“What would a future digital society where the ‘Mydata’ perspective’ has become reality look like?” This was the question behind both the “Imaginarium of MyData Futures” session during MyData 2018, and the “Speculative Data Futures” workshop ahead of the conference.
The rosy ‘MyData perspective’ could play like this: thanks to a combination of innovation and regulation, individuals have access to their personal data, be it data that they themselves have produced or data that organizations have captured about them and their transactions; through convenient services and technologies, they are able to understand their data and, more importantly, to come up with meaningful ways to use them to their own ends; this shift in the data economy enhances self-confidence as well as trust between people and organizations, it produces a new wave of innovations (that, in part, reduces the stranglehold that a small number of platforms exert on the data economy), and the creation of new, shared knowledge through participatory science.
The Happy Self Data Vision
But how could it happen, or happen otherwise, in reality? The workshop participants had been asked to produce a short story envisioning a speculative data future. These provide an insight on the hopes, expectations and worries that we express towards this future. Jesse Haapoja’s story was about a future where humans have internalized the fact that they are so much worse at making the right decisions than well-programmed computers:
“Sometimes I feel lonely. It was recommended that I should start a long-term relationship at the age of 31. [...]
I guess that this is as happy as I can be at the moment. Acting against the recommendations usually leads to suboptimal decisions. Maybe I would be happy for a while, but humans are bad at considering long term rewards and instead look for short term gains.
I don’t really date that much anymore. [...]”
Is that a happy or an unhappy future? The workshop kicked off with a discussion on “personalization”, including when computers make choices that are supposed to be good, or better, for us. What makes personalization “comfy” or “creepy”?
Aside from classic, yet important, privacy consideration, one idea emerges: Personalization should be about opening people’s horizons and deepening their understanding, rather than producing closed and opaque decisions, whatever their intrinsic value. Happily, Oguzhan Gencoglu’s presentation taught us that even machine-learning algorithms were now becoming able to produce “interpretable” results, hence open to feedback and discussion.
So we expect a data-driven digital society to be empowering, rather than to passively steer us into the “right” direction. But how can this come about? Focusing on health-related issues, the workshop’s participants identified 3 scenarios. One is a engineer’s dream: data provide awareness, leading to understanding, allowing for solutions to emerge, that can then be executed through and by programs. The second, focused on preventative healthcare, introduces several countermeasures, such as the need for doctors to remain in the loop. While a third scenario focuses on community action, whereby the production and interpretation of data is done by humans (with the help of computers) and combines personal and collective sense-building.
3 Scenarios Inspired by Health Data
There are several conditions for a MyData future to be truly empowering, the main one being data literacy, or at least a combination of knowledge and tools that allow people to make sense and use of their data, without becoming data scientists. Otherwise, we could produce a “Data Without Self” scenario, in which people are overwhelmed by the data they have access to, consider them as one more digital hassle to manage – or even one more digital risk – and end up distrusting the data economy even more than they do today.
Data Without Self
Julia Velkova’s story nicely captures this situation:
“Linda’s personal data future is getting more and more complicated the more data about herself she collects. In the same way as a manager job requires one doing more administration rather than applied work, so has she suddenly found herself building personal data storage facilities, and doing strategic investments in storage infrastructure rather than actually doing something she would find meaningful with data. […]
She has a pile of unsorted and uncategorised data which she just dumped in a virtual folder the same way one hides a mess in a wardrobe before the guests arrive. She knows she needs to clean-up but there is never time for that. Instead, she needs to hire a data cleaner […]
Her personal data future has become absorbed by her data storage infrastructure.”
Or… we could just enjoy more of the same: Today’s large data platforms leverage our ignorance to lure us into entrusting all our data to them and providing us with ever more convenient services, at the cost of our autonomy as well as an increased dominance over all sectors of the economy.
Now, achieving an empowering MyData future requires a major change in data infrastructures, which are today entirely designed with the needs of organizations in mind and do not consider persons as legitimate data controllers. The most obvious roadblock in the way of this future is the unavailability of data to individuals.
Self Without Data
What are the conditions for a MyData-enabling architecture to emerge? The workshop provided several insights, all of which, interestingly, combine technical, design and policy considerations. In one vision, the pervasiveness of data processing (materialized by bodily implants) requires a super-secure and usable technology stack, combined with “liquid democracy” capable of questioning and/or orienting individual as well as collective data-enabled decisions. In another, a tax on data centers is levied to finance open data-based science and to govern the use of data for the common good, on top of the private use of data for profit.
A third vision focuses on identities, both individual and collective. At a personal level, data would be managed on the edge, at the level of individuals, allowing both deep self-understanding and the projection of public identities through “skins” and masks. At a collective level, anonymized data would be managed as a commons, and used to continuously search for patterns and hidden knowledge, most of which would become public.
But are these infrastructural conditions sufficient to produce a positive Mydata future? Probably not, if only because we might have different visions of what “positive” means. In her talk during the Imaginarium session, Linnet Taylor speculated on 4 different futures related to different human communities.
In the U.S., people gain the right to become their own data brokers, however the data they have access to does not include derivative data. So people need to compete with the data derivatives produced by existing brokers. Most people just can’t, with the exception of a new class of self-data traders. US inequality rises exponentially.
In South-East Asia, people use portability and data sharing tools to influence their social risk scores. Local cooperatives emerge where people create risk scores for neighbours, and are rewarded by government through increased data access. Over time, the surveillance system decentralizes and democratizes. A lot of people like such systems. Those who don’t, too bad.
Stateless communities and refugee collectives use self data to provide self-sovereign identities and related data based on a blockchain (which states might actually love because they don’t have to take responsibility for it), producing a global energy crisis – all the more because when people die, the records would persist.
The fourth scenario is based on the idea of forgetting, where MyData provides both access to data, and a reset button for our lives. Predpol’s possible offenders can be directed to social services then their criminal history is reset. One can use information about genetic risk selectively and make the data inaccessible once they have used it. A somewhat similar, yet more radical idea can be encountered in Sanna Vellava’s “No-Data Nation” story:
“In the future, wastelands of today, filled with electronic devices, will be squatted by no-data activists, […] who are collecting their forces together in order to establish a new nation, starting their manifesto by a motto saying ‘no more data, back to ignorance, into the unknown future’. [...]
The handwritten messages have lured and enchanted no-data activists from all across the globe, having grown tired of their over-intelligent smart-ass machine colleagues, tired of being in the optimal condition all the time, everything being so smooth that no one can tell what is dream and what is real.
So, they decide to conquer iMount and establish there a new nation called No-Data Nation, mining old computer parts from the piles of waste, and trading them with those who still believe that Bitcoin will make them rich and powerful one day.”
Linnet ended her talk by quoting Evgeny Morozov: “The true challenge for the data distributist left is to find a way to distribute power, not just data.” Even a fully realized MyData future may not be inherently good. Relating to the “distributist left” or not is up to the reader, but the question remains: How can we engage with social systems to distribute both data and power? And if this is not the case, what will we have really achieved with MyData?