Emergent Biases

Tay AI
Image from Business Insider UK.

Librarians are (or should) always be thinking of ways to make information and information systems more accessible to more people.  Which can be quite a challenge.  (Databases, anyone?)

Information systems and technology always have biases.  It may be that a technology is biased toward sighted people; for example, Google’s Calendar function has come under fire for not being accessible for people who are visually impaired.  It also might be that the creators have biases, and those appear as pre-existing biases. For example, many dating sites only offer two gender (male/female) options when gender identity and expression are much more varied than that.

Then, there are emergent biases.  These biases do not exist (per se) in the technology straight “out of the box.”  Instead, these biases emerge because of the interaction between the technology and the users.  The smarter, the more interactive, the more self-sustained that technology becomes, the more common these emergent biases are.

Last week, a technology with quite the emergent bias came into the media spotlight.  Microsoft created an artificial intelligence (AI) which would communicate with human users via Twitter.  This AI was named Tay and was meant to mimic a teenage girl.  She came pre-programmed with “teen girl speak” and would learn ideas and ways of communicating through interacting with human users.

Well, as you probably expected, Tay became a Neo-Nazi sex doll.

Microsoft didn’t program Tay to be a white supremacist or a sex object. (Or at least not immediately.)  But they did program her to learn from human users.  And when the human users are trolls, misogynists, and white supremacists, Tay would (and did) quickly learn to mimic her human companions.

This case of bad parenting for a teenage AI created a technology that creates a hostile environment for a (very large) community of users.  It’s a shining example of emergent bias.  When making smart technology, help it to be smart enough not to acquire biases.

Advertisements

Why do college students use the library?

CENGAGE Learning blogged about why university students choose to use the library – and also why many choose not to use the library.

I know I use the library now because I’ve bought into library science, but even before I had the slightest desire to be a librarian, I used my university’s library.  I always found that the library was the safest bet because it provided high-quality materials that I knew I had immediate (and free) access to.

girls going to school
“Children in India Girls going to school” by Sandra Cohen-Rose and Colin Rose, CC BY-SA 2.0.

At the same time, I understand why some users forego the library.  Resources like Amazon and Google are fast and easy.  And even as a librarian, I’ve found that Google Scholar can be an excellent link resolver when your own systems are clunky and not working.

So, why do students use the library? Or why don’t they?  Here are some highlights from CENGAGE Learning’s blog:

Why do university students use the library?

  • They trust the quality of the library’s resources. It’s easy to know that most of the materials are scholarly and peer reviewed – great for college assignments.
  • The library staff is there to help. With how much effort I know librarians put into outreach and getting people to ask for help, I’m glad that at least some students have heard and appreciate the library’s services.
  •  The library has technology and internet access. Having access to computers and stable internet are great draws. My undergraduate university had an information commons with tons of computers – and access to printing. I understand loving the library’s technology.

Why do university students not use the library?

  • It’s easier and more convenient to use something else. Students are much more integrated with services like Google, which can be an effective search tool (duh).  And if you’re networked to a university’s internet, Google can still pull up electronic resources for you – without seemingly having to use the library at all.

Pornography in the Library, Part III

While I openly admit that I’m interested in pornography collections in libraries, I did not expect to find so much material to write about so quickly.  So, here’s another installment of Pornography in the Library.

I am currently in the Information Policy course at the University of Illinois Urbana-Champaign‘s Graduate School of Library and Information Science.  The course talks about all sorts of ways that institutions (governments, schools, universities, libraries, etc.) create policies surrounding information.  For that class, I decided to do a little scanning of the literature related to library policies concerning pornography.  I was surprised to find how few resources focus on pornography and libraries in general (rather than the restriction of child pornography, or restrictions on children’s room internet access).  (Apparently, people don’t want to write about things I find interesting.) The most comprehensive and up to date monograph I found was Libraries, Erotica, & Pornography, edited by Martha Cornog. (And when I say “most up to date” – this book was published in 1991. Yeah, apparently porn and libraries aren’t exciting.)

cover of playboy magazine
Cover of first issue of Playboy Magazine, via https://en.wikipedia.org/w/index.php?curid=12530828.

This post is not a review.  Instead, I wanted to highlight statistics on the iconic serial Playboy in libraries, which Cornog includes in the tenth chapter of Libraries, Erotica, & Pornography.  Sure, they’re a little old – but they still give some fun insights into how libraries cope with their erotic collections.

Here are some fun facts on Playboy in libraries:

  • Libraries most often subscribed to Playboy due to the quality of its content. 34.2% of libraries surveyed listed this attribute as the reason for their subscription.  20.7% subscribed to Playboy because the magazine is indexed (147).
  • Academic users have greater influence over the selection decisions in libraries than public users (147).
  • 88.3% of subscriptions were purchases, rather than gifts.
  • 42.3% of libraries with Playboy subscriptions reported staff complaints; only 25% reported patron complaints (148).
  • 2/3rds of libraries with subscriptions to Playboy restricted access in some manner.
  • Theft seems to be a problem, but it’s usually not the patrons – but the library staff who seem to walk off with a sexy magazine (151).

I always find it fun when it’s the library itself (the staff, the institution) that has issues with pornography – when the users don’t always seem to care.

eBooks and Money

I recently discovered the blog Good e-Reader run by Michael Kozlowksi, which aggregates news related to ebooks.  As I was perusing the site, I came across posts dealing with ebooks and money.  Being in acquisitions, I’m all about money and electronic resources – so this information stood out.

chopped up money
“Money” by Tax Credits. Attribution 2.0 Generic (CC BY 2.0) via Flickr.

Here are all of the different rates publishers charge libraries” lists what the major publishers charge libraries for access to ebooks.  My institution has the money and the clout to buy academic ebooks outright, for the most part, but most libraries have to lease titles.  And popular titles become even harder to acquire, and usually only through a license (i.e. the library cannot buy the ebook).  And these licenses are pretty expensive – like Hachette’s three times the price of print.

So, while I know that publishers seem to be stealing money from libraries with their outrageous pricing, I found it interesting that publishers are reporting plummeting ebook revenues.

How is that happening?  I cannot say for sure, but I have a few ideas:

  • For people reading creative literature and for people who need to read deeply, print is preferred.  (We see that in my library.)  Perhaps more of the marketplace is reading creative literature or needs to engage in deep reading.
  • People are reading less in general, so all sales would go down.  Or a certain demographic is reading less, which affects ebook sales.  (For example, younger readers who would use ebooks are reading less, while older people continue to read but prefer print formats.)
  • eBook pricing is too high.  Though ebooks take the same amount of resources to produce and market as print books, many readers feel that they get less value with the ebook – thus, perhaps they are returning to print because they perceive that they get more value for the same money.

I don’t know for sure.  These are just conjectures.

But one thing is for sure: it’s always interesting in the land of electronic resources.

Right of First Sale for eBooks?

eBooks are quite the conundrum.  Because they’re digital copies of an “original” digital file, most readers don’t own their ebooks but are just licensed to view the copy.  So, readers can’t lend out or resell their ebooks, like they could a print version. But that might change.

Since 2011, Amazon has been filing patents in the United States and Canada around setting up a marketplace for used ebooks.  In January 16, they filed a 46-page document with the Canadian Intellectual Property Office which outlines their plans in more detail…

If Amazon is able to set up this used ebooks marketplace, then we might be forced to rethink how we view these digital copies.  Read more at my guest post at Libfocus.

Luna the Library Horse

I am a horse-lover.  So, when I see horses contributing to, interacting with, visiting, etc. libraries, I always have fun.

In Indonesia, a horse named Luna acts as a library.  Luna and her handler Ridwan Sururi provide library services to communities in the remote region of Purbalingga on the island of Java.  Their library is called Kudapustaka (literally, “horse library”).  They travel to schools three days a week and provide books to children and adults.

Indonesia is working toward increasing adult literacy rates.  The government brought the number of illiterate adults down from 5.4m in 2004 to 6.7m in 2011.  I imagine Ms. Luna is helping out.

The Trouble with Technical Services

Rebecca at her desk
“Rebecca, having threatened to take a cellphone picture, at her desk” by Nada Sweid [2016].
Several months ago, I created this website in order to create an electronic resume and portfolio to aid me in professional development and career advancement.  Sure, the portfolio and the home page was fun enough to create – but the much more fun and engaging part of this site has been continuing to post on the blog.  The blog on this site is truly for me to post about whatever I want, so long as it relates to librarianship.  Librarianship encompasses so many things, that it’s rare that I have writer’s block.  But I’ve found that it’s very hard to write about my area.

LGBTQ services, outreach and marketing activities, privacy seminars, technology gaps, access for underserved populations, emotional labor, approachability (or lack thereof), radicalism – those are a lot of fun to write about.  You’ll probably see posts on these topics here at this site.

But acquisitions workflows? theory? policies?  Cataloging beyond slamming (rightfully) the LCSH?  That’s not nearly as glamorous, though it can be equally fascinating (I think).  And it’s much more difficult to write about.

So, the trouble with technical services is: it’s hard to communicate about what we do.

There are so many reasons for this.  One, tech services is very…technical.  I know so many acronyms and synonyms and technical terms and jargon and product names.  And my reference and instruction colleagues just don’t speak that language; they speak the language of the patron (as they should).  I find that I’m so integrated with acquisitions speak that I have trouble finding more layman’s terms for my work.  I’ll always say SFX rather than the Online Journals & Databases function, OPAC instead of catalog, CARLI instead of I-Share or the library consortium.  I’m speaking a whole different language sometimes.

Then, so often the software tech services people work with proprietary software.  I cannot legally go on my blog website and post screen captures of a lot of my work – because ProQuest owns Ex Libris Voyager and strongly discourages displaying their product without their permission.

And while most of the information I work with is public knowledge, and people are welcome to FOIA it, I feel like airing out institutional finances just isn’t appropriate. Unless I’m whistle-blowing, which I haven’t had a reason to yet.

Then, some of the stuff tech services people have to deal with…we don’t know how to explain these things to ourselves, let alone other people.  I love working with electronic resources…but they’re crazy.  They’re lovely when they work, and a nightmare (sometimes) when they don’t.  And I’ve never met someone who can coherently explain electronic resources.  The things are just so twisty: every single one seems different; the DRMs are important but then you have to coherently explain DRMs; what is SUPO or MUPO?; copyright makes all sorts of stuff crazy; pricing is out of this world.  Electronic resources just aren’t coherent enough to talk about.

Don’t get me wrong, tech services are awesome and I am so glad to work in this field.  But communication with the “outside world” is the trouble with tech services.