I strongly encourage OpenAI to build a persistent, source-grounded learning and research environment comparable to, and ideally substantially better than, Google NotebookLM.
GPT-6-level reasoning is extraordinarily powerful, but reasoning alone is not the complete learning product. Google is building toward something potentially transformative by combining frontier AI with persistent source libraries, research, audio overviews, presentations, visual explanations, and increasingly sophisticated generated video.
OpenAI has an opportunity to build an even stronger version around its frontier models.
I would like to be able to create a persistent knowledge workspace containing books, academic papers, primary sources, notes, websites, lectures, and my own documents. ChatGPT should understand that corpus deeply and then be able to produce different educational experiences from the same verified source base.
The ideal system would support
source-grounded tutoring with precise citations
long-form educational podcasts
complete narrated educational videos
diagrams, animations, and visual explanations
presentations and lecture decks
quizzes and examinations
flashcards and spaced-repetition material
structured curricula from beginner through advanced levels
adaptive lessons based on what the learner has already mastered
persistent learning progress across sessions
the ability to ask questions during a generated lesson and have later material adapt accordingly
transparent separation between source material, model inference, and speculation
The most important part is integration. I should not have to separately ask one model to research, another system to make slides, another to synthesize speech, another to generate video, and another to keep track of what I have learned.
The experience should be closer to this
I provide the source material.
ChatGPT constructs a rigorous curriculum.
It teaches the material.
It produces a podcast when I want to listen.
It produces an illustrated lecture when visual explanation is superior.
It generates a full educational video when the subject benefits from animation.
It tests me.
It identifies weaknesses in my understanding.
It adjusts future lessons.
It preserves the source grounding throughout the entire process.
This matters strategically as well. A company does not necessarily need the single strongest model to dominate an important use case if its competitor provides a dramatically better integrated product around a slightly weaker model. If Google eventually combines NotebookLM with a model that approaches the reasoning capabilities of OpenAI’s frontier systems, the resulting learning platform could become extremely compelling for students, researchers, professionals, and lifelong learners.
I prefer ChatGPT and want OpenAI to remain at the frontier. That is precisely why I am submitting this request. I would rather see OpenAI build this category aggressively than watch another company establish a major advantage in AI-assisted learning while OpenAI possesses much of the underlying intelligence required to do something even better.
Please consider a true persistent AI learning workspace with native audio, presentation, visualization, assessment, and eventually full educational video generation as a major product direction. Thank you. 