In this three-part series, we focus on key elements behind the successful crowd-sourcing project, Decode Surveillance NYC, including the design of the data collection, early analysis, and what it means to build community online.
Sofia Caferri, who joined the project team as a volunteer moderator, reflects on how the discussion forum became a space for building a supportive community and creating new learning opportunities.
Decode Surveillance NYC poses questions about how we can create more meaningful ways for people to participate in activism online
Sofia Caferri
A moderator’s perspective
I joined the project team as a volunteer moderator and later became an assistant. For the 10 weeks, the Decode Surveillance NYC was live, I logged on to the forum daily – welcoming volunteers, answering questions, and moderating discussions.
The Amnesty Decoders forum is a discussion space where decoders can get support from peers, moderators, and Amnesty staff. The support and advice provided in the Decode Surveillance NYC forum focused on the assignments and how to identify cameras.
The forum I moderated was mostly supportive and cooperative. It became a space for peer-to-peer learning, friendship, and conversation.
Peer-to-peer learning
I found the most powerful aspect of the discussion forum was how it enabled peer-learning and a collaboration between volunteers. Decoders engaged with the forum indifferent ways; from asking for instructions and answering questions about the assignments, such as how many cameras were at a given intersection. Discussions also ranged from speculation about small or partially obstructed cameras that proved hard-to-identify, to constructive feedback, and reflective or personal anecdotes on facial recognition technology and surveillance.
When the project launched in April 2021, the discussions were facilitated by moderators, including myself and Amnesty staff members. Over time, decoders increasingly joined the discussion and engaged in more peer-to-peer conversions, on a wider range of topics.
The forum shifted the experience of learning about surveillance technology beyond the passive act of “being taught” how to carry out tasks. Rather, it enabled a collective way of learning how to recognize facial recognition technologies. Each flagged assignment that appeared in the forum was a starting point for discussion and engagement, potentially with a new decoder. Decoders would intervene and support as well as contribute with additional information.
The forum created opportunities to learn from each other’s mistakes. Occasionally, decoders would flag completed assignments, where they mistook surveillance cameras for something else such as streetlights. Other times, mistakes occurred because cameras were difficult to identify due to bad lighting, the image angle or distance.
Either way, decoders were supportive in pointing out mistakes while suggesting the correct answer or recommending tips on how to recognize cameras and tag them correctly. In this way knowledge was pooled about the appearance of cameras, how to spot them and handle uncertainty.
Positive feedback loops
Interacting and collaborating with such a diverse cohort of volunteers exposed me to new perspectives and ways of working. More than 7,000 volunteers joined the project from 153 world countries. Moderators, who were based in different countries, ranged from students to workers, with diverse abilities and experiences.
To a degree, the Decoders forum levelled the power relations between Amnesty researchers, moderators and volunteers by making the learning experience a two-way exchange: a space where Amnesty staff and moderators learned from decoders and vice versa. This combination of a shared learning experience and regular feedback made me competent enough to help someone else in turn.
Building community and kindness
Feeling supported by and connected to others is important to learning in any context. This has been even more the case during the Covid-19 pandemic.
The online forum posed a unique opportunity for fostering an inclusive and equitable space for meaningful participation in Amnesty’s research, discussion, and collaboration.
In Decode Surveillance NYC, working together to achieve a common goal fostered connections between the decoders. The forum required a high level of engagement. To participate fully, decoders had to read posts, give feedback, like and share their opinions.
Decoders mostly encouraged each other to do their best, take care of one another, and be kind by making each other feel accepted or appreciated, regardless of age, gender identity or cultural background. Actively participating in the forum where there was a shared commitment to the wellbeing of others gave me a sense of belonging in a community that shared my same values.
Decoders were invited to respect everyone in the community, acknowledging and respecting other gender identities, sexualities, religions, ages, levels of written English, cultural perspectives, and so on.
This initiative was started by a volunteer who suggested and invited Amnesty staff, moderators and decoders to ensure that non-native English speakers were not alienated by the use of slang or terminology specific to a certain culture or country, and to avoid gendered expressions and use ungendered vocabulary instead.
Making a difference
Decode Surveillance NYC happened during the height of the pandemic, at a time when many people would have been unable to take part in direct action such as protests, and contributing to change felt hard given the scale of the problems we were facing. Here, the forum played an affirming role, thanking and reminding our volunteers for contributing to Amnesty’s research.
With the unprecedented rise of smart technologies and access to social media, activism has, in part, moved online. Over the course of Decode Surveillance NYC, I spent a considerable amount of time helping to generate data for the investigation. Volunteering enabled me to go in-depth on a single issue, beyond other methods of online activism such as clicktivism (e.g. sharing social media posts, signing petitions, creating hashtags). Indeed, working on assignments together with other volunteers from around the world provided me with a sense of achievement and purpose.
Knowing that the data and results are being used to fight for a ban on facial recognition in New York City has given me a sense of participating in real research aimed at real world impact.
Conclusion and future questions
While the data-driven research made possible by the project was extremely important, the social dimension was arguably of equal importance but harder to measure.
To some extent, the nature of the activity involved, completing assignments, reading and commenting on posts, and giving and receiving feedback were important measures of success of the Decoders forum. Despite the fact that the project was only online for three months, these interactions promoted peer-learning, inclusivity, care and knowledge sharing. Individual decoders collaborated by learning from each other’s opinions, mistakes, and feedback.
Glossary
Decoders (usually spelled with a lowercase “d”) are the digital volunteers who contribute to Amnesty Decoders.
Assets are the images, documents or other media analysed by decoders. For Decode Surveillance NYC, an asset was a Google Street View panorama image.
Assignments are the tasks that decoders complete. A decoder cannot get the same assignment twice. For Decode Surveillance NYC, each asset was analysed by three decoders before being marked as complete. As a result, there were three times as many assignments as there were assets.
Topics are collections of forum posts, more commonly referred to as a thread.
Sources: Discourse and Hive
*In the logic of Hive, tasks mean something different again.