Organize your curiosity: Generative AI tools prove adept at structuring volumes of information

I’m using AI to organize dozens of volumes of knowledge into a linear course of study that I’ll pursue over the next year. How I did it could be applied to structuring any large amounts of information, data or dilemmas.”

Chad Davis, chief innovation officer, Nebraska Public Media

Fun fact about me: I love used bookstores and thrift stores where someone behind the scenes has an eye for quality nonfiction. In a world where nearly any body of knowledge is instantly searchable (and even academic work is increasingly legible to lay readers) a finite shelf of books shaped by taste and timing feels like an adventure. 

A couple of years ago I was in a used bookstore in Manchester, U.K., and I ran across a copy of Marshall McLuhan‘s seminal “Understanding Media: The Extensions of Man.” I bought it as a totem, as much as anything, but it unexpectedly kicked off my collecting numerous books about media and its effects on people. I was trying to better understand the purpose (and therefore the future) of public media. If public media is going to remain core to society in the next 25 years, we need to understand the media environments that shape how people learn, debate, gather, belong and trust. To understand public media, I needed to understand media itself: what it does to us, what it makes possible and what it changes about being human.

My curiosity didn’t stop there. It led me to studies of print, of the written word, of language itself. I’ve accumulated dozens of relevant physical and digital books, and the stack keeps growing. To structure the information, I’ve turned to artificial intelligence (AI).

I’m using AI to organize dozens of volumes of knowledge into a linear course of study that I’ll pursue over the next year. How I did it could be applied to structuring any large amounts of information, data or dilemmas. 

 First, choose the learning tool. I chose Gemini because I’ve been increasingly using it as my day-to-day model, and because its ecosystem — especially NotebookLM — fits this kind of media-heavy learning project. Claude is excellent, but Gemini’s multimodal workflow is more useful for my purposes here.

 Give the chatbot your sources. For this learning project, I’ve got a mix of physical and digital books to read and typing in all the titles and authors would have taken a couple of hours. Instead, using my iPhone, I took pictures of the relevant sections of my bookshelves (it helps that I already group my books thematically, but AI should be able to filter out off-target and redundant titles), I took screenshots of my Kindle library, and I took screenshots of my reading wish list on Amazon. I then uploaded all of these to Gemini (limited to 10 pics at a time) with the following prompt:

You are an experienced philosopher of technology and media, and a communication studies professor. I am someone trying to develop a theory of public media in America for the mid-21st century, and I need help developing a self-directed reading list to help advance my thinking in an efficient and productive way.

Attached are photos of physical books I have on my shelf, titles I have in my Kindle library, and titles of interest I have saved in my Amazon wish list. Please read and consider all titles and suggest an order in which they can be read for maximum benefit to someone who is thinking about the future of public media in American society. Feel free to group titles together that might benefit from parallel co-reading. For each title, explain why it appears in the sequence where you have placed it. Feel free to take certain titles off the list, though please group those in a “No Need to Read” section of your response, noting either irrelevance or redundancy as the reason.

Before you begin, do you have any questions? 

I always like to ask the model if it has questions, because it opens a door to further clarification. Gemini asked for clarification, and its initial reply was fine, but not comprehensive enough, so I had to give it some corrective instruction. The new reply was more aligned with what I envisioned, so I tacked on a further request:

Do you have any other authors or titles you’d suggest? If so, please restate the list with your recommended phases and include a section at the end of each under the heading “I’d also recommend.”

Finally, preserve the context. After Gemini delivered on my instructions, I went on to ask if its recommended order was sacrosanct or if I could jump around. I asked it to choose five to 10 books it considered core to the collection. I also asked if I were to use a “parallel pair” approach to how it might pair up the titles on the larger reading list. Finally, I had it convert its suggested six phases of reading into a Google Sheet for me. 

For those seeking to lead in media, this may be one of AI’s more practical near-term uses: not replacing expertise but helping people build it more deliberately. In an age of abundance, your competitive advantage could be turning curiosity into structure, structure into sustained learning, and learning into leadership.

Chad Davis is the chief innovation strategist at Nebraska Public Media, where is he the founder of the Public Media Futures Lab and co-founder of the Public Media Innovators Peer Learning Community. Prior to that he was the inaugural head of Nebraska Public Media Labs, and led teams responsible for published experiments with AI, video games, virtual reality, 360-video and podcasting. He speaks regularly about the intersection of public media and emerging media and publishes a semi-regular newsletter under the Public Media Innovators banner.