Activism and Technology

SpeakOut is a tool for protests that uses the devices we already have in our pockets. 

One person speaks into their smartphone, and others can amplify that live audio over a wider space. SpeakOut is a web app that runs in the browser, with no need to download a native app.

SpeakOut is my graduate thesis at ITP at NYU. You can watch my live thesis presentation here, access my final slides with notes here, and read my accessibility script in Word here.


Role: Design research, opportunity definition, user testing, concept development, prototyping, front-end and back-end programming.


Tools: PeerJS API, library, Javascript, HTML, and CSS.

Civic engagement is up. More people are attending protests, and protests communicate messages and empowerment at scale. Most of all, the U.S. Constitution protects the right to publicly assemble and the right to free speech -- and so should use those rights.

Unique amplification methods at protests are not new. The Occupy movement created the Human Microphone at Wall Street's Zuccoti Park, in which audience members repeat the speaker’s words to the back of the crowd, phrase by phrase.

In relationship to the Human Microphone, SpeakOut is, of course, digital, and SpeakOut has no delay - it’s synchronous, meaning the audio is immediate.


The Story Behind the Project

It can be surprisingly hard to hear speeches at protests. Last spring I was at this rally, pictured above, organized by Transportation Alternatives, a New York City advocacy group for safer streets. I knew TransAlt from organizing with them in the past.

A group of us talked about how frustrating it was to not hear the speeches, even when they used an electric megaphone. I had a feeling I could do something about it with my skills.  And I knew a lot of people have phones in our pockets. Could we use these phones, too?

Opportunity Statement

As I built this idea out to be my thesis, I began to wonder:

“How might protesters hear better by amplifying speeches themselves, using phones and real time communication technology?

Thesis in Three Parts

  • SpeakOut, the web app itself, which I designed in a tight use case with TransAlt.

  • Data rights considerations built into the web app, which stem from...

  • My search for what a phrase like ethical technology might even mean, and how I might design it into SpeakOut.


SpeakOut at Live Rally
In March 2019, I tested SpeakOut at a TransAlt rally to finish a protected bike lane from end to end on Queens Boulevard in Queens (known as the “Boulevard of Death”).

On a rainy Sunday in March in a crowd of 50, people gave speeches into a smartphone on a mic stand. The smartphone streamed live audio using SpeakOut’s web app in the browser. Slowly, a light echo of amplification accumulated in the air as five to ten people chose to amplify that live audio with their smartphones, surrounding the crowd in the message of the speakers.

SpeakOut addresses the problem I set out to solve: people beyond the first couple rows often cannot hear speeches at protests without sound systems. With SpeakOut, people can amplify the speech themselves using the devices in our pockets.


How Does It Work
SpeakOut runs in the browser, meaning you don’t need to download an app to your phone. It uses PeerJS, a peer-to-peer API built on top of WebRTC that supports media streams.

Using SpeakOut, people with even just a few phones can create their own sound system.

User Testing
I conducted nine user tests and sought feedback from fourteen different individuals and groups over four months. User testing greatly informed the current design. Here are two highlights:


  • Regarding rallies, TransAlt organizers said “We want people to show up, not stay at home and listen remotely.”


This feedback went into the structure of SpeakOut. Listeners won’t be able to find a public TransAlt broadcast by just visiting the web app. They need to receive a unique link, which is only possible when the organizers decide to share it.


  • Regarding vulnerability of protesters, an activist said “When we become protesters, we are a constitutionally protected group, yet we are also vulnerable depending on many factors such as disempowered contexts or one’s identity.”

This concern reinforced my decision to add data rights considerations now and in the future (mentioned below), and to make the code open source so that organizers can choose to spin up their own versions with additional privacy protections.

Data Rights Considerations

As for privacy considerations, SpeakOut requires no phone number, no email address, and no account to log in.

There is also a Data Decision Rights page. By Decision Rights, I mean the right to decide what happens to your data. The page explains built-in protections and suggests more protections you can add.

This same content is sprinkled throughout the web app, especially on the Speak page. This is because organizers or people giving speeches are making data decisions on behalf of many others to use SpeakOut, the internet, and their phones. All of these are part of infrastructures that inherently share some amount of data about individuals available to others.


Ethical Technology

The more I became attuned to the design requirements for SpeakOut, the more I found a tension between needs for publicity and privacy. 


Protests are meant to be public and attract attention. Using SpeakOut helps attract even more attention by making it easier to hear speeches, even if you’re walking by.


But there are contexts in which privacy is just as important. People at protests can be in disempowered contexts. In these moments, using one’s phone can result in unwanted data collection by others.


I continue to search for frameworks or examples of technology designed with people as the first priority. The one that’s caught my eye is the Web Consortium’s Accessibility Guidelines, also known as WCAG. WCAG standardizes how to design webpages for people with disabilities.


By designing for these people first, we end up with a better web for them and for everyone.


SpeakOut has led me to a similar conclusion - by designing first for people with the most to gain from adding data protections, we end up with a more creative and inclusive product for them and for everyone else.

Next Steps

Looking ahead, features of SpeakOut might include:


  • Using SpeakOut to coordinate chants, not just speeches.

  • Adding speech to text for people who are deaf or hard of hearing.

  • People have suggested it could be powerful to allow remote speeches. For example, people at our borders, jails and prisons could speak from afar.  

  • I also wonder if a sister app to SpeakOut could let people at protests sign up for future actions in a privacy protected way. That way, when they are most inspired, they can get involved in the future organizing of today’s social movements.

  • Grey LinkedIn Icon
  • Grey Twitter Icon