Artificial intelligence, knowledge management, human experience and libraries

At Knowledge Summit Dublin the overlap of missions, mindsets and challenges between librarianship and knowledge management was an exciting revelation.


Billed as a knowledge management conference, Knowledge Summit Dublin , perhaps inadvertently, revealed important overlaps between the job skills and missions of KM practitioners and librarians. In a panel discussion about real-world challenges facing knowledge managers, Dave Snowden, founder and chief scientific officer of The Cynefin Company, exclaimed that he thought the best knowledge managers came from a library background. It’s the librarians who understand information–its organization, curation, trustworthiness and dissemination.

As the conference progressed, the librarian connection became more apparent. Bonnie Cheuk, Senior Director, Head of Business & Digital Transformation, AstraZeneca, proudly declared that she began her career as a cataloguer in a medical library. Although not a librarian, TJ Hsu, Director, R&D Knowledge Management, Amgen, referred to library literature searching as important to the company’s KM program and its adoption of AI technologies.

Library-oriented presentations

At the Inter-American Development Bank, Kyle Strand, Lead Knowledge Management Specialist and Head of Library, Inter-American Development Bank, described how generative AI is being used to add search functionality to its Publications Catalog. He spoke about how the technology is being used to ask questions and receive answer pulled from thousands of IDB publications, how it is being measured, and the role of knowledge graphs. Impressive was that their review of search results found no hallucinations, all results pointed to real documents.

Ana Sofia Walsh, Knowledge Management Director, and Dominic Dietrich, Lead Analyst, Fragomen, presented on the law firm’s "Immigration Knowledge Library," a comprehensive, internal resource. The library provides detailed information on immigration rules and procedures across many countries, for both internal usage to keep its attorneys up to date and external outreach to clients. The library was recently restructured and relaunched using SharePoint Online to enhance its search capabilities and user experience.

Value proposition

Even when the connection between KM and librarians was not explicit, there were similarities between what was being presented as a KM issue with what corporate librarians have dealt with for years. The conference sessions were in a flipped format–a 20-minute talk followed by breaking into small groups for discussion or an exercise, then regrouping for a final discussion. Thus, many of the overlaps between KM and libraries surfaced in the discussions, as conference delegates related session topics to events in their past (or current) work situations. From my perspective as a former corporate librarian, I found many comments could be equally applicable to KM departments or corporate libraries/information centres.

One example is proving the value of knowledge and information. This has long been a struggle for corporate libraries and the fact that they are dying out is proof that the arguments about their value have not been effective. Knowledge management operations face similar threats but so far have been better positioned to capitalize on their value proposition. Tying knowledge access to profitability and productivity resonates with management.

Sarah Pullin, Global Director or Knowledge, and Sanni Vanttinen, Director of Knowledge Operations, Baker McKenzi weighed in on delivering value through combining content and a data strategy. Their operational model involves reflective thinking, beginning with identifying what business needs and opportunities an be delivered through strategic KM and prioritized AI. They then work out the feasibility of successful deployment and use. Third, they consider the suggested action and plot onto a strategic action grid. However, they cautioned that a rush to AI often results in failure.

Intrinsic to the discussion of value is budget. Like corporate information centres, KM units can struggle to obtain the necessary funding for their projects. Sustainability depends on convincing the powers that be to provide sufficient resources. Paying attention to what is and is not quantifiable helps to justify budget requests. Aligning with the goals of the organization remains a key factor.

Impact of AI

The impact of AI, particularly gen AI, was inescapable. Discussions about whether AI was a strategy or a tool occurred many times without a clear resolution. That it has unrealized business potential was a common theme, but the speed of change could thwart AI initiatives, making them obsolete before they even get started. One suggestion was codifying tacit knowledge by having AI tap into conversations. This, however, raises ethical questions along with the possibility of AI misunderstanding the context of a discussion in a community of practice. On a somewhat more philosophical note came the question about whether humans are adapting to the technology or vice versa. Is the definition of a hybrid workplace changing from in office versus remote working to human versus non-human workers?

Inevitably, the question of management’s desire to replace paid staff with unpaid AI agents arose. Can’t companies replace KM programs (or libraries) with AI. IDB’s Strand confronted that very situation, as another IDB employee came to him with "proof" that Copilot could do what his department was doing and do it faster and cheaper. That proof dissipated when Strand pointed out the hallucinations and mistakes that Copilot made. As knowledge managers (and librarians), it’s critical to understand AI technologies and to know their limits.

AI is top of mind for knowledge managers and librarians. How the technology can help them do their jobs seems at war with wondering if the technology will take their jobs. Most at the Knowledge Summit conference leaned toward the former. They speculated about how agentic AI could be integrated into workflows without losing the empathy that humans bring to the job. Ethics and bias mitigation are beyond the understanding of gen AI, so information professionals provide the critical human in the loop element that validates the information and knowledge created by the technology.

Rebecka Isaksson, Knowledge Empowerment & Copilot Enablement Lead, Sulava Gulf LLC, summed up what many information professionals are hoping for when considering the future of AI at work. She said "Leveraging the best of humans and AI to enhance knowledge sharing".