Originally posted at LinkedIn

By Holly Zanville, Founder/Executive Director, Learn & Work Ecosystem Library 

Sometimes after a Zoom meeting I’ll say, “Thanks—I needed this therapy session.” Everyone laughs because they know what I mean.  We’re not talking about personal issues. We’re talking about making sense of an information world that seems to reinvent itself every day.

We’re comparing notes, recognizing the information bits we each have–realizing no one has the full story. 

We have to find ways to share information. How?  By visiting one another’s websites? Reading more reports and newsletters? Paying subscription fees to obtain access to news sources? Expecting journalism to do the job, of keeping us informed?

Many of us have concluded we need better, more reliable information hubs.  This is what we’re working on at the Learn & Work Ecosystem Library, with a mission to collect, organize, connect, and explain reliable information about the complex ecosystem. 

Each month we review an average of 50 new initiatives, organizations, research reports, technology platforms, coalitions and alliances, emerging terminology, and policy developments. Our goal isn’t only to collect information—it’s also to help people understand how the pieces fit together.

Some days it feels like trying to assemble a thousand-piece puzzle while someone keeps changing the picture on the box. And frankly, I’ve never been a fan of jigsaw puzzles–they require close investigation of irregularly shaped, interlocking, and mosaic pieces, which individually contain only a small portion of a picture, which cannot be understood until you solve the puzzle. 

Many mornings my inbox looks like this:

• “Groundbreaking initiative…”
• “Transforming the future…”
• “Revolutionary AI platform…”
• “New coalition announced…”
• “Learn more…”

Each announcement—each puzzle piece—sounds important and many are. But some overlap with existing efforts. Some are likely to disappear within a year because the project is receiving a 12-month grant so sustainability is questionable in our resource-constrained learn-and-work ecosystem. 

It’s time to make the “human judgement” call: Does the entry go into the Library’s collection? Before anything is added to the Library, we ask questions that many websites were never designed to answer.

• What exactly is this? A research initiative? A nonprofit? A commercial product? A pilot project? A coalition?
• Who created it and who is funding it?
• Is anyone using it?
• How does it compare with similar efforts?
• Is it genuinely new, or a new name for an existing idea?
• Will it likely exist three years from now?
• Why introduce another new term when similar concepts already exist?

Good curation does not seem to be about collecting answers really, but about asking better questions. Those questions define the work of our staff at the Library, where we try to help educators, employers, policymakers, researchers, workforce organizations, technology providers, foundations, journalists, and many others understand an ecosystem that is constantly evolving.

People often ask me two questions. “Isn’t all of this information already on the web?” And increasingly: “Can’t AI just tell us about this?”

These are fair questions. The short answer is ”not yet” and probably “not for a while.” The information generally exists but understanding usually doesn’t.

For example, a press release tells you what an organization wants you to know today. A website explains how an organization describes itself. Research reports answer specific study questions. News articles capture moments (stories) in time. Each serves an important purpose. None of them is designed to explain how everything connects. 

Here’s my simplistic explanation of roles:

• Websites are designed to explain one organization. Websites try to keep you on their site. Websites generally are not trying to explain systems outside their site or inclined to refer you to other sites that might be more helpful on a topic. 
• Libraries are designed to provide access to information resources and to explain the landscape. Curators of libraries try to connect you to resources outside it.

As curators, we often ask questions that are not easy to answer:

• Where did this effort originate?
• What larger movement is it part of?
• Who are the partners?
• How is it funded?
• Who benefits?
• How has it evolved?
• How does it compare with similar initiatives?
• Where does it fit within the larger learn-and-work ecosystem?

Those answers are not usually found in a single document. They’re assembled—from websites, research papers, newsletters, webinars, conference presentations, news articles, and sometimes direct conversations with the organizations themselves. 

Sometimes these different sources tell slightly different stories, and all of them are partially correct. Sometimes we discover there isn’t enough reliable information yet to justify creating a Library entry.

And now, artificial intelligence (AI) has become a member of our information-gathering team, like it or not. AI helps us compare definitions, identify related organizations, summarize lengthy reports, surface inconsistencies, suggest questions we might not have considered, and connect ideas across a rapidly changing landscape. It is fundamentally changing how we work. 

But AI still doesn’t decide what belongs in the Library. It doesn’t know which questions matter most to the stakeholder groups we serve. It doesn’t fully understand historical context, organizational relationships, or why one seemingly small distinction may matter enormously to a policymaker, researcher, employer, or educator. Those are editorial decisions. They require judgment, experience, and sometimes a healthy dose of skepticism. Perhaps most importantly, they require a practice of continuing to ask questions.  

That’s one of the biggest lessons AI is teaching us.  The best way to work with AI is not to stop questioning. It’s to question more—to ask for another source, to ask for another perspective, to ask why two answers differ and what may be missing and whether the evidence is strong enough to accept the information we have received? The quality of the answers often depends on the quality of the questions.

Another lesson learned: AI relies on us for information too. I raised this observation recently with our technology partner, “Maybe we’re trying to organize too much information?”

Their response: “Remember—you aren’t writing for only people anymore. You’re also writing for AI systems that decide for their own work which sources deserve to be trusted, are accurate, and well-organized. You’re writing for Bots too. What you bring to the Bots is contextual information that matters more than ever.”

I think about this conversation often. That we’re writing for Bots too. That our Library is not trying to compete with search engines. That we’re not trying to replace organizational websites. That we’re trying to provide something different—a reference layer that connects organizations, initiatives, glossary terms, research, and historical context so that people can understand individual efforts and the larger ecosystem. 

There is a bottom line here. Information has never been more abundant, and understanding has never been more valuable. Information and understanding are not the same thing. In a world overflowing with information, it’s useful to focus on the role of curation. Curation begins with good questions, and good use of AI does too.

So yes, every once in a while, after a Zoom conversation with trusted colleagues, I’ll smile and say, “Thanks, good therapy session.”  And we share the collective moment of the steep challenges faced in our information “jigsaw puzzle.”