Libraries as AI Model Repositories

Libraries and librarians have a critical role to support and enable explainable artificial intelligence (XAI). Calls for transparency and oversight of the opaque and potentially biased machine learning and deep learning systems have increased as the extent and influence of AI has also increased.

However, transparency often conflicts with the quite reasonable desire for the protection of intellectual property. Hence many regulations (e.g. the EU’s GDPR) have carve outs for IP protection dramatically restricting some XAI options like public access to code or algorithmic models. Third party oversight of AI (such as auditing) is a widely promoted XAI strategy but concerns about secure access are problematic.

Lilian Edwards and Michael Veale (Enslaving the Algorithm: From a ‘Right to an Explanation’ to a ‘Right to Better Decisions’?) offer an interesting option: libraries as secure places for model repositories. The idea resembles how access to sensitive Statistics Canada data is provided via Research Data Centres in academic libraries. The result would allow qualified and certified third parties access to the AI models and training data under conditions that would protect IP.

Three observations from this:

  1. Since third party auditing is a highly favoured strategy for XAI, I’m supportive of anything that furthers this.
  2. This suggestion establishes a new role for libraries as AI repositories. These collections will require increased sophistication in terms of collection management. However, their near-term value as auditing resources and their longer-term value as research collections makes this new role important.
  3. This proposal explicitly identifies libraries as secure and trusted. It selects libraries over other possible sites for such an archive. Edwards and Veale believe libraries are (or can be) ultra secure and have the mandate and capacity to be so. We have a track record with Stats Canada, although this one may require more diligence.

Of course, this option would almost certainly come with additional funding since libraries would be part of a formalized auditing regime mandated through regulation or legislation. For the forward-looking library (or library consortium – I’m looking at you OCUL), this could be an important opportunity and the beginning of a research collection of unprecedented value.

…Mike

Share
Posted in AI | 2 Comments

On Not Travelling to the United States

In 2016 when several US states passed what have become known as the “bathroom bills” prohibiting transgendered people access to appropriate facilities, I joined in on the outrage. I tweeted about not to travelling to North Carolina and other states, withholding my tourist dollars, in protest. Someone on Twitter replied saying, quite reasonably, “yeah, big deal, you weren’t going to travel there anyway.”

He was right of course. It was an empty declaration. It had no real consequences.

At the beginning of this year, I decided that for the foreseeable future one way for me to protest the villainous actions and attitudes of the President of the United States was to refrain from any travel to the US. This includes academic travel as well as tourism.

I have not submitted papers to US conferences, I have not attended any US conferences, I’ve resigned from some associations that refuse to consider relocating their meetings, and I recently turned down an invitation to be a scholar-in-residence at a US liberal arts college. As an academic librarian and a PhD student, this decision has already impacted my ability to engage with colleagues and to advance my research work.

I know my decision can easily be viewed as pointless. I know that what I have given up is nothing compared to those in the US being deluded and oppressed by their own government. But it is something. And I felt I had to do something tangible and with consequences for me, however small and probably ineffective.

…Mike

Share
Posted in AboutMike | 2 Comments

Books and Screens and the Reading Brain

Last week I participated in the Books and Screens and the Reading Brain conference held at Vilnius University in Lithuania. The lead organizer was the EU COST action E-READ network and it drew attendees from across Europe. I was the only person from North America. The E-READ network is a group of researchers exploring the implications of digital reading.Vilnius is a fascinating city and I learned a lot about Lithuania (especially its troubled history and its very proud citizens). Vilnius University (founded in 1579) was a wonderful host. It was refreshing to be outside the US/Canada bubble that dominates most of my professional life.

Paper and Digital

Unsurprisingly, many of the papers discussed the digital reading experience: affect, comprehension, representation, intimacy, and even “aura”. The methodology of choice was eye-tracking. While a few researchers reported on the use of EEG and fMRI, eye-tracking was used by many to explore how we read digitally, especially in comparison to reading on paper.

And this leads to the most common theme: paper vs. digital. And the “versus” part should be emphasized. This was clearly competition. There were many good papers. One especially valuable meta-analysis (presented by Ladisiao Salmerón; a member of Rakefet Ackerman’s research team) analyzed dozens of empirical studies of paper and digital reading. According to the results, there is a clear “digital deficit” with respect to reading comprehension.

I confess to being tired of this debate, largely because it is an “apples and oranges” comparison.

Straight up comparisons of paper and digital reading typically standardize the text being read to isolate the effect of the media. However, few studies looked at e-readers (e.g. Kindles), so most compared paper reading to less optimal digital reading environments (i.e. computers or tablets). Writing on paper evolved to optimize the reading experience. It wasn’t always so (think of early manuscripts with dense typography and no punctuation or paragraphs). These studies change the format but they don’t consider changing the nature of the writing.

Put simply, effective digital reading requires appropriate/effective digital writing. Just as we write differently for paper so should we (in fact, do we) write differently for digital.

A more interesting empirical study would be to compare idea comprehension between a text written for and read from paper and a text written for and read from an e-reader (or an otherwise appropriate device). Still “apples and oranges”, but this time not a search for “deficit” but a search for effectiveness.

…Mike

Share
Posted in Digital Culture, Literacy | Tagged , | 1 Comment

Social Media, AI, and Those Pesky Humans

Sometimes it just falls into your lap….

As some will know, I’m fascinated by artificial intelligence (AI), especially how it relates to literacy and our notion of being literate.

So, when I read about Pi from Post Intelligence I was intrigued. Pi is an “intelligent social media assistant”. Pi will recommend tweets for you, review your tweets for effectiveness, select the best time for you to tweet, and (probably soon) tweet for you.

Pi

Pi from Post Intelligence

But here’s the comment from Post Intelligence founder Bindu Reddy, explaining why we need Pi, that leaped out at me:

“Humans find it very difficult to be good at social media.”

I know Twitter is already bot infested but this takes it to a new level. Give up, let the AIs do all that pesky reading, writing, and posting.

The idea that social media and humans aren’t a good fit both delights and appalls me.

…Mike

Share
Posted in Digital Culture, Literacy | 3 Comments

Neuromancer

It seems this blog and my exploration of SF have gone off the rails a bit. Trying harder in 2017. Back to the SF project…..

Neuromancer (William Gibson, 1984) begins with arguably the most famous opening sentence in SF:

The sky above the port was the color of television, tuned to a dead channel.

It is a sensational read but what does it tell us about post-literacy? Surprisingly less than I expected (and remembered) but only perhaps because so much of it is now commonplace.

RobotThis post is part of an ongoing series about science fiction and post-literacy. What can speculative fiction tell us about post-literacy? Spoiler Alert in effect.

While I’ve berated SF for poor writing, Neuromancer is not among them. This is a wonderfully realized novel with penetrating insights, subtle control of language, and startling images. By combining different genres (Raymond Chandler, Louis L’Amour, and William Burroughs echo throughout the novel) Gibson uses the old and familiar to highlight the new and strange. We appear to be in the known world of hard boiled PIs, outlaw cowboys, and 60s drug culture; but then again, we’re not.

The matrix, which now seems so commonplace, is described in ways that continue to capture the magic and complexity of cyberspace (a term Gibson made popular but did not originate). Gibson calls the matrix a “consensual hallucination” (p. 5) and the “infinite neuroelectronic void” (p. 115).

The battling AIs, Wintermute and Neuromancer, are the main attraction; it’s the type of battle that Elon Musk and the folks at OpenAI are currently worried about.

A few “post-literate” notes of interest:

People who used neuro implanted sockets for “microsofts” (knowledge chips) are described as “gone silicon” (p. 73).

The idea of the “simstim” is interesting but disappointingly one way. It allows Case to track and experience the sense data of Molly but she can’t respond. A bit voyeuristic.

McCoy Pauly (“The Flatline” or Dixie) is interesting because he is dead and exists now as a ROM “construct”: a static intelligence that can interact and accomplish knowledge based tasks but can’t learn anything new. It (he) does seem to have self-awareness, however, and longs to be erased (i.e. really, really dead).

As captured, codified knowledge and experience, a construct is a “live” or “living” resource (but not an AI). Presumably we could all leave/create constructs of ourselves to advise, influence, and even assist those in the present (as does the Flatline when he works with Case on the virus they unleash).

It’s a wonderful novel and it’s depiction of cyberspace is nothing short of poetry. And with Wintermute, Neuromacer, microsofts, simstims, and constructs we have some interesting insights into post-literacy.

…Mike

William Gibson. Neuromacer (New York: Ace Book, 1984).

Share
Posted in Science Fiction and Post-Literacy | Leave a comment

Being Editor-in-Chief

For the past two years I’ve been the Editor-in-Chief of Open Shelf, the e-magazine of the Ontario Library Association (OLA). I’ve decided it’s time to step down.

Previously I was the Editor of Access, OLA’s print quarterly. However, starting and nurturing an online magazine which published every two weeks was a whole new adventure. And it has been a blast.

Open Shelf

What Has Been Learned?

Publishing is an act of passion. If you weren’t a bit obsessed with doing this you wouldn’t put in the necessary time and effort. As they say, it isn’t work if you love what you are doing.

As Editor I’m always worried about the “food chain.” Will enough articles be submitted to keep our twice a month schedule going? And they were. We are a community with many writers … excellent writers … with wonderful stories to tell.

We are also a community that needs a bit of prodding. So, ask, always ask. Many of the articles in Open Shelf were a result of simply asking people to write about their experiences or ideas. People love to be asked and we loved publishing the results.

Fun … have fun with the magazine. It is certainly fun working with the writers and the Editorial Board. Certainly publishing annual April Fool’s Day issues (2015 and 2016) was fun (if occasionally controversial). Put this in the category of not taking ourselves too seriously.

Our “formula” is very simple: short articles written in a conversational tone for a generalist library audience. I think this formula has been the strength of Open Shelf; I hope it continues.

A (Small) Regret

Initially I was obsessed with “readability” – a UX for Open Shelf that made the articles attractive and easy to read on any device. I’d give us a B+ on that (some of this is a limitation of the WordPress theme; some of it is shortcomings on our vision of how to do this). I also thought about all the various digital extensions that were possible online and not available in a print magazine. We’ve experimented with video, audio, long form, mixed media, and others. But so far, only experimented (so a C- on that one). If I have a regret about my time as Editor it is that I didn’t push the envelope further on things beyond text and pictures. The new column by the wonderful Amanda Etches (Incidentally) is a good example of where we could go with creative new ideas.

The Next Editor

I’m excited for whomever takes over the Editor-in-Chief role (OLA has issued a search). It is an extraordinary opportunity to work with many wonderful people and to learn a tremendous amount about libraries and the communities around them.

Sure, I  had to wrestle with WordPress, struggle with the peculiarities of our theme, admit my obvious limitations as a proof reader (I failed Grade 4 spelling … and it shows), confront terrible hosting services (now gone), and learn how to publish issues from hotel rooms with terrible WiFi. But it has been a blast.

…Mike

Share
Posted in Uncategorized | Leave a comment

The Pedagogy of Audio: Making Radio with Undergraduates

Last Friday I presented a workshop at the annual conference of the Society for Teaching and Learning in Higher Education (STLHE) on how I use radio (broadcast and podcast) in my undergraduate teaching.

stlhe2016

It was a blast; largely because I took the advice of Mano Singham (with thanks to Leslie Ortquist-Ahrens for the reference).

So, the morning of the workshop I posted this on Twitter:

Singham’s cri de coeur (Death to the Syllabus!) had an impact on me. If I’m not watchful in my presentations and my teaching, I will fill up the time with my own ideas and my own voice (which may or may not be code for “I lecture too much”).

Viewing my workshop participants as co-creators opened the door for a much more interactive and, I think, satisfying session. “Trust the group” is something I’ve heard over and over from my mentors (thinking of you Kathryn Deiss). And, of course, they came through with questions, suggestions, observations, debates …. all those things that make for an effective learning experience.

…Mike

BTW here’s a link to the PPT for the workshop …. the vast majority of which I never actually used.

Share
Posted in Uncategorized | Leave a comment

SF Writing

I think Buzz Lightyear captured the issue:

Buzz_LightyearBeyond infinity; that’s the issue. A funny, sarcastic line that is suggestive that some, maybe a lot, of science fiction writing ….. sucks. I’ve embarked on this little project to read SF looking for examples of “post-literacy” (examples of literacy practices that extend current practices or are wholly new; it’s background work for the PhD I’m undertaking).

I started with a (“the”) classic, Shelley’s Frankenstein. It was brilliant in every dimension; story, concepts, writing. Then I moved on to some others including Arthur C. Clarke (Childhood’s End),  Walter Miller Jr. (A Canticle for Leibowitz), and Robert Heinlein (A Stranger in a Strange Land). Yikes. Some interesting ideas but the writing is poor (if not horrible in some cases).

I know I’ve still got lots to read and will undoubtedly find some good writers along the way. I’ve just started to re-read Neuromancer (William Gibson, 1984) and it is has startling and wonderfully written as I remember it many years ago. Review to follow soon.

Hope springs eternal.

…Mike

Share
Posted in Science Fiction and Post-Literacy | Leave a comment

Stranger in a Strange Land

On my copy of Robert A. Heinlein’s Stranger in a Strange Land (Ace Book, 1987) the cover boldly announces the book as “The Most Famous Science Fiction Novel Ever Written.” OK, but what does it have to tell us about post-literacy?

RobotThis post is part of an ongoing series about science fiction and post-literacy. What can speculative fiction tell us about post-literacy? Spoiler Alert in effect.

stranger_in_a_strange_land-web

About a human, Valentine Michael Smith, born and raised on Mars as a Martian and now brought to Earth possessing exceptional abilities and a rock star status, Stranger starts well and then drifts off into a thin gruel of libertarianism, spirituality, and social commentary. The characters are cartoons and the writing about women and about sexuality is especially juvenile. Having said that, if you really want to read a real turkey from Heinlein try Starship Troopers. Clearly I’m no fan.

But Stranger in a Strange Land, for all its faults, created one of the most powerful and enduring ideas relevant to my exploration of post-literacy: grokking. The intense connection and deep understanding enabled through grokking is an archetype for what a more profound human interaction could be:

“Grok means to understand so thoroughly that the observer becomes a part of the observed—to merge, blend, intermarry, lose identity in group experience. It means almost everything that we mean by religion, philosophy, and science—and it means as little to us (because of our Earthling assumptions) as color means to a blind man.” Robert A. Heinlein, Stranger in a Strange Land (1961).

Part telepathy, part Vulcan mind-meld, part immersion, part collective conscious, part epiphany; this is powerful stuff indeed.

A very important aspect of Smith’s grokking is that he learned to do it. Smith is a human not a native Martian; this ability was acquired, presumably through education, training, and example. Other characters in the novel learn to grok, although with lesser success. Clearly, we too can learn to grok; this capability and capacity is within human reach.

While grokking and the word “grok” are used repeatedly throughout Stranger (and by repeatedly I mean ad nauseam) it’s surprising that the affect of grokking is poorly articulated. We get the idea of grokking and its intellectual implications, but the emotional, holistic impact is missing.

For a much more convincing rendering of grokking I recommend “Girl in Wave : Wave in Girl”, a short story by Kathleen Ann Goonan (which I wrote about here).

The other literacy practice of interest in Stranger is the nature and affect of the Martian language. It is difficult to master and is described as completely antithetical to English. However it seems to inculcate a telepathic capability in the user. The language brings with it other capabilities (or releases other innate abilities).This isn’t explored much, and it seems light weight and even unnecessary given the more powerful ability to grok. Why use language if you can grok?

Because “grok” has entered mainstream usage (in particular in the coding community), it has lost some of its profound implications. It isn’t just a matter of knowing something; it is a deep and holistic understanding that touches the core of one’s being.

I just wish Heinlein had imagined a more appealing word to describe it. Grok. Yuck.

…Mike

Robert A. Heinlein. Stranger in a Strange Land (New York: Ace Book, 1987). First published in 1961.

Share
Posted in Science Fiction and Post-Literacy | 1 Comment

Solaris

Stanislaw Lem’s Solaris was originally published in 1961 in Polish. Apparently the English translations have been poor until Bill Johnston issued his e-book only version in 2013.

Solaris is part of the “alien contact” theme that is prominent in science fiction. In terms of my interest, the novel explores what happens when intelligences (human and alien) try to understand each other. What literacy practice will they use?

RobotThis post is part of an ongoing series about science fiction and post-literacy. What can speculative fiction tell us about post-literacy? Spoiler Alert in effect.

Solaris-web

Alien encounters are really only of interest to me when they reveal something about a post-literate future. And Solaris does this in a very interesting and compelling way.

The alien in question is a vast ocean on a remote planet. A sentient “plasmic machine” that communicates (thinks) in physical manifestations of itself. The attempts throughout the book for each to understand the other (alien to human; human to alien) are complete failures.

What if we meet an advanced society, but we have nothing to talk about. Or at least no way to talk about it.

One of the possibilities for post-literacy (and I know it’s weird, stay with me), is that it will be brought to us by advanced societies (i.e. aliens). Solaris provides us with that encounter but it leaves us with the profound conclusion that neither party will understand the other. To the frustration of both, communications will fail, misunderstandings will occur, and the opportunity to connect will be squandered.

Unlike the standard sci-fi trope that humans can’t understand the aliens, it is quite clear in Solaris that the alien doesn’t understand humans. The ocean makes all sorts of stumbles in its attempt to make intelligent contact. And this is especially surprising since it can probe our minds.

The “language” of the ocean is comprised of many physical transformations (“mimoids”, “G-formations”, “extensors”, and others) forming an extensive physical vocabulary and grammar. A language that represent ideas, statements …. something (because frankly we don’t know what it really means). The ocean is “a geometric symphony, but if this is the case, we are its unhearing audience” (location 1964).

Making ideas physical (corporal/embodied) is most dramatic when the narrator’s wife, who committed suicide some years earlier, appears on the station and interactions (somewhat successful) with the narrator (Kelvin; interestingly, an allusion to unit of measure with a connection to water). The dead wife has been “incorporated” by the ocean by probing Kelvin’s mind. Things do not go well.

Both parties make attempts normal to their own intelligence and literacy practices; both parties fail miserably and, in some cases, tragically. As Kelvin says “Contact means an exchange of experiences, concepts, or at least results, conditions. But what if there’s nothing to exchange?” (location 2414)

Advanced intelligence is not a linear capacity; we don’t simply evolve along some continuum. Intelligence is constructed within a specific context; a context that may simply be unavailable to the other.

If some of us acquired a post-literate capacity or capability (i.e. sufficient to displace or replace alphabetic literacy) it may be that those with this ability would be incomprehensible to those without it.

…Mike

Stanislaw Lem. Solaris. Translated by Bill Johnston. Krakow: 2013. Originally published in 1961.

Share
Posted in Science Fiction and Post-Literacy | 1 Comment