When generative AI entered our lives just over three years ago, many organizations approached it like any other technology project. Attention quickly turned to identifying use cases, comparing tools, establishing governance, and moving beyond early experimentation.
That work has been—and continues to be—essential. Yet across academic library and information research, another pattern is beginning to emerge. The organizations that are successfully embedding AI are discovering that it is less a technology project than a continuous learning project.
In this article, I explore why implementing technology is only part of the challenge. The harder task is helping people develop the confidence, judgment, and good habits needed to use AI tools effectively—and why that presents a significant opportunity for libraries and knowledge teams.
The Confidence Gap
I was reminded recently why information professionals should be careful about making assumptions.
During a conversation, someone mentioned that they had drafted a rather terse email to a technical support desk. It was exactly the sort of situation where I might ask ChatGPT to make the wording more neutral, clarify the issue I was having, and turn my frustration into a clear, respectful request for help.
Their response was immediate.
“That’s a great idea. I wouldn’t know how to do that.”
It was a moment that made me pause for thought. I realized that I had simply taken my own knowledge and confidence for granted. Those of us who spend much of our working lives reading about AI, experimenting with new tools, and discussing the latest developments can easily assume that everyone else is doing the same.
AI technology has evolved far more quickly than people’s confidence in using it effectively. Many are still taking their first tentative steps, unsure which tool to use and why, when it might be helpful, or how to judge whether the answers they receive are reliable. Or for some, even getting started can be a challenge.
Learning, therefore, cannot be treated as a one-time exercise completed during an implementation project. As technology develops, our understanding must develop alongside it—not simply to keep pace with new features, but to recognize where AI adds value, where human judgment remains essential, and where it should not be relied upon at all.
Every AI Project Reveals That Access Is Not Capability
One of the earliest discoveries organizations make is that providing access to AI tools is only the beginning.
The assumption that people will naturally know how and when to use them rarely holds true. Knowing that ChatGPT, Copilot, NotebookLM, or another tool exists is not the same as understanding which one is appropriate for a particular task, how to ask useful questions, or how to recognize when the answer requires further verification.
Research across library science and knowledge management increasingly describes AI literacy as something much broader than technical proficiency. It includes understanding the strengths and limitations of different tools, recognizing where AI can genuinely improve productivity, and developing the judgment to know when human expertise remains essential. As illustrated above, for many, even identifying a suitable everyday use case is unfamiliar territory.
Perhaps that should not surprise us. When search engines transformed access to information twenty-five years ago, libraries responded by helping people understand how to search effectively, evaluate results, and choose the most appropriate tools for the task. In doing so, libraries demonstrated that embracing new technology and ensuring no one was left behind are complementary goals, not competing ones.
AI is an outreach opportunity for libraries and knowledge teams. For decades, information professionals have helped individuals navigate unfamiliar resources, evaluate information critically, and build confidence in using new technologies. AI literacy represents a natural extension of those existing skills rather than an entirely new responsibility.
Every AI Project Reveals That Learning Is Social
Organizations also discover that successful AI adoption depends less on formal training than on creating opportunities for people to learn from one another.
While introductory courses and online tutorials provide a useful starting point, they rarely build the confidence that comes from experimenting with real work, discussing ideas with colleagues, or seeing how others have approached similar problems.
Across the LIS literature, recurring themes include communities of practice, peer learning, shared experimentation, and encouraging people to exchange prompts, examples, and practical experiences. Learning becomes an ongoing conversation rather than a single event. As confidence grows, so does curiosity, and organizations begin to develop their own collective understanding of where AI genuinely adds value.
Again, this shouldn’t come as a surprise. People still learn in much the same way they always have. We learn by observing others, asking questions, experimenting, making mistakes, reflecting on what worked, and sharing what we discover. Those fundamentals have changed remarkably little, even if the technology itself changes almost weekly.
For knowledge teams, this presents another opportunity. Libraries have long supported organizational learning through teaching, guidance, and trusted advice. The same capabilities are increasingly relevant as organizations learn to work with AI.
Every AI Project Reveals That Human Judgment Becomes More Valuable
The final discovery may be the most important of all. As AI becomes more capable, human judgment becomes more valuable, not less.
Collating information is only one part of knowledge work. Professionals still need to interpret results, understand organizational context, recognize nuance and ethical issues, identify bias, verify evidence, and decide whether AI is appropriate in the first place. These are the qualities that allow people to evaluate AI’s output rather than simply accept it.
Information professionals help individuals, businesses, and other organizations distinguish reliable information from unreliable information, understand authority, and make informed decisions. Those capabilities remain just as relevant in an AI-enabled workplace. If anything, they will become more important as AI-generated content becomes the norm.
Some researchers have asked a more provocative question. If AI performs many of the routine tasks through which professionals traditionally developed expertise, how will future experts acquire the judgement that comes from experience? It is an early debate, but an important reminder that using AI to increase productivity still requires parallel investment in human skills.
A Familiar Role for Information Professionals
None of this suggests that libraries or knowledge teams need to reinvent themselves. Information professionals have guided people through successive waves of technological change, teaching search strategies, source evaluation, digital literacy, and critical thinking long before AI entered the workplace.
What is changing is the context in which those skills are applied. Helping colleagues choose appropriate tools, ask better questions, evaluate machine-generated answers, and understand both the strengths and limitations of AI is becoming one of the most valuable services knowledge teams can provide.
Professional organizations and conferences are already reflecting this shift, placing increasing emphasis on AI literacy, organizational learning, ethical stewardship, and preparing people to face the challenges ahead.
The conversation, then, is no longer simply about adopting AI. It is about helping organizations develop the knowledge, confidence, and judgment to work with it well. That is a familiar challenge for information professionals. The tools may be new, but the underlying mission remains remarkably consistent.
The Access, Capability, and Confidence Spectrum
One unexpected conversation reminded me how easy it is to overestimate where people are on their AI journey. Access to technology is not the same as confidence, and confidence is not the same as capability. Successfully adopting AI depends less on introducing new tools than helping people develop the knowledge, judgment, and habits to use them well.
For libraries and knowledge teams, that realization should feel surprisingly familiar. Helping people navigate new technologies, evaluate information, and build confidence has always been part of our professional role.
Yet learning is only the beginning. As AI becomes embedded in everyday work, organizations will face broader questions about how knowledge is captured, shared, trusted, and applied. Those questions reach beyond technology and into the future of knowledge work itself—and they are the questions explored in the rest of this series.
0 Comments