Founder of NotebookLM: ‘AI is for brainstorming together, not to replace writing’
Steven Johnson is a veteran writer and technology researcher, as well as the author of several bestselling books on innovation. He is currently the co-founder and product management director of the NotebookLM project – Google’s “AI notebook” tool with millions of users, distinguished by its ability to process in-depth knowledge. Last year, Vietnam ranked in the top 10 for the number of users as well as the most viewed documents on the tool.
On the sidelines of the Google I/O conference held in the U.S. at the end of May, Steven Johnson shared with VnExpress about the future of note-taking and writing in the AI era.

Co-founder and product management director of the NotebookLM project, Steven Johnson, on the sidelines of Google I/O 2026. Photo: Luu Quy
Co-founder and product management director of the NotebookLM project, Steven Johnson, on the sidelines of Google I/O 2026. Photo: Luu Quy
– How did the journey of transforming NotebookLM from an internal experiment into a popular product unfold?
– Honestly, when I first arrived at Google, I thought this was just a fun experiment. I told myself that I would have the chance to meet interesting people and get my hands on the latest technologies in the world, and that was enough. The likelihood of NotebookLM actually being released was very slim, as Google has thousands of prototype products being developed simultaneously at any given time.
The turning point came in the fall of 2022. We brought a few early prototypes, still very rudimentary, to present to Josh Woodward (Vice President of Google, in charge of Google Labs) and CEO Sundar Pichai. As soon as he experienced it, Sundar said he was very impressed with the product’s capabilities. Josh came back to tell me that the CEO was genuinely captivated by the idea, and that’s when I woke up and thought: “Oh, so we are actually going to create a real product here.”
Looking at the millions of people globally using NotebookLM, I still feel incredibly amazed.
– As a writer, you must consider every word in your work, but you have created a tool that summarizes and processes information, helping people not to read. What will the future of this industry look like? Will we write for humans or to serve machines?
– That is a very deep question and something I ponder every day as someone who has spent a significant part of my career writing.
When I started building NotebookLM, one of my big dreams was to make it a tool that helps me become a better writer. I wanted it to support data research and inspire ideas. But the core goal and vision from day one have never changed: to create better, deeper, and more engaging books for human readers.
We are moving towards a world where, to some extent, writing must harmonize for both readers and the AI agent. But intellectually, what I crave the most is a connection between human beings. I use AI to amplify ideas, to deeply understand complex material, and to explain it back to readers. But my ultimate goal is to converse with humans, not machines. Therefore, I always use AI primarily for my readers.
– Many people use AI for rapid content creation, resulting in texts that tend to be formulaic and similar. What solutions do you have for this issue?
– The root nature and initial idea of NotebookLM is to personalize the artificial intelligence you are collaborating with. You provide it with your project-specific materials, clearly explain your goals and expectations, and the AI must have the capability to adjust itself to meet those needs and information flows. The next step is to expand its capabilities into creative style and personal voice.
I can take myself as an example. I am writing a new book using NotebookLM and feeding it a lot of data about myself, the works I have published, and how I want to approach this new project. I still directly put pen to paper, but when I ask AI to pose critical questions or brainstorm ideas with me, it responds with a sharp understanding of me.
It no longer provides generic, formulaic answers but actually aligns with my worldview and core vision for the work. I hope this will become the standard for anyone using AI for creativity.
– Although it is called “Notebook,” the writing and note-taking features are not central. Do you have a different vision for the product?
– This is quite a humorous reality. The tool is named Notebook; you can still write and take notes, but that feature isn’t yet front and center. For students, note-taking is a core and inseparable part of the learning and research journey. However, we don’t want NotebookLM to stop at being just a reading and summarizing tool. The goal is to make it a top-tier AI-supported note-taking space, an optimal interactive surface for writing, in addition to the available data analysis features. Users will definitely see significant upgrades in this area in the near future.
– How will NotebookLM change in the trend of AI agents?
– The upcoming experience on NotebookLM will give you the feeling that you are truly working alongside an agent – a collaborator who understands and knows how to flexibly utilize all the tools available within your notebook, from document creation to product output and research tools.
Let’s look at reality: if you want to conduct in-depth research in NotebookLM, you still have to actively input specific commands for the deep query feature. But in the next phase, the AI agent will automatically understand what you have done in the Notebook up to that point. It will scan and recognize where your source document is lacking, which data pieces your essay needs, and proactively search for and fill those gaps. In the near future, users will experience a series of breakthrough functions like this. It is truly exciting.