Session proposal: ethics and data protection in citizen science/open source projects

Workspaces LOTE6 “OpenVillage Festival”
, , , , , , , ,
1

Calling mostly @WinniePoncelet .

I propose that the mighty @markomanka leads a session on ethics and data protection  in citizen science/open source projects. Regulation on ethics in research and data protection is taking shape, with the GDPR coming into force in less than a year in the EU. Marco thinks there is potential for this new regulation to make anyone except very large and well-moneyed orgs viable to do research and data protection. This would be done by making the costs of compliance very high.

Example: if you use algorithms (say, a recommendation algorithm for books) you are supposed to be able to make a case that it is ethical: you are not, for example, systematically recommending books by white authors over minority ones, or male authors over female ones, whatever Bayesian updating might have to say. Which is fine, but where does this leave open source? What happens if I install WordPress on my server, and WordPress comes with a search algorithm? Do I have to audit the code to be compliant with the GDPR?

The outcome of the session could be an ethics/data protection wiki for citizen science projects in care: guidelines that tell projects like OpenInsulin what to do when.

What do you think?

2

Relevant!

This is a tough subject for citizen science projects to look into themselves. It is relevant for those who are already further in the project, like echOpen or Open Insulin, or those starting out, to keep it in mind. When doing something radically open, the last thing you want is to compromise the project by making avoidable missteps, legally or ethically.

I like the idea of a wiki. How would such a session look like? Is it a documentation effort of what can be found on the web, or experts sharing insights and projects judging the relevance, or something else?

If needed, are there other experts to invite @markomanka ?

3

Let Marco explain this

@markomanka , what do you think?

I would certainly welcome an intro to GDPR with an open source perspective.

4

Very relevant

Hi @alberto,

as discussed on hangout I find ethics, in general, to be a very powerful learning point for most, if not all, the open care projects we have met. Well beyond the challenges of GDPR, there is in general a certain semantic confusion about what “open” would mean, and a lack of self-reflections about where value is produced, and for whom… Defending ethics in the absence of such fundamental insights is often challenging, and most conversations get framed as box-ticking before adequate discussion and provocation.

That being said, I would like to ask @winnieponcelet (we can discuss the details tomorrow during our call) to evoke an expression of interest with real questions/stories by the groups that would join this session at the Village.

In facts, we have to be realistic about the time limitations at the Village, and what we can usefully discuss there and then…

On a related issue, I would like to collect questions for the ethics board, from the consortium partners (@Lakomaa , @Rossana_Torri , @Alberto , @melancon , @Costantino ), and from the guests ( @WinniePoncelet ), to hold a private, and a public question times with our advisors… most likely via hangout/skype, during the Village.

Collecting questions/topics is fairly important, as of course the advisors need time to understand the proposed issue, and prepare a conversation. (Let’s give a deadline of at least one month ahead of the Village?)

Talk soon,

5

Ethics of public spending

In CoM’s experience, ethics and data protection in public / open source science projects (@Alberto) belong to public spending decisions. Since one of the major leverage that public administrations can use to include private partners in public policies is to finance them with taxpayers’ money (to keep it simple, I will not mention PPPs, which also have pitfalls), they need to make rational, fair and transparent decisions. This is where public procurement rules apply. However, strict rules often lead to standard, controlled and predictable output. When it comes to innovative_community-based_care related policies, effective solutions can come from spontaneous practices that we have to approach in a different way (what if they do not know there are public funds?), we have to support in a different way (what if they can not write projects proposals efficiently?) and we have to relate to in a different way (what if we don’t understand their languages). In other words, we might be fairly-ineffective or efficiently-unfair. I hope this isn’t too much off topic @markomanka.

6

That’s the kind of provocations I am looking for

Hi @Matteo

thank you for your reaction.

That’s absolutely not off topic, and quite in line with what we have been observing/hearing during OpenCare. Another side of this, correct me if I am wrong, seems to be the fact that most of the times once public institutions get wind of the interesting bottom up activities around them, they often try to replicate/absorb them at face value… investing in a snapshot of what seems to be their current value proposition, but failing to capture the path that attracted momentum, and their value chain.

In facts, many initiatives have little EXPLICIT awareness of their values and situation, and it is perfectly fine for them, as they run on experience and shared stakes… but transfer/reproduction efforts complicate things quite a bit…

…on the issue you point out… Could one challenge be the denial of the need of more than just one model to describe reality?

In care this is quite problematic… bureaucracy tries to optimize for one model, typically fine tuned around middle tendencies of distributions in public health, but this operation introduces important fragilities, as eloquently discussed by Nassim Taleb…

Looking forward to your further considerations :wink:

7

Like this

It latches nicely onto work we are doing on a different project, witn UNDP: @Hasmik , @gazbia-sorour , @Tinatin , @Inge , @Max_Perry , @LilitMidoyan …

The topic is different from care: we are looking at bottom-up urbanism, city-as-a-wiki etc. But it is still community-based, and seeks to empower community capability and enable it towards public good goals. This is very hard for local governments to do, because giving communities space requires, paradoxically, for them to behave in a way that might appear arbitrary. Why are we letting people do things in this park, but not that square over there? Because the people who are interested in this park are aligned with the spirit of contribution to public good, and they have capability to deliver. Who says so? We do. This is the polar opposite of @Matteo 's “strict rules often lead to standard, controlled and predictable output”: if you, as a government, open yourself to bottom-up innovation you pay a price in terms of strict rules.

In David Graeber’s book on debt there is an aside that stayed with me, and might solve the policy maker’s dilemma. It is well known that Imperial China invented bureaucrats. The Mandarins were civil servants, centrally selected on the basis of merit and centrally trained. But they were not given a rulebook to apply. China is big, and communications were poor and slow; also, it was very unstable, with peasant revolts sprouting out every year. The Emperors knew that Mandarins would be on their own in distant provinces, and they would need to take quick action to prevent and quell revolts. So, they were trained in Confucian ethics rather than “law” as we understand it, and instructed to do their best to make sure the people were content and did not starve.

In modern terms, their operating mode was result-oriented rather than process-oriented, and buttressed by ethics rather than rules. I think Amartya Sen’s work shows that this would be enough to ensure accountability. Which means, governments could, in principle, be hierarchical and creative and accountable. But that’s not easy to do in process-oriented, formalistic legal systems and cultures.

Is this something we could be discussing? Maybe invite someone from UNDP?

8

Public spending: how and where?

@Matteo I remember a discussion arising from this post. What is the role of government? The ‘operating system’ plays a major role… It’s almost absurd to expect a bureaucrat doing full time bureaucracy work to speak the language, know the needs etc. of projects that are relevant. Yet if the bureaucrat spends more time on oversight and less on command & control, leaving the latter to the people who have experience in the field, it’s probably more effective and cheaper.

Another thing I’ve been thinking about in the context of Open Insulin. Leaving aside if Open Insulin is the right approach, consider any project that is a long shot, ambitious and with a potential big impact on many parts of the world. It’s hard to disagree (in my opinion) with long shots for systemic changes in eg. pharma industry, based on ethics. I haven’t met anyone who is against it in principle, also not in government.

Though only a smaller part of such an impact is realised in the city where the project resides. In my experience, a city wants to support projects that have an impact in their city, the impact beyond that is of minor importance. Or, more extreme, they will deliberately not support you so that you go search funding up the chain, such as regional grant, because that results in an inflow of money into the city from elsewhere.

Not so popular to be funded by the city, due to the above, but also because they would rather support projects with an immediately visible impact, such as helping disadvantaged children or establishing a development aid link with an African town. This is of course great, but it is a form of symptom treatment, and for the cynical, mainly about the funder’s next election cycle. Policy now seems more about adding new rules and exceptions to help very specific groups who have been disadvantaged, rather than taking ambitions decisions that address the root and affect everyone - from the same disadvantaged to the shrinking group of privileged people - to change the legacy of outdated rules. It should be a balance, naturally, but it’s way skewed to the wrong side now.

A trick is to tie some concrete city level impact to the project - I guess OpenRampette would be a good example. Yet this is not always possible, especially in more complex projects.

Higher up there’s also close to no room for eg. an Open Insulin in more formal and bigger funding programs. Rigid structures, but also lots of lobbying going on from, in the case of Open Insulin, big pharma. There, your chances are tiny and attempts are expensive. Ping @Lucy , maybe you can also pitch in.

Personally I see potential in the role of the city, just because there seems to be more room to be creative for those who dare. These long shot projects are not expensive in money either, especially considering the potential outcome. If five girls & guys in their garage can have a shot at changing an industry, imagine what a city could do. How badass would it be to read in the newspaper: “The city of Milan supports open source treatment for diabetes induced blindness”.

So should a city support what emerges in their city, even if the impact is proportionately realised mainly outside the city? Considering they are dead in the water elsewhere, is it ethical not to support these projects at a city level?

1 Like
9

Maybe…

Maybe, @Gehan, you might have a word of wisdom for me on this topic? :slight_smile:

10

I would like to invite a strategic decision at this point… would we rather like to follow a “seminar+clinics” scheme, as to offer an hands on reflection on real study cases to the participants bringing up their experiences, or running a conference style debate on the topics, with experts of diverse background?

For the latter, I would like to import the form of open peer review traditionally adopted at CERN when taking community decisions: anyone has the right to speak/question, and the speaker “survives” only if the reply is deemed satisfactory… since the audience can interfere, and “counterattack” those asking questions, fairness is pursued by skin in the game, rather than gentlemen’s moderation.

I am quite agnostic about what this session should look like, as I believe any of the alternatives discussed so far with @Alberto and @winnieponcelet would bring important learning opportunities, but a call should be made as early as possible.

My 2 pennies of wisdom: although it takes a bit more to arrange the clinics, they tend to be smoother, because people feel challenged in the 1-on-1 meetings, but pride doesn’t enter the game too prominently… In a CERN-style debate, it is necessary to make sure everyone feels their weltanschauung is equally at stake, and people who have invested large life-capitals on certain ideas, will vehemently try to fight even once the audience has already decided on the sentence of the contestants.
I have had the opportunity to discover the latter is a show not to the taste of everyone…

Your calls?

11

Not my call to make… but I would be really curious to try CERN style review.
Even though it does seem like a methode to decide something. What are we deciding at OV?

12

I second @alberto , I like the idea of doing the CERN style review. This won’t allow us in depth 1-on-1 meetings, but it will allow more people to share in the insights.

Can we do a combination perhaps @markomanka ? The topic of the CERN style debate can be a specific problem of one of the guest projects. Experts in the debate, projects in the audience to interfere.

It hopefully then approaches the usefulness & specificity of the ‘clinic’ format, allows for broader interaction and discussion solutions to specific problems should decrease the influence of ‘life-capital investments’.

There is some preparation work then to find one or two projects and formulate questions well. Those projects ideally have presented themselves already in a different session, before the debate.

Shall we move forward with this?

13

Maybe we can set the expectations and purpose a bit better. I hope to learn about:

  • Requirements for data protection and ethics (legal, like the GDPR, and otherwise – for example ethical oversight is required by many funding agencies).
  • These requirements were dreamed up with universities in mind, so: how they map onto citizen science projects. Badly, I expect.
  • Ways and hacks that citizen science projects can comply with those requirements.
14

Hi @winnieponcelet,

as stated before I don’t see the two formats merging, as the principles of clinics, and total-peer-review, are quite conflicting, as it is the designed relationship of power in the two. In a CERN style event, we have to make sure a majority of the “parterre” has own projects, and stakes, that are touched by the conference topic, but there will not be 1-0-1 engagement, rather all will share their skepticism, and doubts, to converge (hopefully) to viable solutions, or to clearly mark as “BS” all that doesn’t stand peer-review…
Trying to have also a clinic in this context, would defeat the goal of maintaining everyone constantly on watch, peer-reviewing and critiquing/re-framing what is said… imagine a University lecture, at the end of which students wait for the talk to be finished, to approach the professor privately, rather than asking/commenting in public… we don’t want that.

“Speakers”, here, are sort of provokers, offering high quality insights into what’s going to be made, that has to be digested by the conversation of the “public”. Presentations should be short, and to the point, as appetisers for the question time. No project-specific answers will be sought (not by design, at least).

15

Hi @alberto

I hear you loud and clear, but you are not the committer of this session. I would like to collect the same kind of statement about expectations from (at least) all the projects that will come to create the OpenVillage (@noemi @winnieponcelet).

GDPR does not seem a hot topic in our conversations about ethics with the projects/communities we met so far (I tried to comment on this yesterday during the call, but @melancon mistook my opening for a reference to WP7 and I did not want to waste time in polemics)…
Furthermore, people in the legal field I regularly meet, have at the moment conflicting opinions about what to be expected from the new regulations that will soon take course… in facts, taken by the word the new regulations are clearly informed by a rather advanced view of IT, and they would seem to favour a migration to large platforms (please remember EU is working on an “open science cloud” of its own)… however, most legal professionals point out to the fact that the effect of any law will be shaped by the first (and later by the major) sentences by Courts… most envision some leniency in support of a realistic application of the law. An event centered on GDPR will happen, if the community wants it, but I feel it is a bit too early now for it to be anything more than a reading of the norms.

However, ethics in citizen science, something way larger than data handling, is a very appealing topic…

Does this make sense to you?

16

Oh, absolutely. I was just proposing an example of how anyone can declare interests, so that a specific format can be chosen.

1 Like
19

This discussion could easily constitute multiple session proposals rather than just one! I caught it last week but I’m coming to a response late from other things. I don’t know anything about GDPR or much about data protection in general but the discussion touches on a number of topics that interest me; policy re-designed as a tool for a networked and complex world and what Illich would refer to as the ‘radical monopoly’ that our procurement systems create.

But it’s another of Illich’s concepts that seems to be the thread running through this conversation - the systemic problems that arise from growing institutionalisation. As Marco mentions “bureaucracy tries to optimise for one model” and for me this is the nub. It is entirely unsuited to operating within systems thinking perspectives and perhaps this is what is called for beyond creating multiple models.

Systems account for and allow a greater degree of variation and adaption unlike as you say, bureaucracy that seeks to encourage standardisation because this is one of the conditions it needs to operate. This is also how I would understand why institutions can’t replicate bottom up approaches @markomanka - they are founded in two completely different operating systems to use Winnie’s term. It strikes me that the ethics training of Mandarins that @alberto mentions facilitated operation in a living system - and therefore able to cope with the infinite variations that played out across Imperial China. Ethics and data protection compliance are the symptoms in this context but to my mind, the underlying malaise is hyper-institutionalisation.

I also see another dilemma unfolding in the discussion that may be helpful to tease out. There would appear to be an inherent tension not just in this thread but also within Edgeryders itself between responding to immediate needs and contributing to generative creation of completely new approaches and responses. For example there is some necessity in meeting immediate needs of those undertaking Opencare work - to pull in resources to support their work or to understand the implications of future changes such as compliance. At the same time, there is a pull - to work on the ‘long shots’, the longer term, bigger impact projects that @winnieponcelet mentions.

A frame I’ve found useful is Three Horizons. Briefly: Horizon One (H1) is the dominant system, business as usual. Horizon Three (H3) for me represents a different paradigm - or a completely new way of doing things. Horizon Two (H2) is the space of innovation and this can go two ways: either prolong H1 (H2-) or open up H3 (H2+). It strikes me that Edgeryders and OpenCare are operating across all three horizons. Horizon One has been outlined in the proposition for OpenCare as rising health care costs and care systems struggling to meet demand. In exploring DIY welfare, OpenCare researches activity that falls into H2 stretching into H3. I’m not totally clear yet whether the underlying goal for OpenCare is H2-; that is enabling the current way of doing things simply to function better or H2+/H3 - paving the way for the ‘completely new’?

So in this discussion there are the immediate H1 needs of projects to sustain their work. What role does OpenCare (or might a session at OV) play in meeting these needs? But beyond this is there also a role to call for structural changes that get to the root of the problem? And this is where we hit up against mutually exclusive sets of interests. I’m curious as to how OC and EE sees itself in this regard?

1 Like
20

@gehan: the three Horizon model is interesting. Ethics and data protection in citizen science project, however, are not easily classified as H1, H2 or H3. No one seems to care. Data stewardship is normally dismal. Many projects (of any kind, not speaking about citizen science in particular) are playing cowboy. In all this, H1 requires compliance, this is true. But compliance is not where the action is. The action is in self-motivated actions to do things in fairness (in ethnics) and steward the datasets that people helped build together (in data protection).

So, you can take action because H1 demands it as a condition to fund your citizen science project (H1): in this case you will just tick boxes. You can do it to stay fundable, but while you are at it improve your practices a bit (H2-). You can do it mainly to improve your practices, and it’s great that you get extra fundability (H2+); or you can do it because you believe that a world that puts ethics and data stewardship in research is fundamentally different from one that consumes people and data and moves on (H3). In practice, in H2-, H2+ and H3 you will be doing the same things, at least initially, because conditions in H1 are so poor.

21

Let’s make the call then and go with the CERN style review. We can brief participants so that it is generative, as it does seem like a more direct approach that can degrade into a fight instead of a search for insight.

I remember @breathinggames has questions on legislation. Same for @dfko from the Open Insulin side (also what can be generalized across the US and Europe). @olivier also has similar questions on certification and safety.

We’ll be on the lookout for experts to join the session. Do you have a sparring partner (or multiple?) in mind @markomanka ?

22

It’s all good… but are we the only ones engaging in this decision making? Do the people pinged so far all share the same opinions? They did not even reacted to the posts with a “like”…

Is it that, maybe, people would rather have an intrusion of legal/ethical discussions within the other sessions? We could still think of something interactive and pragmatic, if anyone shared their desires and/or frustrations about what we propose…