In “Technology paternalism – wider implications of ubiquitous computing”, S. Spiekermann and F. Pallas deal with how people can maintain control in environments that are supposed to be totally automated. They coined the term “technology paternalism” to describe the situation where “people may be subdued to machines’ autonomous actions“. They take the example of car that beeps when you don’t fasten your seatblet and show how such situations meet the same criteria as the one that define paternalism:

“ the definition of Technology Paternalism extends the general notion of paternalism with respect to two aspects: one is that actions are being taken autonomously by machines. The other one is that by their coded rules, machines can become ‘absolute’ forces and therefore may not be overrulable any more.“

And discuss that in conjunction with Weiser’s notion of calm computing:

“If machines are controlled, then they are not calm any more. There is a clear disaccord between the concept of disappearing technologies and the attempt to remain in control. Control premises attention and visibility whilst Ubicomp environments are designed to be invisible and seamlessly adaptive. Can this dissonance really ever be resolved? “

And of course there is a part about who’s responsible of technology paternalism:

“Of course, this power does not lie in the hands of technology itself. Technology only follows rules implemented into it. Therefore, the question arises: who WILL be the real patrons behind Technology Paternalism if it were to become a reality? Who will decide about the rules, the ‘rights’ and ‘wrongs’ of every-day actions?
(…)
three groups as the potential patrons behind Technology Paternalism: engineers and marketers of Ubicomp technologies as well as regulators influencing application design. “

Why do I blog this? Some relevant issues regarding the notion of control in ubicomp. The authors finally come up with a series of recommendations. The one that strikes me as fundamental is the following: “there should be a general possibility to overrule ‘decisions’ made by technology and any exceptions from this should be considered very carefully“. The notion of exception is a crux issue that is often diminished by lots of engineers I talked to wrt to autonomous technologies such as “intelligent fridges” or location-based services. Exceptions breaks patterns and habits tracked by sensors, disrupt machine learning algorithm and are eventually impediments to prediction-based system that would send emergency messages to 911 because granny did not open her fridge for 2 weeks (because she unexpectedly decided to visit her grandson).

Is Context-Aware Computing Taking Control Away from the User? Three Levels of Interactivity Examined by Barkhuus & Dey is an interesting paper about the users’ experience of different degrees of automation in ubicomp. They investigated this through a user study of a context-aware application in which 3 levels of interactivity are defined:

“personalization, passive context-awareness and active context-awareness. Personalization is where applications let the user specify his own settings for how the application should behave in a given situation; passive context-awareness presents updated context or sensor information to the user but lets the user decide how to change the application behavior, where active context-awareness autonomously changes the application behavior according to the sensed information“

The results are intriguing (please see the details of the methodology in the paper):

“Our study found that users’ sense of control decreases when autonomy of the service increases, as suggested by previous research. We believed that personalization would be preferred and would be more accepted than both passive and active context-awareness, however, the results of our study do not support this. Instead we find that people prefer context-aware applications over personalization oriented ones.
(…)
participants felt they had less control in the context-aware groups but still preferred the context-aware approaches (…) The incurred cost due to loss of control can result in users turning off a service. While the participants initially liked many of the active context-aware services, they might become frustrated by their perceived lack of control and eventually turn the service off. (…) Our conclusion is that users are willing to accept a large degree of autonomy from applications as long as the application’s usefulness is greater than the cost of limited control.
“

Why do I blog this? This is close to the debate about automation that I described here. I am indeed interested in this differentiation between levels of interactivity and how people felt them.

Barkhuus, L. & Dey, A. (2003). Is Context-Aware Computing Taking Control Away from the User? Three Levels of Interactivity Examined, Proceedings of UBICOMP 2003, The 5th International Conference on Ubiquitous Computing, pp. 159-166. October 12-15, 2003.

My interest in “automation” and how technologies enables it through sensors and so-called computational “intelligence” has been attracted to a paper called “Sabbath Day Home Automation: “It’s Like Mixing Technology and Religion” by Allison Woodruff, Sally Augustin and Brooke Foucault. It basically presents a qualitative study of 20 American Orthodox Jewish families’ use of home automation for religious purposes. The two questions the authors address concern the relationship between home automation and religious practice, as well as the relationship between home automation and family life. Then they derive interesting implications from the results about this topic for a broader population.

The paper start by describing the different automation technologies that can be employed on the Sabbath (”The reasoning is that, although an Orthodox Jew should not do anything during the Sabbath that has impact on electrical devices, they can perform acts in advance of the Sabbath“). They found 3 categories: timers (rotary timers), X10 (a system for using household wiring to send digital data between devices) and high-end schedule systems (” rule-based program that uses the Jewish calendar and the families’ specified preferences to dynamically generate a schedule that interfaces with controllers (typically X10) installed by the system developer, as well as with other automation systems in the home, such as lighting, sprinkler, security, entertainment, and security systems.“)

IMO, the “meat” of this research study consists in the description of how automation system became interwoven with the family lives:

“the home automation system reflected and shaped the routines, expected behaviors, and social relations of family life – the social order of the home. (…) Home automation systems are particularly rich organizing systems because they can act “autonomously” to modify the physical environment, and because they are embodied in the objects and infrastructure of the home.
(…)
Automation as a Resource for Influencing Behavior: Automation provides cues as to what actions are expected of household occupants at what times, and it is therefore a resource for influencing behavior – for example, when kids are in the recreation room late at night, a light turning off sends a “message” that it is time to go to bed.
(…)
Interpretation of the Role of the Automation System Interpretations of the automation system and its actions were varied and complex. However, it was quite common for participants to attribute meaning to actions taken by the automation system, and sometimes to associate them with expected behavior. First, participants sometimes oriented to the automation system as an extension or proxy of the
person setting the schedule of the automation system. (…) Second, automation was strongly associated with
caretaking, anticipation, and guidance“

Why do I blog this? What is interesting here is the relationship between a desire for automation and technologies. The solutions described here become part of a ritual and a sign of a community affiliation. This is different form other studies about automation (in which there the desire for automation is more tight to curiosity, for early adopters for instance). The accounts provided here gives a critical overview of specific issues regarding control and automation. I also found intriguing the implications for design:

“traditional wisdom argues for a high degree of end-user control. Our findings prompt us to consider a richer set of options in the design space. Studies suggest that there are situations in which surrendering control offers significant psychological benefit. (…) some individuals might be better served by systems that present a wider range of options for autonomy, e.g. that certain individuals at certain times would benefit from experiences that give them a sense of another entity being in control.
(…)
Plainly, we are not arguing that people should not have any control of their devices. We are however proposing that giving up control can be beneficial or desirable in some situations, and that this is an interesting design space to explore. “

A good quote I’ve read in a french book about foresight (“Fabriquer le futur 2 : L’imaginaire au service de l’innovation” (Pierre Musso, Laurent Ponthou, Éric Seulliet)), translated by myself:

“One can systematically notice a paradox in the public attitude - namely consumers that area comfortable with technological devices. On one hand, one notice the ever-growing needs in terms of comfort, simplification of domestic tasks and distractions. This desire make attractive any propositions that delegate the tedious efforts to machines, that lower the accidents or diseases or that offer improved sensorial performances. On the other hand, consumers always jealously defend a number of principles that corresponds to the traditional representation of how their homeplace should be: autonomy and authority (…) Thus, any automatic device must present some limits in terms of decision autonomy” (Edouard le Maréchal, Tangenciels) “

Why do I blog this? I quite like this ambivalence between a desire for certain delegation and the refusal of automatic systems. That’s definitely one of the greatest challenge in design, which emerge from the problems caused by automation. Too often automation is generalized to lots of contexts in which things do not work properly.