Wearable Diaries Project

Project in a Nutshell
I’m working to produce a series of multimedia diaries that take advantage of wearable technology. The core of the project is the creation of an app for Google Glass that automatically interviews the wearer throughout the day by displaying questions and recording the answers. The app will also occasionally grab video footage as b-roll. The idea is that these video profiles will provide a unique view into people’s lives by capturing moments that might not be otherwise documented and present stories through a person’s eyes.

Progress so Far
I teamed up with Scott Greenwald, a doctoral student here at the Media Lab, and we (mostly Scott) built a prototype app that recorded 10 seconds of video every few minutes (I experimented with different intervals — usually setting it for between 5 and 10 minutes). He coded the app in Wearscript, a system designed by Greenwald and some colleagues to make it easy to quickly prototype Glass apps. The result was buggy, but it let us test the concept.

Primavera’s Biohack Project
Using the app, Primavera wore Glass during three different days as she worked on an art project she is doing called Tree of Knowledge, which involves bioluminescent algae. The idea was to document the project and her vision of it using the POV perspective. Here’s the edited version, with a voice-over interview we recorded later:


I think this video was somewhat successful, but the biggest problem was we missed key moments of action. After running tests using this storytelling approach with the initial app, we learned that taking video every few minutes is too random, and we believed we could find a better way to decide when to turn on recording. I proposed linking the app to the wearer’s Google Calendar. Then we could try to set the app to prompt users with questions a few minutes before each calendar event (while they were likely on their way) and in the middle of events to hopefully get a representative sample of daily activities. Not everyone uses Google Calendar, of course, but the idea would be that I would sit down with the subject the day before they were going to record and help them fill in a Google Calendar with what they expected to do on the day of recording — that way we’re using Google Calendar to custom-program the Wearable Diaries app for each story subject.

We didn’t have time to actually build the prototype app that used this approach. Instead, I simulated it by just asking a fellow student (thanks Leslie!) to manually record moments throughout her day using this rough approach. I also texted her a few questions and reminders throughout the day (using a cell phone rather than Glass itself) to try some prompts that we might program into the app when we do build it.

Leslie’s Wearable Diary
Leslie wore Glass for a day and recorded about 75 short clips (most of them 10 seconds each but some of them longer), for a total of about 17 minutes of footage. I sent her 9 texts throughout the day, roughly one per hour, with prompts such as “Record the next conversation you have,” or “grab 30 seconds of whatever you see.” Leslie did most of the recording herself, though, making sure to get a little video from each activity. I edited the video down to three and a half minutes, with most clips running about 7 seconds each (so that the style is like a series of Twitter Vine videos strung together). Here’s the result:


Leslie was a very good sport, but she reported challenges to wearing Google Glass. “It kind of felt like a chain around my neck,” she said. “It really felt like a collar. It creates this barrier between you and the world.” On the plus side, she said it did give her a different perspective on her day, and she thought the approach could be used for “creating empathy” for someone’s point of view. As she put it, “someone who has a real cause and they want the world to see things through their eyes.”

Important Lessons
1. Many people can’t — or won’t — wear Google Glass.
Several people refused to participate in this project. For instance, I met a street artist last year who I thought would be an ideal subject. I e-mailed him an invitation and made my best case for why he might want to participate. No reply. Then last week, I happened to run into him on the street near my apartment and asked him directly. He declined, saying that “it goes against everything I believe in as an artist,” and that it felt like I was trying to attach a tracking device to him. I told him I respected his decision and I didn’t push it. Two other people turned me down as well in protest of the approach.

So I decided to ask students in the class. Stephen was willing, but he wears eyeglasses, and the Google Glass would not fit comfortably over his specs. While Google now sells its device custom fit to regular eyeglasses, this is too expensive an approach to loan to a subject for a day to record a story.

2. Taking video footage at random is too invasive.
The original plan was to have the camera on Google Glass automatically kick on at various times throughout the day. Sources I’ve talked to about the idea have been most put off by this loss of control, even though I assured them that nothing would be published or shown to anyone else without their permission. As a result, one important function I now plan to add is an opt out before any video recording. In other words, when the app turns on, it will give the user the choice whether or not to record before it begins. That way the source can always opt out of a given prompt.

3. Google Glass too often becomes the story.
Perhaps this will fade in time, but if you wear Google Glass, people will stare, or ask you to try them, or start talking about their views on the technology, or all of the above. That makes it a challenge to try to record a typical day in the life — since on a typical day most of us do not wear a computer and camera and screen on our face. I anticipated that this would be a challenge, but it turned out to be a greater issue than I realized.

Next Steps

I think the approach of recording guided video diaries from a person’s POV perspective is still a promising idea. But Google Glass in its current form may not be the best tool to do it. It’s possible that a similar story could have been shot using a GoPro or other device. Or perhaps I just haven’t found the right story or subject yet. I did apply for a Knight Prototype Grant, so I’m eager to hear suggestions from this group in case I’m able to try to move this project forward.

Final Project Proposal – Wearable Diaries

Note: This is maybe more detailed and involved than needed, so apologies for the length of the post. But I’m submitting a similar version of this proposal for a Knight Prototype Fund grant, so I’d be grateful for any feedback or suggestions.

What is your project?
We are producing a series of multimedia diaries that take advantage of wearable technology. The core of the project is the creation of an app for Google Glass that automatically records 10 seconds of video every few minutes and automatically interviews the subject throughout the day by displaying questions and recording the answers.

For the initial phase of the project, we will choose participants with particularly compelling circumstances and loan them Google Glass, and a producer/editor will work closely with each subject to coach them through the process, explain the privacy implications, and obtain permissions where appropriate. The resulting Web videos will be short – just 2 or 3 minutes each – to give viewers a peek into the day in the life of another person, from their first-person perspective.

In a later phase, we hope to add a system that can automatically assemble the clips into a draft piece and let users do simple editing to cut or reorder the clips. That way, the wearable diaries can be created without the help of a human editor. The app will be free to encourage as many people as possible to participate.

Who is the audience for the project? How will they be impacted?
Our approach is modeled loosely on the Radio Diaries project, a non-profit effort that gives people audio recorders to document their lives (though we have no affiliation with that project). Pieces from the Radio Diaries project reached a wide audience through broadcasts on NPR stations and were also distributed online and were a critical and popular success.

We are reaching out to traditional news outlets (newspaper and magazine Web sites) for possible partnerships to publish pieces made through the Wearable Diaries project. We will also post the pieces on our own Web site, with the hope that those who watch one of the diaries will be curious to see others in the series.

The stories are simple in their structure, but we believe their impact can be profound. By documenting the lives of many types of people in a short and sharable format, showing the similarities and differences in a diverse group of people, we hope to promote empathy and understanding of difference – as well as simply telling compelling stories.

What has come before?
This project can be situated in a long history of “life-logging” efforts, which have been proposed since the earliest days of wearable computers. A 2006 Media Lab project called InSense, for instance, let users make personal multimedia stories with a bulky camera and computer strapped onto a user’s chest. (http://hd.media.mit.edu/tech-reports/TR-599.pdf) Our effort takes advtange of the latest technology, which is smaller and more discrete. It also stresses a journalistic approach to storytelling, and the involvement of a human editor to shape the final piece.

What assumptions will you test?
One key premise of the project is that wearable technology such as Google Glass can help journalists tell stories in new ways.

Specifically, we hope that wearable technology might solve two problems faced by journalists filming documentary pieces about interesting people. One, subjects often act more self-consciously when they are followed around with a camera, detracting from the authenticity of the story. Secondly, the reporter is often not with the subject of a story at key moments, leaving gaps in the final piece. We see Google Glass (and potentially other wearable devices), as a way to follow story subjects with a kind of robotic reporting assistant that will occasionally shoot video and even ask them questions about what they’re seeing.

Who is working on the project?
Jeff and Primavera are working together on this project for the class. We’re also working with Scott Greenwald, a PhD student in the MIT Media Lab. He works in the lab’s Fluid Interfaces Group and is a developer of WearScript, which allows rapid app development for Google Glass with Javascript.

What have you made so far?
We have created a rough prototype of the app using WearScript, a system designed to rapidly prototype Glass apps to test them. We have done a couple of initial tests and created a Web site outlining the project: http://wearablediaries.org/ Jeff also presented the idea at a Google Glass hackathon at the MIT Media Lab: http://bit.ly/weartalk

Taking the project further will require creating a more robust Glass app and purchasing Google Glass that are fitted onto less-conspicuous frames than the original design.

News as Experience In-Class Activity – Twitter

Our group decided to do a write-in candidate and explore the experience of consuming news on Twitter. Since Twitter is so formless (or multi-formed), we first talked about how we typically access Twitter and what we use it for. There were four people in our group, and half of us primarily went to Twitter through laptops, and the other half leaned more to access via cell phone. Some used Hootsuite, some used Tweetdeck, and some used Twitter’s home page.

Then there’s the question is what is “news” on a Tweet. In some cases, the content of a news story fits within the 140 characters of a Tweet itself. More frequently, though, the Tweet is a link to a news story along with some comment about it by the person Tweeting. In that way, reading news on Twitter is like entering a story via the comment section of an article.

All of us follow both individual reporters we like plus the institutional feeds of news organizations (like @nytimes). None of us read all of the Tweets by the people we follow. Instead, we occasionally look at the flow of info into our personal Twitter accounts as if looking out a window.

Here were our brief answers to some questions posed on the assignment:

1. What kinds of values are embedded in this news experience?
Twitter favors “efficiency” of language, conversational, and humor
Twitter’s new Web design (pictured below), released this week, seems designed to attract more general users – to make Twitter look more like Facebook, and therefore more familiar.


2. What is this experience’s “theory of the user”? Who do they imagine you are? Does the experience also have a “theory of change”?
Twitter seems to assume users want to be part of a community, want to do networking or want to be part of a conversation (or all of the above).

3. What is this experience’s end goal? Virality & eyeballs? Deep listening? Exposé for action?
Social sharing. instant communication, breaking news

4. How are you empowered through this experience? Disempowered?
The barrier to entry is low, so it appears to create a more casual, level playing field. It’s possible for anyone to address public Tweets to anyone, no matter how famous. Sure, that celebrity probably won’t read the Tweet, but the system makes users feel empowered.

5. What kinds of stories is this method good for? bad for? underutilized for?
conversation starters

6. What form could you mash up with this to make new product for news?

Twitter wants to be the second screen
the global backchannel

Posted in All

Data Story: Why People Take Free Online Courses (MOOCs)

Millions of people have signed up for Massive Open Online Courses, known as MOOCs. Early studies show that the majority of those who have signed up already have a college degree, and most do not opt to pay for a certificate to prove they passed the class. Put simply, they’re not looking to get college credit in any way. So I’m curious to dig deeper into what motivates these online “students.”

I am late to post because I’ve been digging around for a killer data set on this. I’ve made requests to HarvardX and to some researchers who have a large MOOC dataset, but so far no one has been willing to share their raw numbers. But HarvardX has published some demographic and survey data (not much). My sense, though, is that their data does not answer the question very well (most MOOC surveys only offer a few multiple choices on motivation).

So for the assignment, I’m focusing on playing around with fuzzier “data” – the student postings to forums in a MOOC. In many MOOCs, students post short introductions in the forums at the beginning of the term, usually saying why they are taking the course. I’ll analyze the intro discussion postings in one MOOC and group them into broad categories (my categories won’t capture everything, but there are definitely clear patterns in the responses).

My plan is to pick an astronomy course on edX that just started. https://courses.edx.org/courses/ANUx/ANU-ASTRO1x/1T2014/discussion/forum/i4x-edx-templates-course-Empty/threads/533080e801772bb02e00087f
There are only about 200 intro posts, so it should be do-able in the short time frame.

I plan to pull out one student post that is the best example of each category I create. So the interface will be a simple pie chart with the percentages of each reason for taking a MOOC, but then when you click on a specific group/color, you’ll be taken to that person’s intro post so you viewers can “meet” them.

I’m certainly open to suggestions on tools, critique, etc.

Thoughts on ‘Climate Audit’; Why We Need ‘Fact Fight Club’

How do you fact check a blog like Climate Audit? The site details what the authors see as inconsistencies and exaggerations in the work of climate scientists, so they see themselves as the fact checkers. Yet in many cases, the site simply reprints private e-mails and quotes from climate scientists in which they are revealing the messiness of the scientific process, and suggests that this messiness is proof that the scientists are wrong about their conclusions on climate change.

For example, in one post, a climate auditor posts an e-mail from a scientist and writes: “Not sure what this email is about but it doesn’t sound very good.”

The site is full of details, charts, and graphs. It feels like proof of something. And the site details every time a climate auditor has their FOIA requests declined or redacted, suggesting that such secrecy is in itself proof that the scientists are wrong and hare hiding their true findings.

One thing is very clear: the scientists and the climate auditors don’t understand each other. There’s a culture clash full of misunderstandings.

Facing Off: Why Fact Fight Club

I can’t think of a way to create a single piece of media that can refute the ‘climate audit’ site. But here’s an idea for a service that could make a small contribution.

What if we set up a Web site that could match up strangers who hold opposing views and allow them to participate in a live video chat with each other. The participants would get instructions on how to structure their conversation. They’d be asked to spend the first 5 to 10 minutes answering an ice-breaker question and getting to know each other. Then they’d each give a short statement on why they either agree or disagree with global warming. Then they’d have a chance to give rebuttals. Let’s call it Fact Fight Club, though that name is intentionally provocative and probably not the best name for the actual service.

I built a very simple working version of Fact Fight Club using Blogger:
Screenshot 2014-03-04 09.58.10

The site relies on a service developed here at the Media Lab called Unhangouts, which makes it easy to set up Google Hangout video chats.

I tried to find two people to try this, but I wasn’t able to pull that off by the deadline. The concern, of course, is that the two people would take the “fight” in the name to heart and that the experience could feel more like a live-action flame war than a productive meeting of polite citizens. But I think there’s something to this idea of connecting people to those who disagree with them — to pop the “filter bubble” — and to do so in video so that hopefully people might be more civil because they can see the person they’re talking to.

I’ll be curious to see what people think of this idea.

4 Hour Challenge – Jeff and Primavera

When Biohacking Meets Art

Behind the Story: GoPro and Remote Reporting

This piece represents a reporting experiment, and though we weren’t able to get as much together in four hours as I had hoped, this was a proof of concept of remote reporting with the latest GoPro camera.

GoPro’s Hero3+ camera can stream live high-quality video to a smartphone app, and the camera’s features can all be controlled remotely. So yeah, we strapped the camera to Primavera’s head, and I sat in the next room essentially looking through her eyes and deciding when to record or not as she worked on a biohacking project she’s in the middle of. A picture below shows a screen grab from my iPhone during the recording.


I watched them work for a little over an hour, recording about 20 minutes of footage total. Then I recorded short audio interviews with Primavera and with a scientist she was collaborating with. The narration is an edited version of that interview with Primavera, and I didn’t have time to edit in any of the other interview.

We had two major issues I hadn’t accounted for:

* Slow rendering time: GoPro was designed to capture high-quality footage of high-action scenes. It can record in 1080p resolution, which is overkill for this project. I set it for 960p, but I should have dialed it down further, since the high-res footage created two problems. One, the file sizes were immense. We generated about 15 gigs of raw video. The files were slow to transfer from camera to the computer, so time was lost simply waiting for that. And then when I brought the files into FinalCutPro, the program had to render them, which was also time consuming. This meant I had far less time to pick and choose clips and fine-tune the piece to meet our four-hour deadline. I probably should have used GoPro Studio, free software that comes with the camera, but I wasn’t familiar with it so I went with a program I know better.

* Light issues: We started our project at around 2:30 pm, and the light was excellent then. The building we were shooting in has skylights, so it was pretty ideal. But by 4:30 the light was getting dim. Things still looked fine in the viewfinder and on the app, but once we imported the footage, everything after 4:30 looked so dark you can hardly tell what is happening. We had to scrap most of that footage, and the short clip that is in here looks like we switched to black-and-white.

Bigger Issues

My theory on this is that subjects of a story might feel less self-conscious about having a reporter’s camera present if the reporter wasn’t in the room. That theory was totally wrong. All three of the people involved in this biohacking project were frequently thinking about whether the camera was getting things, and they spent time handing the camera off to each other, trying to get the camera to look through the microscope’s viewfinder, etc. Because as the reporter I wasn’t able to decide where the camera was positioned, this was really a story co-created with the subjects. I did the editing and made decisions about what to put in and what to leave out, but I only had footage that the subjects had decided to take (with some general instructions by me at the outset).

There may be a few situations where it is simply too dangerous for a reporter to tag along, but where subjects are willing to carry a GoPro. But that’s probably a rare case (I’d be curious to hear what others think, though).

For me, this fits into a broader project on having subjects make multimedia diaries of their lives, and it seems like Google Glass is better suited for that (less invasive to the wearer). Still, the GoPro is an interesting new tool.

Jeff’s Media Diary


I decided to present my media diary primarily as a short video highlighting the varied ways media reaches me. Watch it here: http://www.youtube.com/watch?v=TeeHQJxOCtg&feature=youtu.be

This is in no way meant to be comprehensive, and in full disclosure it is a composite of two days.

Putting this together taught me a few key lessons:

* I’m nearly always consuming media — I was surprised by just how often. It depends on your definition of media, of course, but even when I was having a conversation with my wife or a friend, I noticed that there was often some form of media playing in the background (a TV in a restaurant, kids music at an indoor playground, etc.). And as you can see in the video I listen to podcasts on my iPhone whenever I’m walking, riding the T, or doing chores. So it’s rare that I go for more than half an hour without some kind of media input while I’m awake.

* Nearly all media I consume is digital, but I’m holding onto some print, mainly newspapers (Boston Globe and student newspapers) and a couple of magazines (the New Yorker and some cooking magazines). At times that creates awkward workflows, though. For instance, one day this week I picked up a Harvard Crimson when walking past a newspaper box in the law school, and I found an interesting story I wanted to share. So I opened my laptop, directed my browser to the Harvard Crimson online, then Tweeted the link to the story. Switching from print to an online version of the same article was a pain, but a couple of my followers retweeted it and it sparked a conversation with another journalist about the article, so I’m glad I did. I’m not in the habit of visiting the Crimson online, but I try to pick up the print paper while on campus because it’s nice to have the print to read while waiting for class to start. This incident made me realize that it would be easier if I just read the Crimson on my phone or laptop, and I’m sure that’s where I’m headed in the future.

* I guess I’m addicted to podcasts.

Content/Device Breakdown

I did log my computer activity for several days using RescueTime to get a more-detailed sense of my consumption patterns. I also kept a diary manually with major types of activity, since I found that RescueTime sometimes misclassified my usage. Here are some highlights:

* On a typical day I use my laptop for between 5 and 6 hours.

* When on my laptop, my biggest activity is e-mail. I spend about 10 to 20 minutes per day on Twitter (via Hootsuite), and about 15 minutes on Facebook. I do make a point to go to a few newspaper sources throughout the day (The New York Times is my home page and I often check Google News). I’m sad to say I read DrudgeReport every day.

* I use my iPhone to entertain me whenever I’m by myself (this adds up to a couple of hours a day of usage). I either listen to podcasts, check my e-mail, check Facebook, or read an article on the phone. This realization makes me think of that recent Louis CK rant.