Aardvark employees would get the questions from beta test users and route them to users who were online and would have the answer to the question. This was done to test out the concept before the company spent the time and money to build it, said Damon Horowitz, co-founder of Aardvark, who spoke at Startup Lessons Learned, a conference in San Francisco on Friday.

“If people like this in super crappy form, then this is worth building, because they’ll like it even more,” Horowitz said of their initial idea.

At the same time it was testing a “fake” product powered by humans, the company started building the automated product to replace humans. While it used humans “behind the curtain,” it gained the benefit of learning from all the questions, including how to route the questions and the entire process with users.

This is a really good idea, as I’ve argued before on this blog and in a chapter for developers of mobile health interventions. What better way to (a) learn about how people will use and experience your service and (b) get training data for your machine learning system than to have humans-in-the-loop run the service?

My friend Chris Streeter wondered whether this was all done by Aardvark employees or whether workers on Amazon Mechanical Turk may have also been involved, especially in identifying the expertise of the early users of the service so that the employees could route the questions to the right place. I think this highlights how different parts of a service can draw on human and non-human intelligence in a variety of ways — via a micro-labor market, using skilled employees who will gain hands-on experience with customers, etc.

I also wonder what UIs the humans-in-the-loop used to accomplish this. It’d be great to get a peak. I’d expect that these were certainly rough around the edges, as was the Aardvark customer-facing UI.

Aardvark does a good job of being a quite sociable agent (e.g., when using it via instant messaging) that also gets out of the way of the human–human interaction between question askers and answers. I wonder how the language used by humans to coordinate and hand-off questions may have played into creating a positive para-social interaction with vark.

In his new book, How We Get Along, philosopher David Velleman puts it this way:

At one extreme, I have a way of interpreting myself, a way that I want you to interpret me, a way that I think you do interpret me, a way that I think you suspect me of wanting you to interpret me, a way that I think you suspect me of thinking you do interpret me, and so on, each of these interpretations being distinct from all the others, and all of them being somehow crammed into my self-conception. At the other extreme, there is just one interpretation of me, which is common property between us, in that we not only hold it but interpret one another as holding it, and so on. If my goal is understanding, then the latter interpretation is clearly preferable, because it is so much simpler while being equally adequate, fruitful, and so on. (Lecture 3)

That is, one way my self-views can be efficient representations is if they serve double duty as others’ views of me — if my self-views borrow from others’ views of me and if my models of others’ views of me likewise borrow from my self-views.

Sometimes this back and forth between my self-view and my understanding of how others’ view me can seem counter to self-interest. People behave in ways that confirm others’ expectations of them, even when these expectations are negative (Snyder & Swann, 1978, for a review see Snyder & Stukas, 1999). And people interact with other people in ways such that their self-views are not challenged by others’ views of them and their self-views can double as representations of the others’ views of them, even when this means taking other people as having negative views of them (Swann, 1981).

Self-verification and behavioral confirmation strategies

People use multiple different strategies for achieving a match between their self-views and others’ view of them. These strategies come in at different stages of social interaction.

Prior to and in anticipation of interaction, people seek and more thoroughly engage with information and people with self-views expected to be consistent with their self-views. For example, they spend more time reading statements about themselves that they expect to be consistent with their self-views — even if those particular self-views are negative.

During interaction, people behave in ways that elicit views of them from others that are consistent with their self-views. This is especially true when their self-views are being challenged, say because someone expresses a positive view of an aspect of a person who sees that aspect of themselves negatively. People can “go out of their way” to behave in ways that elicit negative self-views. On the other hand, people can change their self-views and their behavior to match the expectations of others; this primarily happens when a person’s view of a particular aspect of themselves is one they do not regard as certain.

After interaction, people better remember expressions of others’ views of them that are consistent with their own. They also can think about others’ views that were inconsistent in ways that construe them as non-conflicting. On the long term, people gravitate to others’ — including friends and spouses — who view them as they view themselves. Likewise, people seem to push away others who have different views of them.

Do people self-verify in interacting with computers?

Given that people engage in this array of self-verification strategies in interactions with other people, we might expect that they would do the same in interacting with computers, including mobile phones, on-screen agents, voices, and services.

One reason to think that people do self-verify in human–computer interaction is that people respond to computers in a myriad of social ways: people reciprocate with computers, take on computers as teammates, treat computer personalities like human personalities, etc. (for a review see Nass & Moon, 2000). So I expect that people use these same strategies when using interactive technologies — including personal computers, mobile phones, robots, cars, online services, etc.

While empirical research should be carried out to test this basic, well-motivated hypothesis, there is further excitement and importance to the broader implications of this idea and its connections to how people understand new technological systems.

When systems models users

Since the 1980s, it has been quite common for system designers to think about the mental models people have of systems — and how these models are shaped by factors both in and out of the designer’s control (Gentner & Stevens, 1983). A familiar goal has been to lead people to a mental model that “matches” a conceptual model developed by the designer and is approximately equivalent to a true system model as far as common inputs and outputs go.

Many interactive systems develop a representation of their users. So in order to have a good mental model of these systems, people must represent how the system views them. This involves many of the same trade-offs considered above.

These considerations point out some potential problems for such systems. Technologists sometimes talk about the ability to provide serendipitous discovery. Quantifying aspects of one’s own life — including social behavior (e.g., Kass, 2007) and health — is a current trend in research, product development, and DIY and self-experimentation. While sometimes this collected data is then analyzed by its subject (e.g., because the subject is a researcher or hacker who just wants to dig into the data), to the extend that this trend will go mainstream, it will require simplification by building and presenting readily understandable models and views of these systems’ users.

The use of self-verification strategies and behavioral confirmation when interacting with computer systems — not only with people — thus presents a challenge to the ability of such systems to find users who are truly open to self-discovery. I think many of these same ideas apply equally to context-aware services on mobile phones and services that models one’s social network (even if they don’t  present that model outright).

Social responses or more general confirmation bias

That people may self-verify with computers as well as people raises a further question about both self-verification theory and social responses to communication technologies theory (aka the “Media Equation”). We may wonder just how general these strategies and responses are: are these strategies and responses distinctively social?

Prior work on self-verification has left open the degree to which self-verification strategies are particular to self-views, rather than general to all relatively important and confident beliefs and attitudes. Likewise, it is unclear to what extent all experiences, rather than just social interaction (including reading statements written or selected by another person), that might challenge or confirm a self-view are subject to these self-verification strategies.

Inspired by Velleman’s description above, we can think that it is just that other’s views of us have an dangerous potential to result in an explosion of the complexity of the world we need to model (“I have a way of interpreting myself, a way that I want you to interpret me, a way that I think you do interpret me, a way that I think you suspect me of wanting you to interpret me, a way that I think you suspect me of thinking you do interpret me, and so on”). Thus, if other systems can prompt this same regress, then the same frugality with our cognitions should lead to self-verification and behavioral confirmation. This is a reminder that treating media like real life, including treating computers like people, is not clearly non-adaptive (contra Reeves & Nass, 1996) or maladaptive (contra Lee, 2004).

References

This name and description makes this sound quite like science fiction. In this post, I assimilate homuncular flexibility to the much more general phenomenon of distal attribution (Loomis, 1992; White, 1970). When I have a perceptual experience, I can just as well attribute that experience – and take it as being directed at or about – more proximal or distal phenomena. For example, I can attribute it to my sensory surface, or I can attribute it to a flower in the distance. White (1970) proposed that more distal attribution occurs when the afference (perception) is lawfully related to efference (action) on the proximal side of that distal entity. That is, if my action and perception are lawfully related on “my side” of that entity in the causal tree, then I will make attributions to that entity. Loomis (1992) adds the requirement that this lawful relationship be successfully modeled. This is close, but not quite right, for if I can make distal attributions even in the absence of an actual lawful relationship that I successfully model, my (perhaps inaccurate) modeling of a (perhaps non-existent) lawful relationship will do just fine.

Just as I attribute a sensory experience to a flower and not the air between me and the flower, so the blind man or the skilled hammer-user can attribute a sensory experience to the ground or the nail, rather than the handle of the cane or hammer. On consideration, I think we can see that these phenomena are very much what Lanier is talking about. When I learn to operate (and, not treated by Lanier, 2006, sense) my lobster-body, it is because I have modeled an efference–afference relationship, yielding a kind of transparency. This is a quite familiar sort of experience. It might still be a quite dangerous or exciting idea, but its examples are ubiquitous, not restricted to virtual reality labs.

Lanier paraphrases biologist Jim Boyer as counting this capability as a kind of evolutionary artifact – a spandrel in the jargon of evolutionary theory. But I think a much better just-so evolutionary story can be given: it is this capability – to make distal attributions to the limits of the efference–afference relationships we successfully model – that makes us able to use tools so effectively. At an even more basic and general level, it is this capability that makes it possible for us to communicate meaningfully: our utterances have their meaning in the context of triangulating with other people such that the content of what we are saying is related to the common cause of both of our perceptual experiences (Davidson, 1984).

References

These human–technology relationships develop and endure over time and through radical changes in the situation. In particular, mobile phones are near-constant companions. They take on roles of both medium for communication with other people and independent interaction partner through dynamic physical, social, and cultural environments and tasks. The global phenomenon of mobile phone use highlights both that relationships with people and technologies are influenced by variable context and that these devices are, in some ways, a constant in amidst these everyday changes.

Situational variation and attribution

Situational variation is important for how people understand and interact with mobile technology. This variation is an input to the processes by which people disentangle the internal (personal or device) and external (situational) causes of an social entity’s behavior (Fiedler et al. 1999; Forsterling 1992; Kelley 1967), so this situational variation contributes to the traits and states attributed to human and technological entities. Furthermore, situational variation influences the relationship and interaction in other ways. For example, we have recently carried out an experiment providing evidence that this situational variation itself (rather than the characteristics of the situations) influences memory, creativity, and self-disclosure to a mobile service; in particular, people disclose more in places they have previously disclosed to the service, than in  new places (Sukumaran et al. 2009).

Not only does the situation vary, but mobile technologies are increasingly responsive to the environments they share with their human interactants. A system’s systematic and purposive responsiveness to the environment means means that explaining its behavior is about more than distinguishing internal and external causes: people explain behavior by attributing reasons to the entity, which may trivially either refer to internal or external causes. For example, contrast “Jack bought the house because it was secluded” (external) with “Jack bought the house because he wanted privacy” (internal) (Ross 1977, p. 176). Much research in the social cognition and attribution theory traditions of psychology has failed to address this richness of people’s everyday explanations of other ’s behavior (Malle 2004; McClure 2002), but contemporary, interdisciplinary work is elaborating on theories and methods from philosophy and developmental psychology to this end (e.g., the contributions to Malle et al. 2001).

These two developments — the increasing role of situational variation in human-technology relationships and a new appreciation of the richness of everyday explanations of behavior — are important to consider together in designing new research in human-computer interaction, psychology, and communication. Here are three suggestions about directions to pursue in light of this:

Design systems that provide constancy and support through radical situational changes in both the social and physical environment. For example, we have created a system that uses the voices of participants in an upcoming event as audio primes during transition periods (Sohn et al. 2009). This can help ease the transition from a long corporate meeting to a chat with fellow parents at a child’s soccer game.

Design experimental manipulations and measure based on features of folk psychology –  the implicit theory or capabilities by which we attribute, e.g., beliefs, thoughts, and desires (propositional attitudes) to others (Dennett 1987) — identified by philosophers. For example, attributions propositional attitudes (e.g., beliefs) to an entity have the linguistic feature that one cannot substitute different terms that refer to the same object while maintaining the truth or appropriateness of the statement. This opacity in attributions of propositional attitudes is the subject of a large literature (e.g., following Quine 1953), but this  has not been used as a lens for much empirical work, except for some developmental psychology  (e.g., Apperly and Robinson 2003). Human-computer interaction research should use this opacity (and other underused features of folk psychology) in studies of how people think about systems.

Connect work on mental models of systems (e.g., Kempton 1986; Norman 1988) to theories of social cognition and folk psychology. I think we can expect much larger overlap in the process involved than in the current research literature: people use folk psychology to understand, predict, and explain technological systems — not just other people.

References

A big theme of the workshop this morning has been transitions among production and consumption — and the critical role of annotations and context-awareness in enabling many of the user experiences discussed. In many ways, this workshop took me back to thinking about mobile media sharing, which was at the center of a good deal of my previous work. At Yahoo! Research Berkeley we were informed by Marc Davis’s vision of enabling “the billions of daily media consumers to become daily media producers.” With ZoneTag we used context-awareness, sociality, and simplicity to influence people to create, annotate, and share photos from their mobile phones (Ahern et al. 2006, 2007).

Enabling and encouraging these behaviors (for all media types) remains a major goal for designers of participatory media; and this was explicit at several points throughout the workshop (e.g., in Teppo Raisanen’s broad presentation on persuasive technology). This morning there was discussion about the technical requirements for consuming, capturing, and sending media. Cases that traditionally seem to strictly structure and separate production and consumption may be (1) in need of revision and increased flexibility or (2) actually already involve production and consumption together through existing tools. Media production to be part of a two-way communication, it must be consumed, whether by peers or the traditional producers.

As an example of the first case, Sarah Lewis (Stanford) highlighted the importance of making distance learning experiences reciprocal, rather than enforcing an asymmetry in what media types can be shared by different participants. In a past distance learning situation focused on the African ecosystem, it was frustrating that video was only shared from the participants at Stanford to participants at African colleges — leaving the latter to respond only via text. A prototype system, Mobltz, she and her colleagues have built is designed to change this, supporting the creation of channels of media from multiple people (which also reminded me of Kyte.tv).

As an example of the second case, Timo Koskinenen (Nokia) presented a trial of mobile media capture tools for professional journalists. In this case the work flow of what is, in the end, a media production practice, involves also consumption in the form of review of one’s own materials and other journalists, as they edit, consider what new media to capture.

Throughout the sessions themselves and conversations with participants during breaks and lunch, having good annotations continued to come up as a requirement for many of the services discussed. While I think our ZoneTag work (and the free suggested tags Web service API it provides) made a good contribution in this area, as has a wide array of other work (e.g., von Ahn & Dabbish 2004, licensed in Google Image Labeler), there is still a lot of progress to make, especially in bringing this work to market and making it something that further services can build on.

References

1. Blogging something at this level of roughness is still new for me…

Over at Captology Notebook, the blog of the Stanford Persuasive Technology Lab, Enrique Allen considers features of Facebook that influence users to update their status. Among other things, he highlights how Facebook lowers barriers to updating by giving users a clear sense of something they can right (“What are you doing right now?”).

I’d like to add another part of the interface for consideration: the box in the left box of the home page that shows your current status update with the most recent updates of your friends.

This visual association of my status and the most recent status updates of my friends seems to do at least a couple things:

Influencing the frequency of updates. In this example, my status was updated a few days ago. On the other hand, the status updates from my friends were each updated under an hour ago. This juxtaposes my stale status with the fresh updates of my peers. This can prompt comparison between their frequency of updates and mine, encouraging me to update.

The choice of the most recent updates by my Facebook friends amplifies this effect. Through automatic application of the availability heuristic, this can make me overestimate how recently my friends have updated their status (and thus the frequency of status updates). For example, the Facebook friend who updated their status three minutes ago might have not updated to three weeks prior. Or many of my Facebook friends may not frequently update their status messages, but I only see (and thus have most available to mind) the most recent. This is social influence through enabling and encouraging biased social comparison with — in a sense — an imagined group of peers modeled on those with the most recent performances of the target behavior (i.e., updating status).

Influencing the content of updates. In his original post, Enrique mentions how Facebook ensures that users have the ability to update their status by giving them a question that they can answer. Similarly, this box also gives users examples from their peers to draw on.

Of course, this can all run up against trouble. If I have few Facebook friends, none of them update their status much, or those who do update their status are not well liked by me, this comparison may fail to achieve increased updates.

Consider this interface in comparison to one that either

• showed recent status updates by your closest Facebook friends, or
• showed recent status updates and the associated average period for updates of your Facebook friends that most frequently update their status.

[Update: While the screenshot above is from the "new version" of Facebook, since I captured it they have apparently removed other people's updates from this box on the home page, as Sasha pointed out in the comments. I'm not sure why they would do this, but here are couple ideas:

• make lower items in this sidebar (right column) more visable on the home page -- including the ad there
• emphasize the filter buttons at the top of the news feed (left column) as the means to seeing status updates.

Given the analysis in the original post, we can consider whether this change is worth it: does this decrease status updates? I wonder if Facebook did a A-B test of this: my money would be on this significantly reducing status updates from the home page, especially from users with friends who do update their status.]

The Lifeworld and ready-to-hand

Heidegger was a student of Edmund Husserl, and Heidegger’s Being and Time was to be dedicated to Husserl.¹ There is really no question of the huge influence of Husserl on Heidegger.

My major introduction to both Husserl and Heidegger was from Prof. Dagfinn Føllesdal. Føllesdal (1979) details the relationship between their philosophies. He argues for the value of seeing much of Heidegger’s philosophy “as a translation of Husserl’s”:

The key to this puzzle, and also, I think, the key to understanding what goes on in Heidegger’s philosophy, is that Heidegger’s philosophy is basically isomorphic to that of Husserl. Where Husserl speaks of the ego, Heidegger speaks of Dasein, where Husserl speaks of the noema, Heidegger speaks of the structure of Dasein’s Being-in-the-world and so on. Husserl also observed this. Several places in his copy of Being and Time Husserl wrote in the margin that Heidegger was just translating Husserl’s phenomenology into another terminology. Thus, for example, on page 13 Husserl wrote: “Heidegger transposes or transforms the constitutive phenomenological clarification of all realms of entities and universals, the total region World into the anthropological. The problematic is translation, to the ego corresponds Dasein etc. Thereby everything becomes deep-soundingly unclear, and philosophically it loses its value.” Similarly, on page 62, Husserl remarks: “What is said here is my own theory, but without a deeper justification.” (p. 369, my emphasis)

Heidegger and his terms have certainly been more popular and in wider use since then.

Føllesdal also highlights where the two philosophers diverge.² In particular, Heidegger gives a central role to the role of the body and action in constituting the world. While in his publications Husserl stuck to a focus on how perception constitutes the Lifeworld, Heidegger uses many examples from action.³ Our action in the world, including our skillfulness in action constitutes those objects we interact with for us.

Heidegger contrasts two modes of being (in addition to our own mode — being-in-the-world): present-at-hand and ready-to-hand (or alternatively, the occurant and the available (Dreyfus 1990)). The former is the mode of being consideration of an object as a physical thing present to us — or occurant, and Heidegger argues it constitutes the narrow focus of previous philosophical explorations of being. The latter is the stuff of every skilled action — available for action: the object becomes equipment, which can often be transparent in action, such that it becomes an extension of our body.

J.J. Gibson expresses this view in his proposal of an ecological psychology (in which perception and action are closely linked):

When in use, a tool is a sort of extension of the hand, almost an attachment to it or a part of the user’s own body, and thus is no longer a part of the environment of the user. [...] This capacity to attach something to the body suggests that the boundary between the animal and the environment is not fixed at the surface of the skin but can shift. More generally it suggests that the absolute duality of “objective” and “subjective” is false. When we consider the affordances of things, we escape this philosophical dichotomy. (1979, p. 41)

While there may be troubles ahead for this view, I think the passage captures well something we all can understand: when we use scissors, we feel the paper cutting; and when a blind person uses a cane to feel in front of them, they can directly perceive the layout of the surface in front of them.

Transparency, abstraction, opacity, intentionality

Research and design in HCI has sought at times to achieve this transparency, sometimes by drawing on our rich knowledge of and skill with the ordinary physical and social world. Metaphor in HCI (e.g., the desktop metaphor) can be seen as one widespread attempt at this (cf. Blackwell 2006). This kind of transparency does not throw abstraction out of the picture. Rather the two go hand-in-hand: the specific physical properties of the present-at-hand are abstracted away, with quickly perceived affordances for action in their place.

But other kinds of abstraction are in play in HCI as well. Interactive technologies can function as social actors and agents– with particular cues eliciting social responses that are normally applied to other people (Nass and Moon 2000, Fogg 2002). One kind of social response, not yet as widely considered in the HCI literature, is assuming the intentional stance — explanation in terms of beliefs, desires, hopes, fears, etc. — towards the system. This is a powerful, flexible, and easy predictive and explanatory strategy often also called folk psychology (Dennett 1987), which may be a tacit theory or a means of simulating other minds. We can explain other people based on what they believe and desire.

But we can also do the same for other things. To use one of Dennett’s classic examples, we can do the same for a thermostat: why did it turn the heat on? It wanted to keep the house at some level of warmth, it believed that it was becoming colder than desired, and it believed that it could make it warmer by turning on the heat. While in the case of the thermostat, this strategy doesn’t hide much complexity (we could explain it with other strategies without much trouble), it can be hugely useful when the system in question is complex or otherwise opaque to other kinds of description (e.g., it is a black box).

We might think then that perceived complexity and opacity should both be cues for adopting the intentional stance. But if the previous research on social responses to computers (not to mention the broader literature on heuristics and mindlessness) has taught us anything, it is that made objects such as computers can evoke unexpected responses through other simplier cues. Some big remaining questions that I hope to take up in future posts and research:

• What are these cues, both features of the system and situational factors?
• How can designers influence people to interpret and explain systems using folk psychology?
• What are the advantages and disadvantages of evoking the intentional stance in users?
• How should we measure the use of the intentional stance?
• How is assuming the intentional stance towards a thing different (or the same) as it having being-in-the-world as its mode of being?

References

1. But Husserl was Jewish, and Heidegger was himself a member of the Nazi party, so this did not happen in the first printing.
2. Dreyfus (1990) is an alternative view that takes the divergence as quite radical; he sees Føllesdal as hugely underestimating the originality of Heidegger’s thought. Instead Dreyfus characterizes Husserl as formulating so clearly the Cartesian worldview that Heidegger recognized its failings and was thus able to radically and successfully critique it.
3. It is worth noting that Husserl actually wrote about this as well, but in manuscripts, which Heidegger read years before writing Being and Time.

1. A general shift from examining non-discretionary use to discretionary use
2. How much easier it is to find (and not train) study participants unfamiliar with a system than experts (especially with a system that is only a prototype)
3. The push from practitioners in the direction, especially with the advent of the Web, where new users just show up at your site, often deep-linked

This focus sometimes comes in for criticism, especially when #2 is taken as a main cause of the choice.

On the other hand, some research threads in HCI continue to focus on expert use. As I’ve been reading a lot of research on both human performance modeling and situated & embodied approaches to HCI, it has been interesting to note that both instead have (comparatively) a much bigger focus on the performance and experience of expert and skilled use.

Grudin’s “Three Faces of Human-Computer Interaction” does a good job of explaining the human performance modeling (HPM) side of this. HPM owes a lot to human factors historically, and while The Psychology of Human-Computer Interaction successfully brought engineering-oriented cognitive psychology to the field, it was human factors, said Stuart Card, “that we were trying to improve” (Grudin 2005, p. 7). And the focus of human factors, which arose from maximizing productivity in industrial settings like factories, has been non-discretionary use. Fundamentally, it is hard for HPM to exist without a focus on expert use because many of the differences — and thus research contributions through new interaction techniques — can only be identified and are only important for use by experts or at least trained users. Grudin notes:

A leading modeler discouraged publication of a 1984 study of a repetitive task that showed people preferred a pleasant but slower interaction technique—a result significant for discretionary use, but not for modeling aimed at maximizing performance.

Situated action and embodied interaction approaches to HCI, which Harrison, Tatar, and Senger (2007) have called the “third paradigm of HCI”, are a bit different story. While HPM research, like a good amount in traditional cognitive science generally, contributes to science and design by assimilating people to information processors with actuators, situated and embodied interaction research borrows a fundamental concern of ethnomethodology, focusing on how people actively make behaviors intelligible by assimilating them to social and rational action.

There are at least three ways this motivates the study of skilled and expert users:

1. Along with this research topic comes a methodological concern for studying behavior in context with the people who really do it. For example, to study publishing systems and technology, the existing practices of people working in such a setting of interest are of critical importance.
2. These approaches emphasize the skills we all have and the value of drawing on them for design. For example, Dourish (2001) emphasizes the skills with which we all navigate the physical and social world as a resource for design. This is not unrelated to the first way.
3. These approaches, like and through their relationships to the participatory design movement, have a political, social, and ethical interest in empowering those who will be impacted by technology, especially when otherwise its design — and the decision to adopt it — would be out of their control. Non-discretionary use in institutions is the paradigm prompting situation for this.

I don’t have a broad conclusion to make. Rather, I just find it of note and interesting that these two very different threads in HCI research stand out from much other work as similar in this regard. Some of my current research is connecting these two threads, so expect more on their relationship.

References
Dourish, P. (2001). Where the Action Is: The Foundations of Embodied Interaction. MIT Press.
Grudin, J. (2005). Three Faces of Human-Computer Interaction. IEEE Ann. Hist. Comput. 27, 4 (Oct. 2005), 46-62.
Harrison, S., Tatar, D., and Senger, P. (2007). The Three Paradigms of HCI. Extended Abstracts CHI 2007.

Wizard of Oz techniques are well suited to prototyping mobile services, especially those using mobile messaging (SMS, MMS, voice messaging). When participants send a request, a wizard reads or listens to it and chooses the appropriate response, or just creates it on-the-fly. Since all user actions in mobile messaging are discrete messages and (depending on the application) the user can often tolerate a short delay, a few part-time wizards, such as you and a colleague, can manage a short field trial. As you’ll see, this can be used for purposes beyond many traditional uses of a Wizard of Oz.

Probing photo consumption needs with realistic motivations
One use for this technique in designing a mobile messaging service is a bit like a diary study. In designing an online and mobile photography service, we wanted to better understand what photos people wanted to view and what prompted these desires.¹ Instead of just making diary entries, participants actually made voice requests to the system for photos – and received a mobile message with photos fitting the request in return. We didn’t need to first build a robust system that could do this; a few of us served as wizards, listening to the request, doing a couple manual searches, and choosing which photos to return on demand. Though this can be done with a normal voice call, we used a mobile client application that also recorded contextual information not available via a normal voice call (e.g. location), so that participants could make context-aware requests as they saw fit (e.g. “I want too see photos of this park”)

In this case, we didn’t plan to (specifically) create a voice-based photo search system; instead, like a diary study, this technique served as a probe to understand what we should build. As a probe it provided realistic motivations for submitting requests, as the request would actually be fulfilled. This design research, in additional to other interviews and a usability study, informed our creation of Zurfer, a mobile application that supports exploring and conversing around personalized, location-aware channels of photos.
It is great if the Wizard of Oz prototype is quite similar to what you later build, but it can yield valuable insights even if not. Sometimes it is precisely these insights that can lead you to substantially change your design.

This study design can apply in designing many mobile services. As in our photos study, participants can be interviewed about the trigger for the requests (why did they want that media or information) and how satisfied they were with the (human-created) responses.²