Marginalia 23 – Technically wrong

I’ve missed a few weeks of Marginalia – sorry about that.

Regular correspondents will be familiar with my views on the invisible work of systems maintenance in libraries. Sorry but it can never be said too much. Here’s Danielle saying it again:

In my 10 years in libraries there’s been a definite move toward vendor supplied solutions for most library systems. And while there’s nothing inherently wrong with that, it’s given libraries the false belief that systems just run themselves. It’s hidden the complexities of technology and the work of keeping them running.

Under the circumstances, it’s no wonder libraries are addicted to the bright and shiny things that get them good publicity and make them popular with funders. But what is often forgotten is that for libraries to do the showy things they need good systems and librarians who can make that happen.

She wrong of course about there being “nothing inherently wrong with that”, but everything else is right 😆

MPOW has just announced a long-anticipated restructure that will ultimately end with 200 people seeking new opportunities. The driver is a sea of red ink in the financial books triggered by Australia’s international student market suddenly closing at the beginning of the COVID pandemic, but there’s been plenty of rhetoric about automation and efficiency. When the filing cabinet was invented, it too was going to make everyone more efficient:

Now, in the early 21st century, the file cabinet is associated with inefficiency. No longer an exemplar of productivity, rationalization, and speed, it instead represents our collective failure to save time and optimize labor. A file cabinet’s storage capacity now embodies the facility of bureaucracies — impersonal and procedure-oriented — to produce paper, to delay, to leave us waiting.

Hmm, maybe not. Craig Robertson’s The Filing Cabinet, from Places journal, does however provide a fascinating history of this ubiquitous office technology. A seemingly simple piece of furniture (though never described as such), the filing cabinet popularised a dangerous idea that we in the “information sciences” have swallowed greedily:

The filing cabinet contributed to the rise of a popular nontechnical understanding of information as something discrete and specific. Critically, it illustrates the moment in which information gained an identity separate from knowledge, an instrumental identity critical to its accessibility... In this historical period, filing technology provided a conceptual gateway for understanding information as a thing that could be standardized, atomized, and stripped of context — information as a universal and impersonal quantity.

Of course, now physical filing cabinets have become rare. Some large government bureaucracies still actively use them, but most organisations migrated their filing practices to computers – which to a large degree still merely replicate the logic of filing cabinets and the files they housed. Still, computers have made work life easier for everyone, right? Alas:

Initially promoted as a kind of personal assistant, the proliferation of computers meant that many high-level white-collar workers no longer had somebody else to do their clerical chores.

Laine Nooney picks up the story in VICE, with How the Personal Computer Broke the Human Body which at the time of writing is returning a 404, but hopefully VICE will pay their web storage bill soon. With the introduction of PCs into offices

Work became a process of filling in blanks; there was no longer anywhere for the clerks to experience decision-making in their jobs. …As intellectual engagement with the work went down, the necessity of concentration and attention went up. What the computer did was make the work so routine, so boring, so mindless, clerical workers had to physically exert themselves to be able to focus on what they were even doing.

Probably not since the automobile has there been a technology that is so insistently reorganized how we use our bodies in day-to-day practice—and the long arc of these transformations are still being played out.

Don’t forget to get up every 20 minutes and stretch.

With a history like this, it’s hardly surprising that Jackie Brown and Philippe Mesly write, in Real Life, that

when social change is framed primarily in terms of choice of technology, the debate necessarily centers the activity level of a productivist society, not a paradigm shift from growth to wellbeing.

Any technology we adopt should be both appropriate to the world as it exists and to the future we desire.

Brown and Mesly’s article explores what E. F. Schumacher’s ideas on “appropriate technology” tell us nearly five decades after his famous book Small is Beautiful. It’s a complicated story, and as usual not as straightforward as either his supporters or his detractors have expected. A couple of years ago Os Keyes, Josephine Hoy and Margaret Drouhard wrote a fantastic paper on Human-Computer Interaction called Human-Computer Insurrection: Notes on an Anarchist HCI. Their paper is a nice primer for anarchism as much as for HCI, and one of things I most like about it is the plain, jargon-free language it uses.

In the long term, there are other ways of running things. Our concern is not organisation but who gets to define the terms under which things are organised, and how consensual participation in and departure from systems is: with autonomy and decentralisation. Rather than an absence of technologies, we are talking about technologies built in a way that centres the communities using them and avoids reserving for some third party the powers to modify, adapt, and repair; about design processes in which the members of that community are treated not as participants but as accomplices.

So let’s look at a couple of tech projects that have been in the news, and the terms under which things are organised. Last year (I think? It was probably longer than that but pandemic-time is confusing), the machine learning model GPT-3 was all over the news. Ed Summers shared some thoughts about it in February:

GPT-3 is an example of an extraction project that has been underway at large Internet companies for some time. I think the critique of these corporations has often been confined to seeing them in terms of surveillance capitalism rather than in terms of raw resource extraction, or the primitive accumulation of capital. The behavioral indicators of who clicked on what are certainly valuable, but GPT-3 and sister projects like CommonCrawl shows just the accumulation of data with modest amounts of metadata can be extremely valuable.

In the last couple of weeks I’ve seen a lot of comments about Microsoft’s new beta product GitHub Copilot, and the observation that it too is an example of “primitive accumulation of capital” or what normal people call “stealing”. But the smartest take I’ve seen, and one that has rightfully gained a lot of eyeballs, is Julia Reda’s GitHub Copilot is not infringing your copyright. Reda notes that many Open Software advocates are concerned that CoPilot uses software published under the General Public License (GPL) as source material, yet CoPilot itself is a closed-source subscription service, which they claim violates the GPL.

Those who argue that Copilot’s output is a derivative work of the training data may do so because they hope it will place those outputs under the licensing terms of the GPL. But the unpleasant side effect of such an extension of copyright would be that all other AI-generated content would henceforth also be protected by copyright.

If we return to the point our anarchist HCI experts made, it should be obvious that the problem not CoPilot, nor even Microsoft using GPL-licensed works as training data. The problem is that gigantic monopolistic billion-dollar corporations are allowed to exist. “Copyleft” licenses are a clever hack, but I’ve always had a concern about the way many Free Software advocates centre their rights as individuals in discussions around freedom. Defending individual rights against copying by a corporate titan chooses the wrong battlefield, even if it’s the right enemy.

On a light note, there’s a beautiful community project that combines all the the things I’ve been writing about above – tech giants, bureaucracy, volunteer-driven tech projects. Mitch Anzuoni’s True Fans Translate: Fansubbing BookStory explains how he recruited a community of translators, using a Google Sheet, an old version of Visual Basic and a lot of spare time to create an English translation of an old Japanese computer game where you play as a book store manager.

Finally, if you liked Lim, Hellard and Aitken’s classic 2005 paper The case of the disappearing teaspoons: longitudinal cohort study of the displacement of teaspoons in an Australian research institute, you’re going to love Confronting Indifference Toward Truth: Dealing With Workplace Bullshit. Not only is this a paper with a fun name, it actually does provide really useful suggestions for dealing with the problem (as well as demarcating the important distinction between liars and bullshitters).

Happy reading 😀