What if we could receive real-time feedback on our social interactions? I developed a system like this for myself using Amazon Mechanical Turk to explore in the form of a performance. During a month of continuous dates with new people I met on okcupid, I streamed the interaction to the web using an iPhone app. Turk workers were paid to watch the stream, interpret what was happening, and offer feedback as to what I should do or say next. These directions were communicated to me via text message.

This was a self-initiated project created to explore a number of questions. What if we could receive real-time feedback on our social interactions? Would relatively unbiased third party monitors be better suited to interpret situations and make decisions for the parties involved? How might augmenting our experience help us become more aware in our relationships, shift us out of normal patterns, and open us to unexpected possibilities? We’ve seen with open source software development, for example, how communities can come together online and create something beyond what any individual could even fathom. What if we applied similar methodology to ourselves and our lives? The challenge I posed to myself was to prototype a system that could test these questions, and create a performance around it that would communicate this idea and experience to others.

This work comes out of feeling like it’s sometimes difficult to connect with people in a really honest and open way because we are so caught up with social routines and expectations. I’m interested in ways that technology can be used to augment, subvert, alter, mediate, and ultimately deepen social interaction. Can we create systems to break us out of our normal systems? Can we use data collected from our daily experience and networks to identify patterns and inspire new ones? I want to make different tools that try to create these situations. In general, I am trying to create situations that temporarily destabilize our concepts of our self and others. Maybe in these moments of slippage, losing grasp of the picture of ourselves in the world we work to maintain and the expectations that go with it, we could stumble upon a little more meaning in our interactions and relationships. The technologies we create shape our experience into rational interpretations. I want to keep pushing on the boundaries of this and asking what kind of experience of reality are we building for ourselves?

In order to do this I moved to a place where I didn’t know anyone (Portland, OR), and went on a date with a different person that I met through OkCupid each day in order to rapidly prototype the system. Using my iPhone, I discretely streamed the date to the web, where MTurk workers followed along and gave directions through a webpage with embedded video feed and submission form. I tested different options for the workers each date -- sometimes they voted to determine how long I prolonged the date, other times they could directly determine things for me to say and do. I received the feedback via text message and had to immediately incorporate the instructions into the date interaction. On the web form, I also had the workers interpret and describe what was happening in an effort to encourage them to really engage with the feed and so I could get a sense of how much they understood from the video. I was surprised how much the workers were able to notice from such low quality audio and video. They picked up on the times when I was nervous, uncomfortable, bored, interested, engaged with surprising accuracy. I considered the question of audience often with this project. In one way, this was a kind of performance or provocation for the workers. I was interested in confronting this group people, signing on expecting to do robot-like tasks, with a very human task instead. I received a lot of unsolicited feedback from the workers expressing their interest and asking to be kept updated on the project and future opportunities to participate. On the last night, I opened my performance to the wider internet audience beyond the workers, inviting my followers on twitter, facebook, and mailing lists to participate. As the month went on the project took on new layers of meaning for me. What at first felt like a game or a performance started to blend into my reality to the point where I wasn’t sure where I ended and the workers took over. Issues of control, identity, privacy, crowdsourcing, surveillance, and augmentation became more poignant as I remained committed to my experiment and performance. Each day I posted the MTurk results, along with statistics including amount paid, number of responses, average response time, average rating, and interface offered to the workers. I also included my own reflections on each experience (see socialturkers.com for full logs).

This piece is a provocation that talks about potential futures through the creation of situations that are real in the present. As a performance, I am able to experiment and prototype this system using myself as the test subject. At the same time, I offer the possibility of this experience to anyone that chooses to build a similar system modeled off of my code and technical setup. I question the direction we are heading with the development of augmented vision systems and networked mobile applications. However, it is not meant to be purely critical -- embedded in the performance is a suggestion of a networked humanity that uses its collective wisdom to improve peoples ability to interact and form relationships. Most importantly, the piece is meant to question how we define our identities, and what freedoms might come from clinging to these ideas a little less tightly.