Long-form inquiries built in Google NotebookLM. Public summaries are free to read. The underlying sources, the chat history with the notebook, the citation map, and the methodology page are available for a one-time unlock. No subscription. You pay per notebook. You keep it forever.
A three-week mapping of what the academic and industry literature on visual AI evaluation actually says about ground truth. Forty-one sources, annotated. The consensus turns out to be much thinner than most practitioners assume.
A reading through a century of Carl Zeiss's technical and philosophical publications on optics and human vision, looking for what they knew about perception that computer vision has quietly forgotten. Some of it is remarkable.
Starting from Ray's Sonar Kella and its measured effect on Jaisalmer tourism, this notebook traces how specific regional stories have materially changed the economies of the places they were set. Includes a working dataset of twelve comparable cases.
§ How the unlock works
Each notebook has a free public summary (what you see on its page). The paid unlock gets you: the full NotebookLM chat history as a browsable archive, the underlying source documents where licensing permits, a citation map visualising where each claim came from, a methodology page explaining how I ran the inquiry, and any spreadsheets or datasets built along the way.
Payment by Gumroad. One-time, no subscription. Complimentary access for working journalists and students by emailing vireshwardas@outlook.com with a short context line.