NGI Ethnography -- Preliminary Results for Community Members

First, some background on the ethnography team’s (@Leonie, @CCS, @amelia) process: we read (and reread and reread) the threads posted on the Edgeryders website. Then we develop codes for the threads that cover the content of the conversations taking place. Sometimes these codes are “in-vivo” reflections of the conversations (direct quotes from community members), sometimes they draw from established terms in the (academic) debates about Artificial Intelligence (AI), Blockchain, Cloud Technologies and other Internet ABC’s (for a detailed description of our coding see our open codebook or feel free to shoot us a message on the platform). To ensure the transparency and methodological rigour of our work, we keep detailed memos in which we explain the decisions we made, identify concerns, and define paths forward. We also have monthly calls in which we discuss difficulties, code-collapsing, and relevant developments in the world of tech and culture (of which we also keep a log in the codebook).

Over the past few weeks the NGI Forward ethnography team has been focussing much of their attention on threads pertaining to issues on the internet, tech and the environment. Our work on these threads has brought up two key insights that will be useful to keep in mind as we develop our methods.

First, those of you community members interacting on these threads appear to largely represent experts in tech; meaning you either seem to be conducting research, developing software, platforms, applications, leading groups of innovators and designers, or engage in activism and journalism around tech issues. This is something to keep in mind both for community managers and ethnographers, as it means we are observing and coding insights based on expert knowledge – we want to tailor our questions to you towards gaining a deeper understanding of emergent trends, policy, design and innovation. This level of knowledge and engagement on your part means we can ask you more detailed questions, as we are interested in your nuanced analyses of technology issues.

Interestingly (and perhaps conversely), explorations of the relationship between the climate and the internet/tech appears to be in its infancy, and though we are working with a community of experts, there often seems to be a struggle to solidly define key terms and concepts. Despite expertise, or perhaps because of it, specificity in defining technologies and technological areas requiring increased research focus persists across stories, meaning that you might find defining targeted and specific problems around these areas difficult as well. Also of interest, many of you find an underlying tension between acknowledging the ecological costs of the tech industry and findings ways to develop technological tools to protect the climate. These two points seem to characterise a caution on the part of the community when it comes to making bold or definitive claims about tech and the climate. They also indicate a desire to define the terms we are using to make sure we are on the same page, talking to each other about the same things, before we try to give solutions to problems. This may explain why codes thus far are more topical (sketching the broad areas that you have been interested in discussing) than analytical or solutions-focussed. We don’t necessarily see this as an issue, but instead something to track and pay attention to moving forward, as it may tell us all something about the challenges we face and approaches we take when trying to imagine the future of internet technologies.


Several current threads revolve around the notion of Deep Green Tech. While there is a lot of interesting discussion around this concept, it has yet to be clearly defined by the community. From what we can gather so far, Deep Green Tech is similar to Deep Tech, describing mainly startup companies based on substantial scientific research and innovations in tech engineering. Deep Green Tech, broadly, seems to describe technology that is ecologically sustainable, innovative and based on scientific advancements. The concept of Deep Green Tech appears to be a central area of focus for the NGI Forward community, but there seems to be ongoing discussions to clarify confusion around the term’s definition, its ramifications and implications. As a result, we want to help cut through the opacity of these discussions by working with you to understand how you define the term while also asking you to give concrete examples of its application, e.g. in design, research, tech innovation etc. In this way, we think we might all move toward a deeper and more productive discussion of the futures we want to see.

A notable tension that arises in community discussion on tech and the environment is that between navigating the ways in which tech/internet can help combat climate change and protect the environment and the energy and ecological costs that technological advancement has on the environment. We are interested in further exploring your understandings of this tension.

Finally, an initial visualisation of the codes pulled from the NGI Forward threads revealed a range of central, yet general themes. This may be again based on the cautious atmosphere of community interactions at this stage, but it also seems to be linked to the absence of concrete examples and personal narratives. We want to start to understand these issues on a personal level, beyond just drawing upon general and objective material. Going forward, we want to gather your expertise by asking more biographically oriented questions, which may help us tease out personal narratives and interpersonal connections around these technologies and help you share what inspires you to think the way you do about these issues. Asking you what motivated you to work in tech, how you use the internet on a daily basis, etc. can help us capture a more dynamic picture of the internet of humans.

Looking back on Open Care, it’s rare that people have detached and impersonal stories about healthcare experiences. We need to find ways to explore similarly impactful or meaningful avenues around the Internet— to ask you to share stories about experiences or opinions formed from experiences, as well as abstract ideas.

One of the central things we’re taught when doing the interviewing /question asking part of ethnography is to ask our participants to be specific by asking specific questions (this is what community managers on ER are always trying to do) — to move away from the general and ask what you did yesterday, to tell a story about a specific time something affected you, and so on. In Open Care, instead of asking ‘what do you think of healthcare’ it was ‘tell us about a time you interacted with the healthcare system— what did it look like? How did you feel? What issues came up for you?’

If we can find such parallels on the Internet (rather than ‘what does the future look like to you’?), instead ‘tell us about the last time something online upset you, and why?’ (or other questions along those lines, around sustainability for example) we might all get a deeper understanding.

In short, we hope to move away from asking how you think and feel in general and toward asking you about specific experiences, recollections that prompted emotional reaction, opinion forming, or action from them. We can’t wait to hear more of your fascinating thoughts and experiences on these issues!


@nadia @MariaEuler @inge

1 Like

Great, this is the rewrite for sharing?

1 Like

edit at will :slight_smile:

1 Like

Yep! I don’t know if @nadia wants to look it over first, but this is my take on what she asked for :slight_smile:

1 Like

Im good.

1 Like

Wow, super interesting!

First, a methodological point: the very high modularity that appears in the first picture (which represents all co-occurrences) is an artifact of SSNA itself. The cliques like the one highlighted in blue are induced by a single, dense post which is coded with 5-10 codes, which then, by definition, form a clique. On the other hand, the fact that the graph is connected (all nodes in the same in connected component) is not artifact of SSNA. The graph that has a more meaningful interpretation is the second one, where only the edges between codes that co-occur at least twice are preserved.

It’s just an intuition, but I expect this finding to be representative of the broader debate, not just of the NGI Forward community.

One for @johncoate and @MariaEuler!

This is also very useful analysis, to inform community management in the months to come.

I completely support this idea!

1 Like

I have a question regarding how we should try to gently push the development of a certain type of post creation as you touched on here. I fully understand the importance of the connection of personal experiences and examples to the abstract discussions. The texts that are developed from the interviews such as this one: Can tech design for survivors? How sex, violence, and power are encoded into the design and implementation of data/AI-driven sexual misconduct reporting systems can provide this especially since @inge did a great job getting it out. However, I would like to ask the Ethnography team (@amelia) on their opinion on the influence of the length of a post on its ability to create ongoing interwoven discussion. Do you have any insight on that? There are certainly posts that are too short to lead to meaningful discussions, but could it also be that if a post gets too long and dense, that kind of decreases the ability of the community to engage into a discussion of the different aspects since the discussion will develop along the lines of the first few comments which in return will become the source of following comments which will mainly zoom in on one aspect of the original post and morph that on further?

1 Like

Agree. Posts that are too long are also hard for people to respond to. I think what might be useful is to go to the Open Care page and sort the posts by most replied, and look at the posts themselves — check them out for length, detail, and content. If you get people talking about focused, specific things, chances are the length won’t be too long (and not too short, as long as the question is asking for a story or experience or example). I find that usually the question shapes the answer length, so asking the right scale and scope of question is key.

1 Like

@nadia and @inge, should we take this in regard when doing the longform deep dive interviews and correlating posts? Maybe making more than one post out of and interview? Could that be realistic? Or should we keep them this rich and dense? Maybe we could do the longform posts but than take 2-3 paragraphs to make into their won posts/ask the inved person to do so?

I remember that @johncoate has ideas on people’s attention in terms of post length, clicks, etc as well!

when you look at journalism, it doesn’t really matter, as long posts often get longer responses and are actually shared more: you weed out those who are just there to be heard themselves and those who care about a subject.

1 Like

Some thoughts about this, possibly disjointed.

Regarding length: I think that after about the first paragraph, and maybe just halfway through it, people tend to spot the writer their attention one sentence at a time. If something is interesting enough, the reader will stick with it, but even then often not to the end.

This is a big reason why journalistic writing was developed and why you shouldn’t “bury the lede.” They knew for years that most reader don’t go “past the fold” in a newspaper story unless they are on a long bus ride or in a waiting room. It is not so different online really. Except for the concept of click bait, which didn’t exist before. I think a lot of bloggers mislead with their headlines and then take as long as possible to get to the point so they will get a better impression count for their ad revenue. Also, they might just be lousy writers.

I know I was for a long time. Learning to write clearly and succinctly in this semi-spontaneous posting medium, while bringing out a compelling story worthy of a reader’s attention takes skill and practice.

But I very much like the exhortation that we continue to try to bring out more the personal side of people.

I think there are a couple of challenges in doing it. One is that the ER community uses the platform for quite business-like or project-specific purposes. It is not a social place or a hangout. This is by design for the most part, and just clearly not the preference of the main participants. The ER community does a lot of getting to know each other, socializing, meeting up, etc. but it is almost entirely offline and in-person. I think this works well for the regulars, but does not tend to create that encouraging of an atmosphere for comfortable personal sharing (stories) as a default. For my part, I try to throw a story or an anecdote into every post I make if I have one i think is worthwhile. And encouraging more of it in others is something I agree we can do to help enrich the dialogue.

The other challenge besides this ER platform not really being seen as a hangout, is the nature of the subject matter in NGI. Amelia describes the problem well. Experts likes to share their expertise. But we have two kinds of experts here: policy and engineering. I think the engineering participants do inform their work with what they see as an ethos compatible with NGI goals, but like the policy people, they come in and say something then go back to their project. Maybe asking for updates is a good approach to keep them engaged.

The policy experts for sure want to see things made that reflect NGI principles, but there seem to not be many of those projects out there to point to. Finding and highlighting them whenever we find them is, I think, one of the things the EC would like us to be doing here. Generally, the policy experts, and I see myself as one of them at least to some degree, point out the problems. I think we have done a good amount of that here.

Tunneling our way to solutions or attempts at solutions or raising the profile of individuals who the EC ought to know more about are big parts of what makes this project worthwhile.

I do think that the people interviewed by Peter and Zenna are all interesting and they do tell their own stories in those talks. Interviews are of course one of the best ways to draw them out of people.

As for deep green tech: very challenging indeed to define it and we did try to get some better definition of what the term even means specifically, but I think we need to try to get more clarity there.

And I like Amelia describing tensions inside some of these issues. I think none of us believe that tech is going to “save” us, but I also think that if we did not have at least some belief in clever tech solutions we would not be involved in these conversations. I know I feel pretty conflicted about it myself.


well those posts have a specific context/purpose - the practitioner workshop on AI and Justice. We base the development of case studies on these deep, rich posts. And it is a deeper conversation amongst people who are very interested in these topics that will generate value for everyone.

Perhaps an approach to try is to develop one page articles presenting the synthesis we make (i.e the cases) with a question and invite more people to contribute to that conversation?

The other is a series of solarpunk stories that take on a more personal entry point to all this…

1 Like

Pinging @mdroemann to see our current insights.