There is continued talk about how robots and other forms of automation, such as artificial intelligence, may be going to take most of our jobs. Some more informed observers, however, are sceptical, pointing out the enduring limits of AI. I want to add to this scepticism by talking about an employment sector where automation and IT have already drastically changed the content of jobs, but without eliminating the profession: librarianship.
I went to library school over thirty years ago (I studied at CLW in 1987-1988) , at a point when the cutting edge storage technology was the CD-ROM and to access remote databases, you used a dial-up modem, running at 300 bps (as opposed to modern broadband rates in the UK in the tens of megabytes per second). As for the internet, I was very proud of myself when I worked out how to use JANET (the Joint Academic Network) to search a library catalogue elsewhere in the UK, but it was far too complicated a procedure to think of doing regularly.
I started as a librarian, therefore, about ten years BG (before Google) and indeed even before the World Wide Web was invented. By the time I’d been a librarian for around fifteen years, the changes were drastic. In 2001, my job was managing the University of Hertfordshire’s electronic journals, a form of communication that was still only in the experimental stage when I was at library school. And several other traditional library roles and tasks had already pretty much vanished.
The most drastic was the widespread disappearance of the “reference librarian”, the specialist in finding information for users from print sources. In library school I remember reading several books of case studies by Denis Grogan of how one answered a variety of reference queries. That specialism was already being threatened by online databases thirty years ago: it’s now effectively obsolete, or at most only a small fraction of library work.
Online searching in turn has largely ceased to be an activity for library professionals. Twenty years ago, when you paid for online databases based partly on the length of time searching them and used a command line interface, online searching was almost entirely done by professional library staff, and it could be an important part of a librarian’s role in academic and specialist libraries. Nowadays, with the partial exception of some specialist medical, business and legal databases, most database searching is done by the end users themselves.
By 2001, I could also see the rise of the paraprofessional: members of library staff who did not hold a specialist library qualification, but who were carrying out tasks that had once been reserved for professional librarians. At the University of Hertfordshire, there were only one or two specialist cataloguers, rather than a whole team: most of the acquisitions came “shelf-ready” from the supplier, supplied already with a catalogue record and with only minimal changes being made by non-professional staff. The library service had also “converged” with the IT department, with frontline staff trained to deal with straightforward library and IT enquiries, and qualified librarians reserved for higher-level and more specialist roles, such as academic liaison.
Given all this change between c. 1987-2001, what’s surprising is that technological changes (at least in academic libraries) have had far fewer disruptive effects on library roles and jobs in the period 2001-2020. The biggest change has been self-issue, which has affected library assistants’ jobs quite a bit, although I’m not sure the extent to which there have been job losses as a result. But otherwise, few roles have been eliminated through technological advances, as opposed to universities’ continuing efforts to cut staffing costs. Library services have been reconfigured in many ways, but the main roles are still familiar.
What changes there have been in academic libraries since 2000 are in new roles developing. There are a few new posts in designing library apps and data analysis. Advances in reading list software have created a whole new aspect of library work. (I was initially employed by the University of Hertfordshire in 2000 on a project to see if it was possible to implement online reading lists with their existing library management software. Spoiler: it wasn’t, instead it needed the specialist software that has been developed since).
One of the most significant developments in terms of new jobs in the UK is managing open access, repositories, scholarly communication and research data. These roles developed primarily because of policy changes by research funding councils, rather than any technical developments, although new systems are now appearing to help manage the research process. Here, there’s a blurring between traditional library work and more general information and knowledge management roles.
I know the other sectors of library work much less well than academic libraries, but apart from self-issue, I haven’t seen any recent technical changes in public libraries that would result in job losses. The widespread job losses that have occurred are overwhelmingly because of funding cuts, rather than library staff becoming obsolescent. In contrast, libraries in the commercial sector did take a big hit with the coming of Google, which supposedly made them unnecessary. There are roles in knowledge management in some companies, but they’re not necessarily held by people from a library background.
If you put all these developments together, what you can see is a profession that has been significantly changed by developments in IT, but not one that has been eliminated. And it’s difficult to see technological changes coming that would drastically affect employment. There are now book sorters available, but I’d be surprised if we saw robot shelvers any time soon: physical books and shelving are non-standardised enough to make robots less effective than humans. And the physical book is still here for the long term, even if electronic journals have largely won out over hardcopy in academic libraries. One of the early services that many academic and public libraries are implementing during the Covid-19 epidemic are kerbside pick-ups of print books.
In most areas of library work, we’ve now got a substantial number of IT systems in place, which are increasingly integrated, even if they don’t always work as well as you’d hope. But you still normally need someone there acting as a minder or advisor for when there are problems with the self-issue or the online room-booking system, for example. Could customer service chatbots replace frontline staff? Again, only to a limited extent, until you can create a robot that can both deal with a scanner problem and understand natural language reliably.
A recent webinar I attended on AI in libraries was mainly focused on the possibility of improving discovery of scientific articles via automatic summarising and indexing. That might mean job losses within abstracting and indexing services, but it’s still going to need research support librarians training academics in how to use these fancy new tools effectively, Nor will AI or other technologies easily replace human trainers in academic libraries. Most library instruction isn’t purely generic, but needs to be tailored to particular subjects and particular libraries to be effective: “University X has access to these databases and if you want to find books and articles on topic Y, this is what you need to do.” There isn’t really much opportunity for MOOCs here, and although librarians can create and reuse online and VLE content, creating good quality distance learning materials takes time and expertise.
In my own current job (Digital Resource Manager) I spend a chunk of most days solving problems about access to particular online content. This needs a combination of general IT and library skills (knowing how remote authentication works and how to interpret error messages) and specific knowledge of our own systems and entitlements. Even when I’ve being working from home for five months, the idea that I could easily be replaced by someone cheaper in India is ridiculous.
I don’t think there are many librarians among us who are longing to go back to large-scale local cataloguing or paper reading lists, even if we may still sometimes feel nostalgic for them when the systems crash. Our concerns instead are funding cuts and the ever-increasing costs of e-resources, or the type of information provider branding that obscures the library’s role in paying for and administering resources. But so far we’ve adapted well to IT and now AI, transforming our services rather than having them eliminated. And I think we’ll keep on doing so long after my library career has ended.