Today’s blog post is a playful yet insightful reaction to an article by Katrina Pugh, Jonathan Ralton, Marc Solomon, Andrew Trickett and Eve Porter-Zuckerman. '“Knowledge Managers Bring Collectivity, Nostalgia, and Selectivity to the AI Ecosystem.” Kate et al. (a friend and former colleague) repurposes“nostalgia” to mean the curation of historical exemplars in knowledge management – a smart strategy to improve AI outcomes. Here, I pivot from her technical use of nostalgia to the importance of emotional connection in knowledge management, coining the term “knowstalgia”.
Key points:
Nostalgia for Pugh et al.: In their framework, Nostalgia is about curating exemplars – leveraging past knowledge and best practices to train AI and guide current work. It’s a clinical approach, stripped of sentiment, ensuring AI systems know “what good looks like.”
Emotional Nostalgia in KM: By contrast, knowstalgia (knowledge + nostalgia) emphasizes emotional connections to knowledge. Human memory is profoundly boosted by emotion – as Don Draper famously said in Mad Men, “in Greek, ‘nostalgia’ literally means ‘the pain from an old wound.’ It’s a twinge in your heart far more powerful than memory alone.” Integrating this emotional aspect can make organizational knowledge more memorable, meaningful, and motivating.
Why Feelings Matter: Research shows we remember emotional events more clearly and longer than neutral events. Stories that move us are recalled “long into the future”. By weaving emotion into knowledge sharing (through stories, personal context, pride in past wins, etc.), we create durable knowledge that sticks.
Human-AI Connection: As AI grows more present in our work and lives, people increasingly use it for personal, even emotional support – in fact, “therapy/companionship” is now the #1 use case for AI in 2025[4]. This trend underscores that emotion is a key part of how we interact with tech. Knowledge managers should ensure that our AI systems and knowledge bases don’t ignore the human heartbeat behind the data.
Takeaway: Knowstalgia is about marrying Kate’s smart data-driven nostalgia with true nostalgia’s heart. The future of knowledge management – especially in an AI ecosystem – will benefit from bridging head and heart, combining factual exemplars with the emotional resonance that makes knowledge truly come alive.
From Nostalgia to “Knowstalgia” – A Playful Pivot
Confession before we begin: I've been tweaking this post for over a year now — way too long for a piece about not letting good knowledge gather dust. There's a certain irony in sitting on a blog about knowstalgia while it quietly ages in my drafts, so I'm finally hitting publish.
Last year, my former colleague Katrina Pugh, along with Jonathan Ralton, Marc Solomon, Andrew Trickett and Eve Porter-Zuckerman published a new paper, “Knowledge Managers Bring Collectivity, Nostalgia, and Selectivity to the AI Ecosystem.”. To sum up:
Collectivity: Working together with AI to make smarter decisions, while relying on real people to check and refine what the AI suggests. It’s about building trust and keeping humans in the loop.
Nostalgia: Training AI with carefully chosen past examples to help it learn “what good looks like.” This keeps valuable old knowledge alive and useful, not lost in the digital shuffle.
Selectivity: Being precise about what information and data gets fed into AI, so we don’t lose creativity or miss fresh ideas. Humans help by highlighting what really matters and spotting hidden gems the AI might miss.
When Kate et al. wrote about nostalgia as a crucial behavior for knowledge managers in the AI ecosystem, I was intrigued. Their nostalgia isn’t about sentimentality at all – they define it as “curation of exemplars”, meaning the deliberate gathering of high-quality historical examples to train and guide AI. In her article, an AI is continuously trained on human-vetted exemplars so that its outputs become more accurate and understandable. This Nostalgia approach essentially gives the AI a rich memory of “what good looks like” by feeding it curated past cases. As Kate and colleagues describe, this ensures even older-but-gold knowledge (which AI might overlook due to its age) is not forgotten.
Some records are oldies but goodies
In my experience with records management from organizations like Microsoft and AIIM I know there is often great value in older information. But often that older-but-gold knowledge is unassayed.
Most modern integrated generative AI systems, like Google Gemini or Microsoft Copilot, use some form of retrieval augmented generation, or “RAG”, to select a limited information substrate to expose to LLMs as they create responses to knowledge queries in chat. Most of these RAG systems are “graph-based” – meaning they determine relevance based on recency, frequency, and proximity. Usually this ensures results that are more relevant and personalized.
However, in practical terms this also means that the version of a contract which had hundreds of edits and dozens of views may look more relevant than the final, signed copy, which gets a record flag for retention and preservation.
Most graph-based systems use semantic indexing, meaning they’re based on multi-dimensional vector storage. However, much metadata is stored as lexical information in tabular or columnar form, including the record flag. Integrating lexical and semantic data is complex, but it offers the way to make sure that older data is when appropriate, is given its proper prominence in the world of generative AI. (This was finally added to M365 Copilot in October 2025.)
From nostalgia to knowstalgia
Kate’s team’s reimagining of nostalgia is smart – it removes the rose-tinted glasses and focuses on knowledge assets. But being a pun-loving knowledge manager myself, I couldn’t resist flipping the script: What about nostalgia in the classic sense? You know, that sentimental longing for the past that packs an emotional wallop. Is there a place for that feeling in knowledge management? I’d argue yes.
Let’s call it “knowstalgia” – where we intentionally evoke and harness emotional connections in managing knowledge.
Why? Because while Kate’s nostalgia gives us the brain (structured past examples), knowstalgia gives us the heart. And effective knowledge management in the age of AI needs both.
The Power of Emotional Connection in Knowledge Retention
Think about your own most vivid memories from work or life, Chances are they aren’t just spreadsheets and procedure manuals. They’re moments infused with emotion: the thrill of a project success, the sting of a failed experiment (and the lesson you learned), the camaraderie of a late-night team effort. Emotion is the secret sauce that transforms information into lasting memory. In fact, numerous studies have shown that our clearest memories tend to be of emotional events, remembered with more detail and longevity than neutral events.
In the realm of knowledge management, this means that knowledge attached to a strong emotional context is more likely to be remembered and reused. As one KM resource notes, “we connect with stories emotionally, and a story that has had an impact on us will be easily recalled long into the future.” A dry lessons-learned document might be read once and forgotten, but a compelling story with challenges, triumph, maybe a touch of humor or personal pride will stick. It’s why storytelling is such a powerful tool for organizational learning: stories create an emotional connection that pure data can’t match.
Don Draper’s famous quote from Mad Men captures this phenomenon beautifully:
“Nostalgia – it’s delicate, but potent… in Greek, ‘nostalgia’ literally means ‘the pain from an old wound.’ It’s a twinge in your heart far more powerful than memory alone.”
That twinge in your heart he describes is exactly what we, as knowledge managers, should not underestimate. It’s more powerful than memory alone because it is memory – supercharged with feeling.
When Knowledge Hits the Heart
Let’s put this in concrete terms. Imagine two knowledge base entries at your company describing how to handle a client crisis:
Entry A: a factual, step-by-step description of the crisis resolution process.
Entry B: the same steps, but introduced with a short narrative: “In 2019, our team faced a similar crisis. It was 2 AM when the alert went off… By dawn, we not only solved the issue, but half the team was cheering because we set a new standard. Here’s how we did it…”
Entry A gives you the what. Entry B gives you the what and the why it mattered. The second version might spark a sense of pride and camaraderie, even for someone who wasn’t there – it provides context, emotion, and a sense of shared history. Years later, which version of the knowledge are people more likely to recall when another crisis hits? Likely the one with the human story attached.
Storytelling and feelings matter. Remember the 2008 financial crisis? Andrew Ross Sorkin’s Too Big to Fail treats the 2008 crisis as narrative nonfiction, not a market recap. Instead of replaying the Dow’s daily swings, he reconstructs scenes from interviews, emails, call logs, and contemporaneous notes - recreating conversations and layering in concrete, human details (the late‑night takeout, the stale conference‑room air, the rain outside, the clipped asides between exhausted principals). That scene-by-scene approach lets readers feel the pressure, power dynamics, and tradeoffs as they unfolded — who paused, who blinked, what was said and what went unsaid — turning abstract balance‑sheet risk into character, conflict, and consequence. The result is a fly‑on‑the‑wall narrative that explains not just what happened but why it felt inevitable in the moment. Far more compelling (and illuminating) than any end‑of‑day DJIA printout.
This is knowstalgia in action: infusing knowledge with narrative and emotion to make it resonate. It’s not about indulging in fuzzy sentimentalism for its own sake; it’s about leveraging our human nature – we’re wired to remember feelings and stories – to make knowledge management more effective.
Benefits of emotional connection in KM include:
Better Recall: Knowledge that carries emotional weight has bookmarks in our brains. We can retrieve it faster because it made us feel something. As neuroscience and experience both show, those emotional “tags” are potent retrieval cues.
Knowledge Stickiness: Just as a catchy tune stays in your head, a powerful story or meaningful anecdote will stay with an employee. That means critical lessons or best practices are genuinely learned, not just filed away.
Human Context for AI: If we document not only what worked, but why it was important to people, we provide richer context. In the future, AI that parses our knowledge repositories might discern not just patterns in data, but patterns in what humans value and celebrate. (More on this shortly.)
Engagement and Culture: Sharing knowledge in an emotionally engaging way helps build a knowledge-sharing culture. People bond over stories. A repository of sterile documents doesn’t spark anyone’s imagination, but a living library of team “war stories” and proud moments can inspire new employees and veterans alike to contribute and learn.
Bridging the Gap: Data, Emotion, and the AI Ecosystem
The Pugh et al. article ultimately is about how humans and AI can form a better partnership in knowledge work. Their collectivity, nostalgia, and selectivity are three ways to keep AI’s outputs accurate, relevant, and innovative by injecting human judgment and curation. In practice, her nostalgia-as-exemplars approach is already an antidote to AI’s tendency to lose context. For example, large language models often prioritize recent or surface-level info over the truly relevant knowledge. By training AI on a trove of vetted historical cases, we counter that recency bias and ground the AI in enduring principles and past wisdom
Now, consider layering knowstalgia on top of that. If those exemplars were not just data points but carried annotations of their significance (“this project was a breakthrough that saved the company” or “this design was beloved by customers, we still talk about it”), then the AI isn’t just learning from a knowledge base, it’s learning from a knowledge culture. It starts to see what humans found meaningful, not just what met a KPI.
We’re venturing into speculative territory, but it’s not far-fetched. AI researchers have noted that as people use AI more intimately, the lines between cold data and warm emotion blur. In fact, a recent analysis of generative AI usage showed a “marked transition from primarily technical and productivity-driven use cases toward applications centered on personal well-being, life organization, and existential exploration.”
Today’s AI can be coach, companion, and confidant. Case in point: therapy and companionship have hit the #1 spot in HBR’s list of top AI uses, with life organization and finding purpose also in the top three. People are pouring their hearts out to ChatGPT and its cousins for advice, ideas, or just to vent – effectively building an emotional rapport with a machine.
What does this mean for knowledge management? It tells us that people crave a human touch, even from high-tech systems. If our knowledge platforms and AI assistants at work remain soulless data dumps, they’ll be underused and uninspiring. Conversely, if we design them to acknowledge and incorporate human emotion – from how we curate content to how AI responses are tuned – they stand to engage users much more deeply.
To be clear, I’m not saying your company’s AI should start cracking jokes and giving hugs. But it could prioritize answers that come with relatable examples or highlight the human impact of a piece of knowledge. An AI-powered knowledge tool might say, “Here’s a solution, and by the way, it’s the same approach that won us Client X’s appreciation last year.” Little touches of knowstalgia can remind us that knowledge isn’t just an abstract thing – it’s tied to real experiences and values.
Cultivating “Knowstalgia” in Practice
So how might organizations bring a bit of heart into their knowledge management? A few ideas:
Story Banks: Encourage teams to document key learnings in story form. Instead of just writing “Did X, got Y result,” ask: What was at stake? Who was involved? How did it feel when it succeeded (or failed)? Even a short paragraph of context can turn a bland wiki page into a memorable tale.
Capture the “Why”: When saving best practices, include a note on why it was a “win.” “This process saved 50 hours of work and everyone was relieved.” Over time, patterns of what your organization collectively cheers or laments will emerge – an emotional layer of metadata.
Use Media: Don’t limit knowledge to text. A 2-minute video of a team lead excitedly explaining a breakthrough, or a snippet from a town hall praising a project, can be gold. Voices and images carry emotion that text sometimes can’t.
Tacit Knowledge Sharing: Create forums (online or in-person) for employees to share war stories and lessons learned. Often, the most valuable knowledge is exchanged through anecdotes over coffee. Try to capture that somehow – perhaps an internal podcast or newsletter featuring these stories.
AI Training with Care: If you’re using AI to surface knowledge (say, an agent that answers employee questions), train it on not just the policy documents, but also on those stories and narratives. Test if it can cite examples that resonate, not just recite the rule. The goal is an AI that can say, “Here’s the info you need, and here’s a quick story of how it helped in the past.”
Conclusion: Embracing Knowstalgia – The Past, with Feeling, for the Future
Knowledge management has always been about connecting the past to the present to improve the future. What Kate and her co-authors remind us is that we need to do this thoughtfully: using our collective wisdom (collectivity), our curated past successes (nostalgia-as-exemplars), and our strategic insight (selectivity) to guide AI and ourselves.
My friendly addition to the mix is to remember that nostalgia isn’t just a repository of examples – it’s also that wistful affection for the past that Don Draper waxed poetic about. That emotional side of nostalgia isn’t a weakness; it’s a powerful human feature that can make our knowledge endeavors more robust.
In a sense, knowstalgia is about building a memory palace for your organization – not just a database, but a place where each important piece of knowledge has stories and feelings attached. It’s a place people visit and come away not only informed, but inspired or moved. And in the age of AI, where the risk is that technology makes everything impersonal and instantaneous, cultivating an emotional connection to knowledge ensures we don’t lose the meaning in our information.
To all the knowledge managers, leaders, and technologists reading: next time you celebrate a win or learn from a failure, think about how you can bottle a bit of that emotion and store it in your knowledge shelf. Years later, when someone uncorks it, that echo of human experience might be exactly what they need to truly understand and remember the knowledge inside.
Nostalgia as Kate defines it will make our AI smarter. Knowstalgia, as I propose, will make sure our knowledge ecosystem – AI + humans – stays wise, warm, and authentically human.