012 per megabyte [60]. The internet–where, effectively, all data are priced the same–lacks these architectures of control, and this:

“…helps explain the push in the telecommunications industry for new network architectures that would provide service providers greater control of what customers do, and would deviate from the ‘stupid network’ model of the Internet.” [60]

Regulation and control of users’ behaviour in other telecommunications networks can also be a commercial necessity, especially where bandwidth is considered to be at a premium–for example, wireless grids, “challenging environments in which users’ strategic behaviours are crucial to system performance.” [61]
Lee McKnight,William Lehr and James Howison have reviewed technical, social, legal and economic methods to regulate behaviour in wireless grids (paralleling Lessig’s four regulators), with the technical methods most closely corresponding to architectures of control as examined here–as they put it:

“Appropriate behaviour can be ‘hardwired’ into the network through hardware and software design… The key is to define open interfaces that provide sufficient assurance as to the functionality that will be supported to allow interoperability without dictating detailed implementation rules that might limit innovation.” [61]

On a similar issue, ‘internet appliances’–the hardware through which a user (or a device) connects to the internet for a dedicated “particularity of purpose”–come into the picture as products for enacting architectures of control on behalf of telecommunications companies or ISPs: since the architecture of the internet itself is difficult to change, control may be put into the components which face the user.
Sharon Gillett, John Wroclawski, David Clark and William Lehr again (all MIT) have examined a range of internet appliances, the architectures of control built into them, and threat posed to “internet innovation” by the devices; their conclusion is that because of the enormous number of existing PCs, largely free of restrictive control, connected to the internet, any new, restrictive device will be at an immediate disadvantage–even if simply because users will be dissatisfied by these devices’ lack of forwards compatibility:

“Truly fixed function appliances… are unlikely to place a serious drag on innovation, because they are only likely to succeed in the marketplace if they are inexpensive, frequently replaced devices [e.g. phones].” [62]

Nevertheless, they accept that devices where the function can be remotely changed throughautomated software updates (such as the TiVo) are more of a risk–but note consumers’ lack of enthusiasm:

“Consumers do seem to be quite aware of, and unenthused by, the extent to which high switching costs would lock them into particular devices and services. This kind of lock-in may be a policy issue in its own right, especially if coupled with other products, services, or practices that raise antitrust concerns.” [62]

Although not expressly considered by Gillett et al, trusted computing would seem to fall squarely into this category, and, assuming that the majority of existing PCs are ultimately replaced by ‘trusted’ computers, the threat suddenly gains considerable momentum.


Previous: What things regulate? | Next: Everyday things & persuasive technology