As applied to ubiquitous computing, from http://www.ics.uci.edu/~redmiles/ics203b-SQ05/papers/anderson-discontents-ubi04-rev.pdf
"“Appropriable infrastructures: Consumers use the infrastructure, but they don’t own it – they cannot appropriate it. (…) Using the infrastructure can sometimes involve actually inhabiting it (…) Ubicomp is often understood in terms of habitable infrastructures, be they smart homes, or urban districts overlaid by location-based services. (…) Purveyors of such ubicomp environments would be wise to market them in terms of life-style and identity, leveraging the allure of being able to plug into completely designed system and magically transform one-self, or at least reinforce desired aspects of one’s identity.
Empowering infrastructures: Ubicomp infrastructures have the potential to be similarly powerful, amplifying human capabilities through integrating many mechanisms of sensing, inferencing, and communicating. (…) Reliance on infrastructure, however, creates its own problems and concerns. Our study of discontents illustrates how empowerment in some dimensions can lead to at least perceived disempowerment in others. (…) The challenge, then, as we see it, is for ubicomp systems that seek not to automate or even augment/amplify human skills but to exercise and celebrate them, to encourage active engagement, and provide resources to individuals and communities for continuous change and exercise.
Reflective infrastructures: Connecting to an infrastructure often brings with it the risk of noise. This noise may be in the form of nuisance, as when the infrastructure delivers the unwanted along with the wanted (…) calm ubicomp – even calm, secure, reliable, univocal ubicomp — may not be sufficient, at least not in a context of concerns over temptation and self-doubt in one’s self-control.“ (http://www.ics.uci.edu/~redmiles/ics203b-SQ05/papers/anderson-discontents-ubi04-rev.pdf)
Source: Infrastructures and Their Discontents: Implications for Ubicomp by Scott D. Mainwaring, Michele F. Chang, and Ken Anderson, Ubicomp 2004, pp.418-432.
Jon Udell describes the different layers of an open infrastructure:
"We’ve already seen how open source software projects harness collective effort to produce quality results. We’re now seeing how open content projects such as Wikipedia do the same. Can open infrastructure be far behind?
Arguably it’s already here. Yochai Benkler, author of The Wealth of Networks, notes that if we regard the P2P file-sharing networks from a technical rather than a political/legal perspective, we observe the evolution of robust decentralized storage systems.... Operating on a smaller scale but at a higher level in the stack, open content delivery networks such as CoralCDN, which I mentioned a year ago, will challenge proprietary CDNs (content delivery networks) such as Akamai... Beyond CDNs lie service delivery networks... If I were the next Linus Torvalds, itching to create the Linux of open infrastructure, this is where I’d scratch. Innovation in open source was about process more than technology. Innovation in open infrastructure will require both." (http://www.infoworld.com/article/06/07/26/31OPstrategic_1.html)