A recent event hosted by the M.S. in Human Capital Management (HCM) program sought to get to the bottom of a very important issue: What if the most valuable insights in your organization are already there—just hidden within your existing learning materials?
For example, until recently, mislabeling a file and saving it to the office database might have spelled the end of that content item’s life cycle. Countless images, videos, or PDFs with vague file names were lost to time. But with recent advancements in AI that can scan and summarize vast amounts of information in mere minutes, organizations are rediscovering troves of buried treasure in the form of repurposable past content.
The hybrid event, Smarter Workdays: AI-Driven Content Intelligence for Learning Management, was held on March 5 at the Lee C. Bollinger Forum, a unique community gathering space located on the corner of 125th Street and Broadway on Columbia’s Manhattanville campus, and simultaneously live-streamed for online participants. Industry practitioners Nabeel Ahmad and Josh Bellis led the presentation.
Through an interactive demonstration of a special-purpose large language model (LLM), the event offered practical skills and training for anyone in the field of learning and development by showing how AI can transform the way organizations understand, use, and learn from their own content.
The “Black Box” of Learning Content
To ground the discussion, Ahmad began his presentation by screening a short video on neuroscience and relationship-building. Participants were asked to identify key takeaways and quickly responded with a few themes of the clip—highlighting how easily humans can extract meaning from content.
On the other hand, enterprise content is often a “black box,” identifiable and sortable only through metadata and manual tagging. Learning management systems (LMS) store vast libraries of videos, audio files, and courses, but lack the ability to fully interpret what’s inside them. As a result, organizations often operate with an incomplete picture of their own knowledge assets.
Using a live demonstration of the platform changeforce, Ahmad showed how newer AI can analyze a file—one generically titled with no description or metadata—and generate a comprehensive understanding of its contents in seconds, producing a suggested title, a detailed description, relevant keywords, a structured summary outlining key themes, and takeaways. It can also map the content to user-selected specifications and draw on established frameworks such as O*NET and other skills taxonomies.
The implications are substantial. Rather than manually reviewing and categorizing learning materials—a time-consuming and often inconsistent process—organizations can use AI to generate a first draft of insights, leaving humans to refine, validate, and apply judgment.
People are beginning to move beyond fears that AI will replace them and are seeing how “it's going to help me do the things that I want to do, which is more judgment instead of filling in content,” Ahmad said.
According to Ahmad, individuals within organizations are now asking: What can I do “if I can understand the skills that are being learned or acquired through this content? Can I repackage it and do something else with it to give it a longer shelf life?”
Reimagining Existing Content
Once organizations understand what their content contains, new possibilities emerge: repackaging long-form courses into shorter modules, aligning materials to evolving workforce needs, or extending the life of existing resources.
Ahmad demonstrated some surprising integrations of the changeforce platform within a learning context—displaying how the AI was quickly able to generate sample test questions about the content of the video. The discussion also touched upon how AI could be used to adapt content to different learning styles, preferences, and performance levels—an area already gaining traction in other educational contexts.
Faculty in attendance noted the potential for mapping connections across curricula: using AI to identify how concepts build over time, where redundancies exist, and where new opportunities for integration might lie. Others pointed to applications in human resources and professional development, such as analyzing large volumes of podcasts or training sessions to identify recurring themes and skill gaps.
From auditing existing course libraries to identifying gaps in training programs, when applied to learning systems, human capital databases, and myriad other organizations dependent on large quantities of content and information, AI offers game-changing ways to move from passive storage to active knowledge management.
About the Program
The Master of Science in Human Capital Management at Columbia University prepares graduates to be world-class HCM strategists able to address changing needs in building and motivating talented, engaged workforces in the private, public, academic, and not-for-profit sectors.
The program is available part-time, full-time, on-campus, and online. Learn more about the program here.
