Dec. 6th, 2018

greywash: A lounging pinup girl, holding a cocktail. (Default)
[personal profile] greywash

I'm making this as a sort of a placeholder post because an issue that keeps coming up—both in things I want to say (in comments to [profile] recentlyfolded's mod posts from yesterday) and in private discussions with other fans—is that what we're discussing is on the bleeding cutting edge of both technical and technology-and-society research. The thing I want doesn't exist; it might port from some technologies that do exist, but it's not a social network, it's not an archive, it's not a data center—it's some weird chimerical beast that draws from all of those things, and also is going to require input from experts on lots and lots of things that I, at least, am not an expert on.

I would very much love to crowd-source some research in, and also ideally get in touch with some fan-friendly academics doing research in, the following areas (WIP):

  • Computer science:
    • Machine learning: supervised learning for (e.g.) abuse detection in both text and images; community-owned machine learning data sets; machine learning that is transparent and responsive to community stakeholders.
    • Distributed systems: distributed networks/"the swarm", archival concerns and persistence of data in swarm computing, legal concerns and privacy in swarm computing.
  • Technology and society (I feel like this department goes under many names, so—): the internet and the historical record, machine learning that is transparent and responsive to community stakeholders.
  • Library science (I think?): archival concerns and persistence of data in swarm computing, balancing archival concerns (citations/a persistent record) with privacy concerns (the right to be forgotten), legal concerns for material stored in libraries that are accessible world-wide (when the material may be illegal in some countries).
  • Law: legal concerns in swarm computing, privacy in swarm computing, legal concerns for material stored in libraries that are accessible world-wide (when the material may be illegal in some countries).

Please leave a comment here (comments are screened) if you are or know of an academic or academics doing research in these or other relevant areas I haven't thought of. If you are okay with me unscreening your comment, please actively say so in the comment itself; otherwise, your comment will remain screened and be viewable only to me and the other mods. However, if the information you reference is publicly available, I will link it in the directory (below) without referring back to you.

For example, if you say, "Dr. Jane Foster wrote this great paper on swarm computing, and has also written about the fannish internet here, , her contact info is available on the Unseen University website here: http://whatever - please don't unscreen my comment" — I would post this as "Dr. Jane Foster: "Title of swarm paper" - URL, "Title of fan paper" - URL," with no reference to your username; and I would contact Dr. Foster privately to explain what we're working on and ask her if it was OK to link her contact info before I did so (if I did so). You do not need to tell me whether or not you are, yourself, Dr. Foster; I will assume that you are not her, but you want to remain anonymous, and I will not mention you when I contact her. Research Directory )

gclane: 19th century woman, looking stern (Default)
[personal profile] gclane
Hello. I'd like to float some ideas and see what people think. 

I've read all of the posts, particularly recentlyfolded's issue-by-issue breakdown (so good!), but I haven't branched out into the various links that have been shared. Apologies if I repeat what's already been discussed/decided on elsewhere.

What we have here is an information problem. And, when you have an information problem, the current state of things invites you to think about it like a developer. Our aspirational models are large ICT (information communication technology) companies. Which, I mean, of course - developers know how to solve information problems in a way that's satisfying to their employers and often satisfying to the general public. Their employers make money, money drives attention, so we're all very attached to their version of solutions to problems of information. We're too attached, though - we've really begun to think that these large companies have the best/only solutions to information problems. At least until one of their solutions jumps up and bites us, as it has with Tumblr. (ETA - I've gone back and read my own words and I'd like to make it terribly clear that I have big respect for developers and the many things they are capable of. I think, to put a fine point on it, I have less respect for their employers and people with an excess of decision making power in the places where developers often work.)

I'm a librarian at a very large institution with a lot of money (relatively speaking). My chosen career and where I work means I get to spend a lot of time in rooms with brilliant people who think about information problems in ways that entities like Google and Facebook and Tumblr wouldn't recognize. And, in the past couple of years, a recurring theme of these conversations in these rooms has been bemused despair at everyone else's total shock that Google, Facebook, and Twitter are politically compromised, bigoted, nightmare pits. Because of course they are! Mark Zuckerberg didn't have a single thought in his head beyond all the money he could make when he started expanding Facebook. Google's mission statement and assertion that they aren't evil are laughable. These entities exist to turn a profit, which is just fine - capitalism, right? - but it is their central motive. It is what they wanted to do when they sat down and decided to put on a show. And, as many people have identified, that's what's happening with Tumblr right now. Their porn problem got them kicked off the App Store. They want to be on the App Store because that's where the money is. Ergo, in their heads, the porn must go! But they've never given a thought to what porn means. Or who might be implicated in their choices. Or how their choices might land differently on different communities. Or what a real technological solution to their information problem might look like.

They do not want what we want and have not thought about it in the ways that we've thought about it. Because, really, based on what I've seen here and the critiques I've read on Tumblr, we want to think about it like librarians. We're not foregrounding profit or uncomplicated solutions to difficult questions. We're doing the actual work of solving an information problem. 

All of this is to say that I think we'll be most successful if we focus on what we want, rather than what the solutions look like. When librarians solve information problems, good librarians anyway, we tend to foreground the following - 

  • Access - big, complex, lovely access; not "eh, it's in there somewhere, let's pretend our shit tagging system and mysterious algorithm constitute access"
  • Privacy - How does a user put information in? How does a user take that information back out? How much distance can a user put between themselves and their identity? What if someone uses that distance to do harm? How easy is it to connect someone to their use of an information system? How often is the record of use destroyed? 
  • Marginalized Users - How will any decision made fall on the most vulnerable user of this information system? What will we do when one of those vulnerable users tells us we made a mistake and have harmed them? 
  • Education - How do we bring new people into this information system? How do we teach community norms? How do we teach people how to use the solution we've devised to our information problem? 

Notice the lack of reference to profit, shininess, or scale. I think this is why AO3 keeps coming up in these conversations. They've done this work for the particular problem they wanted to solve. They don't want to do the work of a social network, which I completely respect, but their solutions are very... librarian-ish. And I really appreciate that. 

That's what I've got. I hope it helps. 


Some discussion space for where we go next.

March 2019


Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags
Page generated Apr. 18th, 2019 02:26 pm
Powered by Dreamwidth Studios