Open Humans

I'm Mad Price Ball, PhD, co-founder and Executive Director of Open Humans. My work focuses on enabling people to access, use, and share their health and personal data to advance individual understanding, collective empowerment, and research.

This work primarily occurs through the nonprofit project Open Humans, which combines technology and community to advance an open and individual-centered approach to make discoveries with personal data.

I've led Open Humans since winning a prestigious Shuttleworth Foundation Fellowship in 2017, to help individuals contribute their data to research. Prior to Open Humans, I served as Director of Research for the Personal Genome Project. I also currently serve as a Director on the membership-elected board of MyData Global, a Finnish nonprofit engaged in personal data rights advocacy.

We all have questions about our lives: “Is social media making me unhappy? Does this diet help my condition? Does this medication work for me?”

Answering these questions can be hard – and it's especially hard because everyone starts from scratch. At a collective scale, people might pool their data to ask a shared question – but this rarely happens. We need to collect and understand information about ourselves, but we lack an ecosystem for creating and sharing the tools to do that. A lot of questions go unanswered. This isn't a reality we should accept.

With Open Humans, I want to create that ecosystem: a place where people do these things together, sharing and adapting their tools and solutions.

Humanity deserves an alternative to our personal data dystopia. From health to civil society, the ability to collect and understand data about ourselves is vital in an increasingly digital age.

The phenomenon of collecting and analyzing personal data is not new. Quantified Self has recorded hundreds of "show & tell" talks where people share their discoveries. The subreddit "Data Is Beautiful" was so flooded by personal data analyses, it limited such posts to Mondays.

These are the creators that could be sharing advice and tools – not merely their results. The communities that stand to benefit are far broader.

Chronic, complex diseases affect millions of individuals, and often require personal solutions – and finding them involves self-tracking and personal data empowerment. Nightscout Project and Open Artificial Pancreas System are examples: open source projects where a few individuals created tools used by thousands of Type 1 diabetes patients and caregivers.

Outside health, personal data can be used for other types of wellbeing and insight – and for advocacy and collective action (e.g. by Uber drivers). Understanding (and potentially using) the data others collect is part of being informed and engaged digital citizens. Thus, all "digital citizens" are potential beneficiaries, as our personal data is inextricably entangled with so many aspects of our lives.

I want everyone to be more empowered to use data about themselves, to ask questions and make insights.

This data – personal data – can range from health-related data (e.g. activity trackers or diet logs) to data about almost anything (e.g. music listening habits). It can also be highly individual! Recording when you had a compulsive thought, or took a particular medication. 

Often there exists no pre-made solution: we must adapt existing tools to meet our needs. To make that easier, we need an ecosystem where people are sharing solutions.

Open Humans already has powerful features, and a nascent community. People can import data, explore it, and create group projects. But we need to go further, to do:

• self-research with personal data

• easy sharing & re-use of ideas, tools, and solutions

• "scale-up" to group efforts

We must connect what we already have, to enable an ecosystem for personal data empowerment and discovery. By working closely with Quantified Self, patients, and other communities, we must ensure that it's useful and usable. We can make a place where people want to do their self-research – and easily share and re-use each other's work – to have the ecosystem we need.

Ultimately this project serves a broad, disempowered group: people that aren't currently able to explore and use their personal data, and would benefit from being able to do so. (That might be you!) The relevance ranges from health to civil society – from patients struggling to manage chronic conditions, to advocates investigating the data corporations collect about us.

We know solutions exist: self-research that creates and uses personal data already happens, but is rarely shared in ways that others can build upon. In a 2019 ethnographic report on the Quantified Self community, Nils B. Heyen observed: "it is noticeable that, so far, no knowledge accumulation can be observed in the QS movement … each self-tracker starts more or less from scratch."

We can achieve more together – as with software (Linux & open source) or knowledge (Wikipedia & Open Street Maps), we need a similar "open" ecosystem for personal data: where solutions are being created and shared.

We must engage people already creating solutions in a way that encourages sharing – and sharing in ways that make it easier to re-use and adapt solutions. As the barriers to doing these things get lower, empowerment expands.

Digital citizens are disempowered in how people create and use data about them.

In general, our personal data is collected by others. We are not creating the insights – at best, these are created for us. (More often made about us, and not seen by us at all.) We rely on others to create solutions.

The less tech-savvy you are, the more this disempowers you.

Fixing it requires an "open ecosystem" of sharing solutions – where do it ourselves, rather than being dependent on others. The innovation that creates can scale-up in numerous ways to benefit many beyond it.

I originally co-founded "Open Humans" to "share your data to advance research" – academic origins, spun out of George Church's Personal Genome Project.

Separately, I had friends active in open source & free culture (e.g. Wikipedia and Linux). I saw how technology can be democratized, and the barriers. The undemocratic nature of researching "humans" (ourselves!) became clear: we rely on institutions to do it, even if the only "equipment" needed is computers.

So Open Humans enabled community projects, just as academic ones – but we had no community. Bastian Greshake Tzovaras joined the project and observed: "you go to Open Humans and you don't see the humans".

A key insight came from one of those friends, Mako Hill, who studies peer production ecosystems (e.g. Linux and Wikipedia). Mako noted a common pattern: one person does something to solve their own needs, and is happy to share. Usually that's it – but sometimes it grows.

That's brought us to our current effort, which Bastian now co-leads. We're striving to support self-research with personal data such that people interact and share solutions, and to be a place where solutions can sometimes grow.

It is a travesty that humanity is so disempowered in studying ourselves.

I did my PhD at Harvard University, and I've stepped off that academic path to tell you: you don't need a PhD to do science. We don't need to rely institutional authorities, to hope they do good things for us. We can do it ourselves. This disempowerment isn't intentional: by and large, academics would love to see more people learn. Technologists are similar – I can code, and my spouse does it for a living. Using our data is harder than it should be, and technologists don't mean it to be inaccessible – again, an unintended outcome.

Solution-creators in this space are often moonlighting academics and technologists, applying their skills to explore personal questions and challenges. They'd like their work to help others too, and we need to make that easier.

It should be an obvious statement: we should be able to collect, understand, and use our personal data. Laws like GDPR promote it. But it's not happening, it's not going to happen unless and until we have an ecosystem for doing it.

I'm passionate because I know there's potential good, and I know why it's not happening.

My experiences and work have encompassed data & research ethics, human subjects research, open source/culture, as well as web/data technologies, operating a nonprofit, growing a highly skilled community, and applying funding previously won to accomplish what we have already. It's a unique combination – and I've been working on this problem for a while: I've learned what won't work.

I've brought Open Humans to where it is now. What I propose is what I know is missing and needed to achieve our mission – "empowering individuals and communities around their personal data, to explore and share for the purposes of education, health, and research".

Launched in 2015, Open Humans has evolved substantially – especially after I assumed leadership in 2017.

Technological features for people to create, share, and use:

• tools to retrieve personal data via APIs (e.g. Fitbit, Spotify, Twitter)

• personal data analyses

• activities that invite data sharing (academic studies or otherwise)

Community aspects:

• Twice-weekly community calls

• Online chat (Slack)

• Community governance (activity reviews & community-elected board seats)

A recent example of how these combine is "Quantified Flu", which started as a question during community call – can we use wearable devices to understand and anticipate when we're getting sick? Co-created with code/data contributions from volunteers, individuals can visualize ongoing symptom tracking alongside data (e.g. resting heart rate).

Few others have achieved what we have, learned the lessons, and have the connections to innovators and stakeholders. I and Open Humans are uniquely positioned to be the foundation for much more.

At the start of 2020 we launched a project for people to do self-research projects together. I wanted an open-ended exploration of what self-research looked like – by helping people as they do it – to understand needs and challenges.

Then, the pandemic happened.

Projects ground to a halt – for understandable reasons. Questions and curiosity lost importance in the face of radically changed circumstances. It seemed our plan had lost relevance in the face of global challenges.

But our data is still relevant! Something new emerged, a question that arose during a community call: can we use wearable devices to understand and anticipate when we're getting sick?

Thus, "Quantified Flu" was born. We drafted an initial tool for basic data visualization. Now, thanks to substantial contributions of code, expert advice, and the ongoing participation of scores of volunteers, people can visualize ongoing symptom tracking alongside data (e.g. resting heart rate). It continues to grow and improve.

But Quantified Flu isn't just about the pandemic. I've ensured it's flexibly built, because symptom tracking is also a key need for chronic disease patients. The project isn't merely creating something relevant for the moment: it will remain relevant beyond it.

I co-founded Open Humans, but didn't always lead. I began working on it in 2014 – thanks to substantial awards I helped win – and worked alongside a team, but not managing it. When funding neared completion in 2017, I was the one seeking new funding. I succeeded, winning the Shuttleworth Foundation Fellowship, a prestigious award that catapulted me into the leadership role. I had to learn a lot and do it fast.

For me, a leader can be measured by their ability to inspire others to join them in their efforts – especially other leaders. Dana Lewis of OpenAPS has joined our board of directors, as has Gary Wolf of Quantified Self.

Leadership also means elevating others to become allies. When I recruited Bastian Gresake Tzovaras to join me in 2017, he was finishing graduate school. As an open science advocate and co-founder of the grassroots openSNP project, I knew he had potential. In 2019, Bastian won a three year academic fellowship in Paris, with funding for postdoc and students. His academic work is what we're trying to do now: making Open Humans more collaborative. He's brought students and collaborators to the project, which he now co-leads.

There's nothing like Open Humans – it's applying aspects of "open source" / collaborative projects like Linux and Wikipedia to a new domain: personal data.

Because people keep personal data private, open innovation in that realm has floundered. From data standards to tools to collect, analyze, and visualize data: all struggle from the effective silos that exist.

Open Humans acts as a secure steward to enable these things while protecting privacy. People can also create opt-in group projects to aggregate community data, while interacting with contributing members anonymously. People can use Open Humans to share tools that collect data, and code that runs on personal data – without needing to share the data itself with others.

An ecosystem like this is unique and potentially disruptive (in a good way) to seed innovation for uses for personal data that are desired by the individuals that data comes from.

The ultimate goal is to achieve positive use of personal data, collected and used in ways desired by the people that data came from.

Short term goal: To work with the people that are already engaged in self research that collects and uses personal data, to increase collaborative approaches.

The goal is to make working collaboratively to be sufficiently useful that expert self-researchers will want to work this way. To strategy to achieve this is to make it easy to share and adapt tools, methods, and ideas that make their own projects easier.

Mid-term goal A: Work with individuals that wish to perform self-research (but haven't done so before) – e.g. patient communities.

Shared tools for self-research need to at least be useful for expert users (i.e. already motivated to do self-research) before I expect newcomers to find it useful. Also, a community of experts is needed to produce solutions that can be shared with newcomers.

Mid-term goal B: Extend individual self-research to collective activities (e.g. using aggregate data from contributors).

This is how "open" or "peer production" ecosystems often work: scaling up from individual projects. Also, an example of how this otherwise struggles was the Galileo platform developed by Vineet Pandey (UCSD) for "empowering regular citizens to design experiments" – I attended his thesis defense, the few people that created structured research questions were mostly already familiar with research. To solve that: the first & easiest way to learn to do research "about people" is to do it about yourself.

Long-term goal A: To see "mainstream versions" of tools and methods that are more accessible to others that need them. By working with newcomers that try to use tools and methods developed by self-researchers, design issues will be uncovered – which can be addressed via projects or products that have specific funding for development.

Long-term goal B: To see collective activities produce positive outcomes for communities that contribute to them. I don't anticipate all (or most) community projects will succeed, but we will be learning what factors help make this happen.

A major issue that occurs throughout our work is the need to engage with experts and community leaders while not being seen as parasitic – but instead, acting in service to them.

They need to want to work with us – with the platform and community using it – to see it as beneficial to them, and themselves represented within it – not seeing their own achievements as co-opted or exploited.

Beyond being "nonprofit", the main way I address this is by emphasizing community governance, through formal and informal methods, to ensure that our work is in service to the stakeholders that contribute to it.

Formally, in governance, a third of the seats on our board of directors are elected by members of Open Humans. In addition, all candidates for directors are publicly shared prior to election, to allow for feedback from the community regarding support or concerns.

Also formally, group activities on the site (projects that can invite members to share data, and/or tools that add data to a member account) undergo a consensus community review process. These are performed publicly via our forums.

Informally, we have a community chat (Slack) and forums where members can interact with each other, and can openly ask questions, share ideas, and raise concerns. We also have community calls (Zoom) twice weekly where community members can similarly engage with each other and ask questions (the first focused on the overall platform / group activities, the second on self-research).

All code produced by our project is open source (GitHub), including the platform itself as well as projects we supported (e.g. mobile apps for self-collected GPS data, or to export data from Apple Health).

We also try to highlight projects and experts that are engaged with the platform (should they wish it!), e.g. via blogposts or community newsletters, to bring attention to their work.

With Quantified Self, since Gary Wolf joined our Board of Directors in 2019:

• We've supported Gary Wolf's Article 27 effort for "advancing the practice of everyday science through education, advocacy, and open tools"

• Worked jointly on the Quantified Flu project to advise and promote

• Planned a joint "Show&Tell" event and online self-research meetings

Furthermore we have patient groups we work with, using Open Humans to support their work:

• The Nightscout and OpenAPS Type 1 Diabetes communities operate "data commons" activities in Open Humans, and have developed tools to connect continuous glucose monitor data via Nightscout servers and mobile devices.

• The cluster headache community, Nobism, uses Open Humans to connect data for symptom self-tracking, and has operated two "data commons" activities patients can contribute to.

• In collaboration with the Alan-Turing Institute and the Autistic Foundation, developing a community-led research project to allow autistic people to share their experiences with each other

And with Bastian Greshake Tzovaras, at the Center for Research and Interdisciplinarity in Paris, we collaborate on additional projects:

• Just One Giant Lab (JOGL, a collaborative task-solving platform), a project to connect group activities to JOGL to see how social/collaborative features for Open Humans projects increase engagement

• OpenCovid, survey-based research that will use Open Humans to recruit participants and return data

• Extensive joint work on Quantified Flu, including managing technical contributions (Apple mobile app, support for Garmin / Google Fit data), as well as student projects for group data analysis and interactive visualization of integrated symptom-tracking and wearable data.

I think Open Humans has made amazing progress in gaining key allies and stakeholders, and in developing the platform and community – BUT we haven't had our work promoted much more broadly. Because of this, there are many potential opportunities and contributors are unaware of it.

That's partly because I'm concerned that me driving our own promotion single-handedly risks me being seen as having co-opted the incredible work that stakeholders and contributors have made in the community. (It's also because I've never done it before.) Amplification would be wonderful. I've not had mentorship/coaching, and potentially other advice (e.g. funding and revenue model) might also be valuable.

The minimum prize would support my work on the project for coming years, and provide funding for a couple others to help implement project goals, potentially via collaboration with existing allies and partners.

I would like to partner with Quantified Self, led by Gary Wolf (or Article27, their nonprofit effort), as they have enormous experience in this area – with self research, commercial and academic stakeholders, and patient-led research.

I hope to work closely with them on the project as a whole, and specifically on enabling individuals to engage in self-research using shared tools and methods.