Last week Library 2.0 hosted a webinar on "invisible labor" in libraries (very well presented by Sonya Schryer Norris) — the work that librarians do that isn't in their job descriptions, isn't reflected in their compensation, and largely isn't acknowledged by the institutions they serve and that fund them. It was a thoughtful and somewhat sobering presentation, and it stayed with me afterward.
It made me want to apply a framework I've been developing — one I call idealized narratives and actual functions — to the history of public libraries over the last thirty years, and see if it illuminates anything useful about where the profession has been and where it might be heading.
I want to be clear about what this is and isn't. It isn't a critique of librarians, who are by and large among the most dedicated and genuinely caring professionals I've encountered in twenty-plus years of working with them through Library 2.0. It's an attempt to map something honestly, in the hope that the map is more useful than a more comfortable picture would be.
Every institution carries two layers that are worth distinguishing.
The first is the idealized narrative — the story the institution tells about why it exists and what it does for people. Schools educate children. Hospitals heal the sick. Courts deliver justice. These narratives are not lies exactly. They describe something real and something genuinely valued. They attract people into the work and sustain their commitment to it.
The second is the actual functions — what the institution actually does that keeps it alive, what its participants genuinely depend on it for, why it persists even when the idealized narrative is being challenged. Schools provide childcare, credentialing, and social sorting. Hospitals are organized around billing codes and liability management. Courts process plea bargains. The actual functions aren't cynical substitutes for the idealized narrative. They're the real work the institution performs, and they serve genuine human needs — even when they're rarely named as such.
The gap between these two layers is not corruption. It's the basic architecture of how institutions function. And it becomes analytically useful when a technology arrives that challenges one or both layers.
There are essentially four possible scenarios when a technology disrupts an industry. The technology can challenge the idealized narrative but leave the actual functions intact — in which case the institution absorbs the technology, narrates it as innovation, and continues. It can challenge both the idealized narrative and the actual functions — in which case the institution faces genuine existential pressure. It can leave the idealized narrative untouched while quietly undermining the actual functions — a kind of silent disruption where the story still sounds credible while the floor drops out. Or it can challenge neither, leaving the institution essentially unchanged.
Which scenario applies determines almost everything about what happens next. And the library's story over the last thirty years is, I think, a revealing case study in how this plays out in practice.
It's also worth saying at the outset that this pattern is not unique to libraries. Skilled, respected professions have faced this dynamic before, and the experience of navigating it is part of a longer human story about what happens when technology moves the ground beneath genuine expertise. I've included an appendix at the end of this essay with several historical examples, for readers who want the broader context.
The idealized narrative of the public library is one of the more beautiful ones in American civic life. Free, equitable access to information and knowledge for all citizens. A democratic institution that says: regardless of your income or your background, the accumulated knowledge of civilization is available to you, staffed by professionals trained to help you navigate it, open to everyone.
That narrative drew people into the profession for generations. It still does.
Beginning in the mid-1990s, the internet began dismantling the primary actual function that had sustained the public library: the management of physical information scarcity. The library had the books, the databases, the periodicals, the professional expertise to navigate them. You came to the library because that's where the information was.
The internet didn't make libraries disappear. But it removed the information scarcity function that had been the institution's primary justification. And what happened next is worth examining carefully.
Libraries didn't contract. They expanded — into computer access, into meeting spaces, into community programming, into social services navigation. Each expansion was narrated as a natural extension of the mission. And each one was genuinely responsive to community need.
But it's worth asking an honest question about some of those expansions. Makerspaces, for instance, became a significant trend in school and public libraries through the 2010s. They're creative, they're engaging, they provide hands-on learning opportunities. Many librarians embraced them with genuine enthusiasm, and that enthusiasm was real.
At the same time, makerspaces are more naturally a vocational education function than an information access function. The honest observation isn't that makerspaces were wrong. It's that the library was the available institution — present, funded, staffed by people constitutionally inclined to say yes — rather than necessarily the right institution for that particular function. The pattern of reaching for adjacent activities when the core function erodes is worth noticing, because it has continued, at greater cost, in the years since.
Here is the most direct way I can describe what I observe in the library community today, drawing on some years now of organizing professional development for librarians and paying close attention to what they tell us they need.
Our most popular programming at Library 2.0 (after AI right now) is not about information literacy. It's not about collection development or community engagement. Our most in-demand offering has been "Dealing with Difficult Patrons," part of an extensive body of work around library safety and security developed by Dr. Steve Albrecht. We have also added a mental health and wellness programming track with Loida Garcia-Febo because the demand was there.
We didn't create these offerings speculatively. We followed what the audience was asking for. And what the audience is asking for tells you something that the idealized narrative doesn't.
Librarians are increasingly reporting that they don't feel safe. They are managing mental health crises, addiction, homelessness, behavioral situations, and even sometimes violence, for which their training did not prepare them and for which their institutions have a hard time providing adequate support. The invisible labor conversation from last week is really a conversation about this: the actual functions of the public library have expanded well beyond the idealized narrative, without corresponding expansions in funding, staffing, training, or compensation.
The funding gap is worth naming specifically, because it sits at the center of the difficulty. Libraries are not perceived by funding bodies as the appropriate institutional home for social services. Mental health systems, social services agencies, and public health departments are. But those systems are overwhelmed, and the library — open to everyone, staffed by caring professionals, physically present in communities — often quietly absorbs the overflow. The functions arrived without the funding that should accompany them, and there is no obvious political mechanism by which that funding will follow.
There is also something worth naming about the professional culture itself, and I say this with genuine appreciation rather than judgment. Librarians are caring people. A significant portion of the profession holds values oriented toward service, inclusion, and advocacy for the vulnerable. Those values are real and they produce real good. But they also make institutional boundary-setting genuinely difficult. The same orientation that makes librarians exceptional at their work makes it hard to say: this function belongs elsewhere, and absorbing it without resources is not serving either library patrons, libraries, or library staff. Within the professional culture, that kind of boundary-setting can feel like abandoning the mission. The result is a profession that keeps saying yes to functions it cannot adequately resource, sustained by an idealized narrative that doesn't fully account for what the work has become.
The consequences are visible. Compassion fatigue is a recurring theme in library conversations. Attrition is increasing. The gap between what people imagined the profession would be and what it looks like on any ordinary day has widened considerably.
The internet challenged the library's idealized narrative while leaving some significant actual functions intact. Reference expertise, information navigation, the professional mediation of complex research questions — these survived the internet era as genuine differentiators. A trained librarian navigating specialized databases, evaluating sources, and guiding a patron through a difficult research question was doing something Google couldn't fully replicate.
AI challenges those remaining functions directly.
The reference interview — understanding what a patron actually needs, translating a vague question into a productive search, evaluating the quality and relevance of results — is very close to what a well-used language model does. Not perfectly, and not without limitations, but well enough for the majority of reference interactions, which are not complex research questions but relatively routine navigation tasks. And AI is only going to get better. Maybe dramatically.
It's also worth being honest about something the profession's idealized narrative has tended to overstate. Librarians have championed information literacy as a professional value and a public mission for decades. That commitment is genuine. But the reach of that expertise beyond the profession's own community has been limited for some time. The audience for information literacy training has been largely other librarians, library school faculty, and professional conference attendees. The general public has not become demonstrably more information literate as a result of library programming. This isn't a failure of effort or intention — it's a structural observation about where the expertise has actually landed.
The deeper issue is that the cultural conditions which would reward careful, critical thinking about information are not currently being cultivated by the broader environment. The attention economy is optimized for fast, confident, frictionless information consumption. AI adoption is being driven by the same logic. The librarian's instinct to slow down and evaluate — which is genuinely valuable and genuinely rare — is structurally at odds with how information is being produced, distributed, and monetized at scale. The expertise doesn't lose because it's wrong. It loses because it's slow in a fast system, careful in a careless one, and the business model of the attention economy depends on people not doing it. There is a tragic irony here — the moment when careful evaluation of information arguably matters most is arriving precisely as the institutional structures that housed that expertise are under their greatest pressure.
This places the library firmly in the second scenario of the framework: the technology challenges both the idealized narrative and the actual functions, and leaves the institution without a strong layer underneath to hold it up. What remains — the social services absorption, the safety management, the community anchor function — arrived by default rather than by design, without funding or cultural mandate, and without the kind of public recognition that would sustain political support for library budgets.
That is a genuinely difficult position. And it's one that, in my observation, the profession has not yet found a clear way to talk about.
None of this is a verdict on individual librarians, who are navigating real choices in real institutions with real constraints. The institutional trajectory and the personal trajectory are different things, and it's worth separating them.
Looking at the library community as honestly as I can, I see three paths emerging.
The first path is staying and championing the evolution toward community services. Some librarians are genuinely suited to and fulfilled by the expanded mission — the human contact, the social services navigation, the role of community anchor. They find meaning in the work even as it changes, and they are not wrong to. The community need is real. The question for people on this path is whether they can go into it with clear eyes about the funding reality, the compassion fatigue risk, and the gap between what the institution is being asked to do and what it is being resourced to do. Sustained by an honest assessment rather than an idealized narrative, this path is viable and valuable. Sustained by the fiction that the resources and recognition will eventually arrive, it carries significant personal cost.
The second path is staying and fighting for the critical information function — the expertise in source evaluation, research navigation, and information literacy that the profession was trained to provide. This is a genuine fight, and the expertise being defended is real. But the honest observation is that the institutional battle for this function within the public library is probably not one that will be won, at least not at scale, in the current cultural and funding environment. Where this expertise has a more promising future is in adjacent spaces that are actively looking for it: health systems navigating AI-generated medical information, corporate settings grappling with information quality, educational institutions that are actively trying to prepare students for an AI-mediated information environment, and policy contexts where careful evaluation of AI outputs is genuinely valued. The expertise is worth carrying. The public library may not be the institution that allows it to flourish.
The third path is moving on — leaving the library for settings where the training and values translate into work that feels more aligned with what drew people into the profession in the first place. This is not abandonment. It is a reasonable response to an honest assessment of where the institution is heading and what the individual needs in order to do good work sustainably. The people who make this choice would not, in my view, be failing the profession. They are exercising the self-knowledge that the profession's own idealized narrative sometimes makes difficult to claim.
Institutions are not the same as the people inside them, or the values they carry, or the expertise they've developed.
The public library as an institution is under genuine pressure, and the trajectory I've described here doesn't resolve itself easily. The funding gap between what libraries are being asked to do and what they are being resourced to do is structural, not incidental. The cultural conditions that would elevate and reward careful, critical information expertise are not currently strengthening. The profession is navigating a transformation that arrived without a roadmap and without adequate support.
But the people who went into library work because they believed that careful thinking about information mattered, that equitable access to knowledge was worth defending, that the vulnerable deserved patient and competent help navigating complex systems — those people are carrying something worth carrying. That expertise and those values have a future. It may not be inside the institution as it currently exists. But it exists.
The most useful thing I can offer, after twenty years of watching this community, is simply the honest map. Not because the map is encouraging in every direction, but because people who know what they're looking at can make better choices.
The invisible labor webinar from Sonya was, in a way, an early version of that honest map. The profession is naming what it has actually been doing. I think that has value.
The pattern described in this essay — a skilled, respected profession finding that technology dissolved the specific function that made its expertise valuable — is not new, and it is not unique to libraries. The following examples, researched and written by Claude.ai, are offered not as context. Understanding that this dynamic recurs across very different fields, affecting professionals who were genuinely skilled and genuinely committed, makes it easier to see the structural forces at work rather than looking for failures of individual effort or institutional will.
Typographers and Compositors
For centuries, the craft of typesetting — arranging type for print — required years of apprenticeship, genuine expertise in typography, layout, spacing, and print quality. Compositors knew things about how text worked on a page that writers and editors did not, and that expertise was respected and well-compensated.
Desktop publishing, arriving in the mid-1980s with the Macintosh and programs like PageMaker, moved the typesetting function to the person generating the content. It didn't augment compositors — it made their specific function unnecessary by relocating it. The expertise was real. The cultural moment when it might have been most valued — the explosion of desktop-published material that was typographically poor precisely because the craft knowledge had been discarded — arrived just as the profession was being dismantled. What survived was a smaller, higher-specialized tier of typographic designers working at a level where the craft still commanded a premium.
The parallel to libraries is close: caring professionals with genuine craft expertise, a technology that moved the function rather than improving it, and a cultural moment where the expertise arguably mattered more than ever arriving at the same time as the institutional disruption.
Accountants and Bookkeepers
Before the spreadsheet, a significant layer of the accounting profession existed specifically to manage the labor of numerical calculation and ledger maintenance. This wasn't clerical work — it required training, accuracy, professional judgment, and deep familiarity with financial records and their relationships. The expertise was real and the function was genuinely essential.
VisiCalc, released in 1979, and later Lotus 1-2-3, dissolved the calculation and ledger function almost immediately. What remained — tax judgment, financial strategy, audit, complex analysis, regulatory compliance — was genuinely more skilled work. But the middle layer, the professionals whose expertise was primarily in accurate calculation and ledger management, found their specific function gone. The profession restructured around what the spreadsheet couldn't do, and shed a substantial portion of its workforce in the process.
The lesson is not that accounting disappeared. It's that the technology didn't augment the existing profession uniformly — it eliminated specific functions and left others intact, and the people whose expertise was concentrated in the eliminated functions faced a genuine reckoning about where their skills translated.
Telephone Operators
Perhaps the cleanest historical example. Telephone operators were skilled, professionally organized, and genuinely essential — routing calls through complex manual switchboards required training, spatial reasoning, memory, and the ability to manage multiple simultaneous demands under pressure. At their peak in the mid-twentieth century, telephone operating was one of the largest skilled female employment categories in the United States.
Automated switching didn't make operators faster or more productive. It made their specific function unnecessary entirely. The transition happened over decades rather than years, which provided some cushion, but the endpoint was the same: a profession that had been essential found that the technology had moved the function, not improved it.
What makes this example useful is its clarity. There was no ambiguity about the expertise being real. There was no argument that operators were doing their jobs poorly. The function they performed was simply no longer a function that required a human.
Legal Stenographers and Court Reporters
A smaller but precise contemporary example, and one that is still unfolding. Court reporters are skilled, certified professionals whose expertise in stenographic transcription and the official legal record is both genuine and legally mandated. The idealized narrative of the certified official record remains intact and culturally credible.
Voice recognition and AI transcription are dissolving the actual function — accurate real-time transcription of spoken proceedings — while the legal and professional frameworks that protect the certified record are slowing the transition. The profession is fighting on accuracy and certification grounds, which is exactly the institutional resistance the framework would predict. The outcome is not yet settled, but the direction of travel is clear, and it follows the same pattern: genuine expertise, a technology that performs the core function adequately enough for most contexts, and a profession defending its idealized narrative while the actual function erodes.
Radiologists
The most instructive contemporary parallel for a professional audience, because it involves a high-status, highly credentialed profession in the middle of this transition right now.
The idealized narrative of radiology — expert medical interpretation, years of specialized training, the irreplaceable judgment of an experienced physician reading complex imaging — remains fully intact and culturally credible. Radiologists are respected, well-compensated, and central to medical practice.
AI diagnostic tools are already matching or exceeding radiologist accuracy on specific imaging tasks, particularly in pattern recognition for conditions like certain cancers, diabetic retinopathy, and pneumonia detection. The actual function — the expert interpretation of medical images — is being quietly undermined while the idealized narrative remains strong and the legal and institutional frameworks still protect the role.
The profession is responding in the ways the framework predicts: emphasizing the complexity of judgment that AI cannot replicate, the importance of the physician relationship, the cases where AI fails. These arguments are not wrong — they describe real aspects of the work. But they concentrate on the top of the expertise distribution while the middle is being compressed. The radiologists whose work involves the most routine pattern recognition face a different future than those working at the frontier of complex diagnostic judgment.
Radiology is an early and high-visibility example of what may become a broader pattern in professional services: the idealized narrative of expert judgment holds, the legal and institutional protections hold, but the actual function is being performed increasingly by AI, and the profession is restructuring around what remains.
Travel Agents
A more familiar example, and one that has already completed its arc. The idealized narrative of the professional travel agent — expert knowledge of destinations, relationships with hotels and airlines, the ability to construct complex itineraries that a layperson couldn't navigate — was credible and the expertise was real.
The internet didn't improve travel agents. It bypassed them by giving consumers direct access to the booking systems and fare information that had previously required a professional intermediary. The actual function — access to information and booking infrastructure that the public couldn't reach directly — dissolved almost completely. What remained was a smaller tier of high-end travel specialists serving clients for whom the complexity of the itinerary or the value of the relationship justified the fee.
The travel agent example is instructive because it completed quickly and completely enough that the outcome is clear. The profession didn't disappear — it contracted to a specialized tier and restructured around the functions that direct consumer access couldn't replace. The people who survived were those whose expertise was concentrated in judgment and relationship rather than information access.
These examples share a common structure. In each case, the expertise was real. In each case, the technology didn't simply augment the professional — it moved or eliminated a specific function that the professional had been trained to perform. In each case, the profession responded by emphasizing the idealized narrative of expertise and judgment while the actual function eroded beneath it. And in each case, the outcome for individual professionals depended heavily on whether their specific expertise was concentrated in the function the technology eliminated or in the functions that remained.
That structure is what makes these examples relevant to the library situation. Not as a prediction of exactly how things will unfold, but as a reminder that this dynamic has a shape — and that understanding the shape is more useful than being surprised by it.
