Immersion (A brief history)
The Intimacy of Creation
A little less than a year ago Daniel Smilkov and Deepak Jagdish joined my group at the MIT Media Lab. Daniel had impressed me during the interviews with his sharp analytical mind, while Deepak had the emotional intelligence needed to be a gerat artist and designer.
This was during the Fall of 2012, an unfortunately over-traveled semester. During a five month period I visited Belo Horizonte, Tianjin, Kyoto, Tokyo, Brussels, Santiago de Chile, Lisbon, New York and Washington D.C. It was an overwhelming time, and keeping up with emails was yet another head ache. It was in that context that I pitched the idea of Immersion to Daniel and Deepak. I had been thinking about the need for a visual inbox for a few years now, but I have not yet had the opportunity to assemble a team that coule make that happen.
Both Dani and Deepak liked the idea from the beginning, and started working on it right away. By the 2012 Media Lab Member’s Meeting they had a working demo (see Video 1 below), which Deepak cleverly called Immersion.
During the Fall Semester Deepak and Dani also participated of a course called Data Centric Projects (DCP). I designed the course together with Ethan Zuckerman, Catherine Havasi and Sep Kamvar. The course was designed to satisfy a number of constraints, such as having a course that empowers rather than distracts from research. DCP was not a series of lectures of seminars, but a place where we coached student projects. The idea was that Media Lab students are talented enough already, and that a little bit of coaching could help them distill the essence of their work. DCP required students to present a different aspect of project every two weeks. These included technical exercises, such as summarizing the biases of their data and the methods used to mitigate them, and narrative exercise, such as writing about their project as if they were a journalist reporting on it. Although my traditional academic friends will shriek at this idea, the course was actually inspired by American Idol, since I have observed that periodical doses of audience exposure and tough love could hone talented people into professionals. The course was essential for Immersion to evolve as a project and an idea. Below is a video of what Dani and Deepak had to say about Immersion at that time.
During that semester I became the first heavy user of Immersion. As I played with the tool the network that existed inside the screen started to slowly move into my brain. The act of sending an email became an explicit act of weaving. The fact that we nudge webs of people to accomplish goals becomes transparent, not implicit in our communication technologies. Immersion also started to slowly change the way I saw the world.
One thing that becomes conspicuous when you experience Immersion is that we participate in a number of clusters. So when we interact with people we are not having a dyadic experience, but interacting with a member of a well defined cluster. What also becomes clear to me is that each of these cluster had its own emotional and intellectual identity. A local zeitgeist if you may. In my case there was one particular cluster that was emotional poisoned, not entirely, but at its core. So Immersion helped me visualize that it was time for me to reorganize my interactions and extirpate the emotional cancer. Immersion helped me understand how my network would evolve in that situation, and what were the links that I needed to strengthen to eventually reconnect the network, and heal that wound.
To try Immersion out we created a small exhibit in the Media Lab lobby, which sat next to the Mediated Matter group’s majestic Silk Pavilion. The exhibit was presented during the Media Lab Member’s event in April, and was well received by a number of people that embraced the “social nudity” of exploring their personal data in front of others. It was also then when we started to realize the emotional reactions that Immersion was able to generate.
In mid June we participated from the Science of Success workshop organized by Dashun Wang. I was checking my email on my phone when I received a note from a Boston Globe editor. He had seen the exhibit in the Media Lab lobby and was interested in Immersion in the context of the NSA scandal.
I was well aware of what metadata means–I have been working with metadata since 2004. I was also well aware of the scandal, since I had been doing a bunch of press that stemmed from a paper I recently co-authored with Yves-Alexandre de Montoje, Vincent Blondel and Michel Verleysen. The paper looked at the probability of identifying an individual uniquely using cellphone metadata. In my opinion, the project was almost ready, and the only thing that we needed was that final deadline to make the final push. The Boston Globe exclusive sounded as a good opportunity, so we decided to give it a go.
On June 30, 2013 Immersion was opened to the public and unveiled through a small interview at the Boston Globe. The reception was modest during that Sunday, and I tweeted the project in different ways to try to get some people to adopt it. The Monday right after that–July 1st–went more or less the same until mid afternoon. I remember checking Google Analytics at 3 pm and thinking that we were peaking at 265 people an hour. Yet, I was totally wrong. A short article on NPR ripped Immersion out of the intimacy of the process in which it was created and launched into the mainstream. A couple of hours later, our server collapsed. The project that let’s you spy on yourself was everywhere, except in the web.
The next three days were days of intense reengineering. Dany Smilkov rewrote most of the backend of the website in a matter of a few days. The problem was that there was one operation, the parsing of names, that was done on the server side. This was enough to jam the computer when we started to receive thousands of people per hour. The new version of Immersion pushes everything that it can to the client. Immersion serves your web browser a file with your emails, and the code to parse it and visualize it. So the server is basically queuing the collection of email metadata and sending you a bunch of text files that help your browser create pictures.
We relaunched Immersion at 10am of July 4th. After a day of good traffic (18k visits) the site exploded against on July 5th with 60k visits in one day. I guess we were lucky that a July 5th Friday was a perfect day for online procrastination. The site is now serving tens of thousands of people every day. We peaked after Wikileaks tweeted us late on Friday, and received nearly 7000 visits between 6 and 7pm (EST). We also experienced our second outage at that time .
After the Tide
So what is left after the tide has receded? What lessons can we rescue? Certainly, each of us will rescue something different from the project, but there are a few bullet points I would like to highlight while I am still thinking with my heart.
First, by using Immersion, and interacting with Dany, Deepak and the other members of my group, I realized what type of advisor I want to be. I don’t want to be the academic advisor that is travelling all the time, and coming back to MIT briefly so students can report back to me. I have to be busy for other people, but not for my students. Although nobody knows my students, they are more important to me than the well-known people that invites me to speaking engagements, and I have to act accordingly. With my students I can work together to do things, not just talk about them, and that is simply a fantastic relationship to have. With Immersion I learned to prioritize my interactions and focus on strengthening the core of my network. Explore the periphery can feel “exciting”, but it is distracting. Ultimately, a strong network core is needed to create content, while a periphery is needed to share it. If you don’t put your core first, you wont have good content to share. You can still be successful, but you will be a phony boloney.
Second, I learned that art, science and technology are not different disciplines. This is something that I have always believed but that the project reinforced. The creation of Immersion required strong technological skills, which both Dany and Deepak have, but also artistic sensibility and a scientific mind. Certainly, the timing of Immersion was a big key to its popularity–thanks Snowden–but I believe that the project worked because it was executed without compromises. It was executed by putting art, science and technology in the same ground. I have worked in the past with people that talk about the importance of having diverse teams, but does not what the walk when pushes come to shove. You know them, these are people that believe that other disciplines are there to serve their goals. The people that see other disciplines as tools rather than content. Immersion was executed properly because a conversation about the choice of colors was as important as a conversation about the technologies and algorithms used.
Finally, I have learned the importance of letting things go in two important ways.
First, although Immersion was executed at my group, after suggested it, in a deep way it is not my project. It is Deepak and Dany’s project. What I had a few years ago where ideas, and ideas are a dime a dozen. I am positive that there are hundreds of people out there that thought about a visual inbox, and the idea of using a network to visualize emails is as old as emails. What we have now, however, is a reality that works, not just words about it. The idea of a visual inbox for emails would not have been visited by hundreds of thousands of people in a few days. It would not be worth a tweet. The crystallization of this idea, however, ended up being more powerful than the idea itself. It provided the project a life of its own, and allowed the idea to break the prison of our minds. So, the intimacy of making this idea into a reality is what makes Deepak and Dany the experts here. Immersion is now full of their ideas. Some of this are visible in the front end, and others are seemingly embedded in the back end.
The second way of letting go connects to the framing of the platform. Ultimately, an idea becomes alive when it brakes out of the prison of our minds, a break that is made possible by the encapsulation of the idea in a “tangible product” that allows other people to experience the idea in absence of those that gave the idea form. For us, Immersion was about art, technology privacy and strategy. For the press, it was the tool that allowed you to PRISM yourself. A tool for self-spionage. In a week Immersion made a dent in the discussion of the Government’s use of metadata, and has given a new meaning to the “it’s just metadata” argument voiced by governments. Is this what we wanted? Who cares! We are happy to let go and ride that wave, since we also have something to contribute in that arena.
But what makes me happiest about Immersion are two things: the faces of Dany and Deepak when I see the excitement caused by the attention that their work has received, and their professional development, since now I know that they are professionally much stronger that when I met them a year ago. Certainly, this team has a long way to go, but I am happy to know that the Immersion project is now one they will take with them for the ride.
Cesar A. Hidalgo
July 6, 2013