In my earlier Lucidea series, we explored what generative AI can do (and what it can’t), where it can add value, and how knowledge teams can adopt it safely and responsibly. Those questions remain important, but the conversation is beginning to change.
Across libraries, archives, museums and the wider knowledge profession, researchers are looking beyond AI’s capabilities to its consequences. Rather than asking, “What can AI do?” they are asking, “How is AI changing the assumptions that have shaped knowledge work?”
Libraries, archives, and knowledge organizations have long been built on shared assumptions about how information is discovered, organized, trusted, and shared. AI has not overturned those foundations, but it is challenging many of them—creating new expectations, responsibilities, and opportunities for information professionals.
Here are seven assumptions that recent research suggests are beginning to change.
1. Knowledge Discovery Begins with Search
The assumption
For generations, library catalogues, enterprise search systems, and knowledge repositories were designed to help people find information. Success was measured by how efficiently users could retrieve relevant documents.
The challenge
Many users now expect to ask questions in natural language, refine their thinking through conversation, and rely on systems to interpret their intent. Discovery is becoming less about retrieving documents and more about exploring ideas. Yet structured search remains essential for many research tasks.
What this means
Libraries do not need to replace traditional search; they need to support multiple ways of discovering knowledge. As conversational discovery becomes more common, the knowledge structures beneath it become even more valuable, even if users interact with them less directly.
The future of discovery may be conversational, but it will still depend on trusted, well-organized knowledge.
2. Better Access Leads to Better Decisions
The assumption
Improving access to information has long been one of the primary goals of libraries, archives and knowledge centres. Better search, broader collections, and faster retrieval were expected to support better decisions.
The challenge
We are drowning in information, but easier access does not automatically produce better decisions—we are still reliant on human authority and understanding.
What this means
As information becomes easier to generate and consume, professional value shifts from providing access to applying judgement. AI can summarize, compare and identify patterns across vast collections, but people still determine what is meaningful, what context is missing and which conclusions deserve to be trusted.
Access is becoming easier. Judgement is becoming harder and therefore more valuable.
3. Knowledge Systems Are Passive Repositories
The assumption
Knowledge repositories have traditionally been places to preserve institutional knowledge until someone needed it. Their role was to store information, support retrieval, and maintain a trusted organizational record.
The challenge
Knowledge systems are evolving beyond passive storage. AI-enabled platforms can recommend relevant content, surface relationships between information, and support decision-making through semantic search and AI-assisted discovery.
What this means
Knowledge systems are moving from the background to the foreground of organizational decision-making. As they become the interface to institutional knowledge, repositories stop simply storing information and begin shaping the answers people receive.
The future belongs to organizations that treat knowledge quality as strategic infrastructure rather than administrative housekeeping.
4. Professional Value Comes from Providing Information
The assumption
For many years, the value of information professionals was closely associated with managing information. Organizing collections, maintaining metadata and answering enquiries formed the foundation of professional practice.
The challenge
AI can summarize documents, generate metadata, answer routine questions and retrieve information in seconds. As these tasks become automated, the distinctive contribution of information professionals shifts from managing information to interpreting it.
What this means
Managing information is becoming automated. Managing knowledge is not. As AI takes on routine work, professional value moves further up the value chain—from organizing information to shaping trusted knowledge environments.
The organizations that thrive won’t simply have better AI; they’ll have information professionals who provide the judgement, context, and stewardship that AI still can’t.
Related reading: When AI Sounds Right But Isn’t: Simple Verification Habits Every Information Professional Needs
5. Expertise Develops Through Routine Work
The assumption
Many professions have traditionally relied on an apprenticeship model. Early-career professionals develop expertise through practical experience.
The challenge
AI is increasingly performing routine cognitive tasks. While this improves productivity, it also raises questions about how future professionals will develop expertise. Experience is built by questioning, comparing sources, recognizing uncertainty and learning from mistakes—not simply by receiving answers.
What this means
Organizations will need to become more deliberate about developing expertise. Mentoring, critical thinking, and opportunities to exercise professional judgement will become even more important.
The challenge is not simply preserving jobs—it is preserving the pathways through which expertise is built.
6. Technology Can Be Adopted Without Changing Institutional Values
The assumption
New technologies have traditionally been evaluated in terms of efficiency, cost, and productivity. Organizations could adopt new systems without fundamentally changing their mission or professional values.
The challenge
AI is different. It influences how information is discovered, interpreted, and trusted. Decisions about AI increasingly become decisions about privacy, transparency, intellectual freedom and accountability. Choosing an AI system is no longer just a procurement decision—it reflects what an organization values.
What this means
AI strategy is becoming inseparable from institutional strategy. Libraries and knowledge organizations must decide not only what AI can do, but what it should do.
As AI becomes embedded in everyday services, professional values move from the margins of technology projects to their very center.
7. The Goal of AI Is to Reduce Human Effort
The assumption
The promise of AI is simple: remove routine work so people can achieve more with less effort. The more decisions AI can make on our behalf, the more productive we become.
The challenge
As systems become “more intelligent”, there is a risk that people become passive recipients of knowledge rather than active participants in creating it. Some commentators describe this as a growing crisis of agency: a gradual shift in which the machine does more of the intellectual work while people become increasingly detached from the process of enquiry itself.
What this means
Libraries have never existed simply to provide answers. Their purpose has always been to cultivate curiosity, critical thinking and informed judgement. Preserving that human agency may become one of the profession’s most important responsibilities.
Success should not be measured by how much thinking AI can do for people, but by how well it helps people think for themselves.
Thoughtful Adaptation, Trusted Values
Libraries, archives, and knowledge repositories have long been built on shared assumptions about how information is discovered, organized, trusted, and shared. Those foundations remain remarkably resilient, but AI is challenging many of the assumptions built upon them.
The result is not that the profession needs to abandon its purpose, but that it needs to rethink how that purpose is fulfilled. As AI reshapes how knowledge is created, discovered, interpreted and applied, it also reshapes where professional expertise creates the greatest value.
Understanding AI now means understanding more than the technology itself. It means recognizing how it is changing the assumptions that have guided knowledge work for generations. The organizations best placed to realize AI’s potential will be those that adapt thoughtfully while remaining true to the values that have always underpinned trusted knowledge services.
0 Comments