KMWorld Conference Wrap-Up and Look Ahead
Text analytics is another area impacted by advances in GenAI. The Text Analytics Forum theme of New Synergies and New Solutions aptly described the current situation. Several talks, notably by David Seuss, Northern Light, and Alice Chung, Genentech, dealt specifically with the inte- gration of text analytics and GenAI. Seuss added tagging, machine learning, and search to the mix, stating that, working together, these technologies deliver the KM benefits expected by today’s users. Another trend, stressed by Chung, was the need for a realistic assessment of new technology. Understand limitations and don’t overhype what it can do is her advice.
Not particularly new or revolutionary were the constant reminders that data quality is essential. Lacking clean data, reliable information, and relevant knowledge, no addition of GenAI to a KM platform will result in high-quality results. GenAI deals in predictions to come up with its plausible answers. RAG is an enormous help in making those predictions more accurate and actionable. What GenAI can’t do is determine whether the LLM on which it has been trained contains correct information. That is up to human beings.
Evident in all the tracks and conferences was the trend toward recognizing the foundational value of KM to enterprises large and small, nonprofit and for profit. Without seamless access to knowledge—access that stretches across data silos and transcends formats—enterprises cannot please customers and employees as they would like. AI-based technologies contribute to elevating the value of KM, but they are not the only driving force.
Insights from the closing panel—Dave Snowden, Cyne- fin; Julie Mohr, Forrester; Dan Rasmus, Serious Insights; and Ross Smith, Microsoft—included their view of KM as exciting, recursive, and intriguing. We need awareness of the risks of AI. Tech is too focused on text, which is a fraction of relevant content. Humans don’t think like AI does and GenAI can’t handle novelty. Critical thinking is necessary, but COVID caused a delay in critical thinking and encouraged students to cheat, which has consequences for the future workforce. We need to re-examine measurement and rethink content. KM is evolving and pulling information from many different sources. People now do their own due diligence. Where is the trust? Discovery is more important than search. Connections and networking remain critical to knowledge sharing.
Returning to the KMWorld conference theme—KM & Enterprise Intelligence: Human or Artificial?—Lee Rainie, Elon University, posed a related question in his talk at Enterprise AI World: What is human knowledge good for in the age of GenAI? He sees hybrid intelligence on the horizon, which makes the special skills of KM’ers critical. The ability to be stewards of knowledge, promoters of truth, and creators of trust is what powers human intelligence in its relationship with AI. My takeaway is that it requires both human and AI for KM to succeed.
Companies and Suppliers Mentioned