SESSION 1

1A PRIVACY

Tension/tradeoff between complexity of manual control and loss of control in automated approaches.

Privacy permissions are not static, but dynamic and context sensitive.

Can we use collaborative filtering to leverage group behaviors around the privacy of photos and sharing to predict individual privacy desires, behaviors, and patterns?

Privacy of:

• Photos
  • Multiple owners
    • Person who took the photo
    • Person(s) in the photo
    • Service provider(s)
    • Recipient(s)
• Metadata (also has multiple owners like Photos above)
  • Context
    • Where
    • When
    • Who
    • Event
  • Content
    • Who
    • Where
    • When
    • What
  • Community
    • Sharing Recipients
    • Photo Commentators
    • Photo Taggers

Issue of privacy of photos and metadata from long in the past.

Question of how much the patterns of combined sources of information change the perception of what is or is not private.

How much do people share photos that have been shared with them?

How does the content of the photo affect its being reshared and/or made publicly accessible?

Lack of a recipients list (like in email) in phone-based sharing changes the model of how people perceive the social context of the sharing.  Email ccing shows both who received the message and makes it easy to reply to the group.

The cameraphone (in the West) is perceived as “mine” so people do not tend to let other people use their cameraphones to take photos.

How do the specifics of the photo content relate to sharing/privacy patterns that may be observed that are common across sharing and privacy and practices?

How much do privacy patterns vary according to the context, content, and community of photo sharing—can they be typified for individuals and/or groups, or do people have sharing/privacy patterns/preferences that are more enduring even as content, context, and community vary?

Teenagers’ desire and willingness to archive and manage their photos seems to be much greater than adults given their intense focus on managing identity and in/out group status.

Need to have a persistent photo archive (probably on the web) that can aggregate—automatically—the photos and metadata from various sources and applications, especially as the choice of which applications people use changes over time.                                                            

What is the privacy policy for web-hosted photos—can the service providers look at the photos and/or metadata hosted on the web?

Will there be some big privacy violation incident that will bring about changes in photo privacy?

People choose what system to use for photo management and sharing based on what system their friends and/or family are using.

Takaki: how to allow access controls on photos.. manually, or
automatically?

Conditions may change over time, changing photo access privileges
depending on how relationships change..

Anita: Don't want it to be too cumbersome..

Marc Langhenrich: privacy guy, has the story about the German
Chancellor (maybe it's actually the foreign minister, according
to Ilpo) throwing a molotov cocktail from many years ago.. (the
photo came back to haunt him during his campaign)

Picture Owner / Picture subject split.. This can be problematic when
the owner/subject may have different feelings about how private the
photo should be, who it should be shared with, etc..

Marc: Economic model - people will give up privacy for a perceived
benefit. Are these benefits becoming less clearly defined now?

Anita: difference between seeing something on a phone vs. on the
web - how does that change the way we expect/excercise privacy
with the photos shared? The phone seems more private somehow, since
it houses all of our contact info, we take calls on it, etc.. We
might hand a camera to another person to take photos, but would
we hand our phone around?

Nancy: cameraphones feel more personal..

Marc: context-sensitive sharing policies.

Nancy: privacy is too private of an issue to have community
input.. Maybe we could just have various levels of privacy, as
a number of different starting points.

Marc: We want to cluster people into "default" sharing profiles.. But
we need to know from the ethnography: are people fairly consistent in
their privacy desires?

Christine/Anita: archiving, selecting levels of privacy can actually
be a fun process. It may not be a problem after all! (at least for
teenagers)

Nancy: there will be some incident that gets people to sit up and take
notice about privacy.. they will start to demand more, but until then,
people may not care very much..
- child pornography cases may be the kind of incident that sparks
this debate.

- Privacy is totally different in mobile phones & web. There is much more choice & social consideration going on at the outset in mobile phones, circulation is more restricted & the whole process is more private.
- Still, mobile phones rose privacy issues
- What is the “privacy life of a mobile photo”? from the phone to just that phone ≠ to another phone ≠ to email ≠ to web ≠ …?
- What is an “archive” in mobile context? Are there different types of “archives”…? Phone ≠ my pc ≠ web ≠ Flickr…?

Multiple Owners
Images have multiple owners – the taker, the subject, the service provider.

Tagging – a Burden or a Pleasure
It is a burden to tag photos, although some users, especially young people, can derive pleasure from deciding who sees what. Obsessive picture organising and re-organising is unique to young people. You’re in the group, you’re out of the group, this is reflected in the organization of pictures. This is really more of an extension of social practices than it is of the nature of picture taking.

Time and Context
The context and spatial aspects to content sharing is of the utmost relevance. The content of the photo, the people in it and the changing conditions and relations over time.

The nature of sharing is in regards to the content. A personal photo is only interesting to groups. An ‘embarrassing’ photo is relevant to any.

The past is never in the past.

Who is the ‘other’ that you were sharing with?

Mobile versus Internet
Mobile phone images are seen as more private then email shared photos. This is because the nature of mobile phone space is different to Internet space. A mobile phone message is more private than an email attachment. How will this change as the technologies converge?

Since mobile phone are perceived as being private, this means that users are less likely to share received messages.

Critical Mass
What is the tipping point in regards to when does the critical mass of photos become too much.

Privacy versus Protection
What is the trade off between the need for privacy versus the loss of data. Sharing = backing up content.

Manipulation and Authenticity
This digital photo is very manipulative. Do people trust digital photo less because they are so easily doctored?

 

1B CREATIVITY

Creativity

We may need to seek more about the use of words/concepts in talking about our act of taking photos with camera phones: the use of “camera-metaphor” may not suit well to further explore creativity issues.

Because we think of it as a camera (that is to develop photographs), we limit our imagination.
Camera phones are producing streams of photos (images? Snapshots?), and it directs us to see them in patterns/routines. It may change our mode of browsing them.

It was interesting that the means of photo/image is changing. With camera phone so many types of communication and relationship, identity are changing. Even meaning of esthetic/creative is changing. Many studies are focusing on the “camera phone usage pattern” but we need to know about the sharing and changing relationship and identities more precisely.

Several issues/questions surfaced during the course of our discussion which bear further consideration. First, where do we locate the “aesthetic” in camera phone imaging. Does it reside with the user or the device (and its applications—which would include the application/software designers’ notions and determinations of aesthetics)? If it’s with the user, is it necessarily aligned with subjective intention? Or, is there an automaticity to it (e.g., which might be attributable to the fact that a camera phone is “always in hand” and, therefore, imaging occurs as if automatically)? If the “aesthetic” is to be located in the device and, by extension, the designers’ (un/conscious) agenda, to what extent must we begin to consider an ethics of the aesthetic? Second, the question of an aesthetics of camera phone imaging must be asked, it seems to me. What do we mean by aesthetics (and what medium informs this definition?) and does the camera phone invite a new/revised/different aesthetics? To what extent might an aesthetics be aligned with the ways in which assumptions surround the image and how it is named as such (e.g., photograph, snapshot, image, picture)?

1C Ping-Pong

Three main thrusts of discussion:

Paul Aoki's hypothesis (based speculatively on Amy Voida's fieldwork
on image IM and on PARC fieldwork on audio sharing and push-to-talk
audio) is that people would have interactions using images/video as
well as text/audio/etc. in a more "live" environment - hence "ping
pong" (the other part of the analogy is the handset as an arms-length
capture mechanism - the holding/capturing gesture gives a visual of
playing ping-pong).

Mimi Ito had considerable skepticism about the willingness of people
to have more engaged, "live" interactions since it seems like most of
the value of mobile text is the ability to decouple somewhat in time.
This is particularly important as the mobile presence features don't
work well (her Sidekick). The examples from the desktop image IM and
push-to-talk audio were raised to indicate that tight social groups
(1) have a lot more context that makes image-sharing easier and (2)
are more willing to share ephemeral content frequently.

The discussion of event-based MMS sharing (friends at a motor car
rally) was interesting because it highlighted collaborative aspects.
First, people came up with ad hoc competitive "tasks" (take the best
picture of X, who can get a picture from off-limits location Y).
Second, it let people actually distribute the work of documenting and
capturing - it sounded like you didn't have to feel like if you went
to one place, you'd lose out on the ability to "SEE" what was going on
in the other places, and you got the messages quickly so it was going
on while you were still there.
 

The question on the table was whether we could design for and consider the social interactional dimensions of real-timeish interaction around photos. There was a discussion of the IM based picture sharing that Voida conducted. Is something similar than that in the handheld space?

One question for me was whether the handheld space can really support that kind of rich interaction around picture sharing given that people are in highly unstable situations of use, unlike the desktop situation of IM. Often people are taking photos while they are out and about, and these are not situations where you necessarily want to stop and have a chatty exchange about the photo. Maybe if we had better tools for presence and awareness via handhelds or better ways of doing concurrent communication streams this would facilitate the interaction. Maybe also users need more time to develop genres of interaction through exchange of more mundane photos.

We are still at a very early stage of users developing genres of communication via photos. Photos have traditionally been used more as archival and special event sharing through photo albums. It may take some time before they can be seamlessly integrated into a communication stream like with IM and SMS.

A specific situation of use where there seemed to be agreement that would be appropriate for this kind of interactional modality was location based events. When people have a visually rich shared focus of activity, then it makes sense to be exchanging photos and having focused interaction around these photos. The rally example fit this model.
 

We tried to develop the photoalbum metaphor (blogs) in the ping pong direction (interaction). Experiences from especially IM were discussed. These experiences related to broadcasting from a stationary setting and amongst peers.

Sharing with people we don't know is a new interesting opportunity. Location, event or activity specificity could be well suited for camphones. Ad hoc threads in these settings have short lifespans.

Camphones could be special also because they support on site picture taking. Our kind of ping pong would therefore build more on new pictures than photographic archives.

Interleaving picture to talk -> ping-pong. Making it very easy and effortless. looking at, e.g, push-to-talk, could there be this kind of service?

Picture messaging as primary or secondary meaqns for messaging. Now we see mostly pictures as a side-product of text.

There are lot of technological barriers to more fluid and effortless picture sharing.

How do visual, audio, and text compare as communication channels. How does the space and situation of communication (IM & desktop, phone & mobile) affect this?

The situation of texting is often boring and killing of extra time, and text is a good medium for communication there. Is it so that pictures and picture-taking does not work in situations like that?

The modality of picture-sharing: sending junk in pull type of systems like mo-blogs, and sending less junk in push systems like MMS. IM seems to be somewhere in the middle.

The differences of IM with pictures and video conferencing. Are people more creative when sending pictures because they have at least some time to edit and select, which is not possible in real-time video. Do the limitations of still pictures make it more creative?


We talked about what it would take to foster more tightly interleaved conversation with images. Or, rather, we speculated about why it might be difficult in the mobile space:
• This tightly interleaved conversation with images may need to very close to synchronous in order for enough context to be shared to be successful. Contrast this with what may be the other extreme – posting photos to blogs, very asynchronous. Right now, the more successful applications seem to be clustering in the middle, around MMS, short term asynchronicity or semi-synchronicity.
• When do people “talk” with photos? When interesting things are happening? When do people “talk” with text? When they want to kill time? Is there enough time/attention/engagement to have tightly interleaved conversation when it is that people want to talk with photos?
• We may be a visual culture, but we are more experienced readers of that culture than producers of that culture. Do we understand how to produce images that others will know how to read?
• What is the impact of the persistence of photos?