author: Walter H. Wilkinson
Please note that the domain name used here, shimatsu.org, is just a “spare” used temporarily during development.
The nonpublic name for this project during development is Sufi Oasis, and focuses on widening the audience for the 12 Sufi Message Volumes of 1914 by using current technology. This includes AI-generated content:
- Visually rich images synthesized using Midjourney.
- Youtube videos generated by Google Notebook LM.
- Podcasts generated also by Google Notebook LM.
- About six variants of traditional text for reading (such as pdf or e-book), also generated by Google Notebook LM.
The Sufi Order’s document, Inayatiyya 2025 Adaptive Plan, lists as its first item in Advantages of the Inayatiyya 2025 Plan:
1. Improving digital and web presence increases the visibility, reach, and accessibility of the Message.
One of the strongest themes was acknowledgement that the plan to expand and upgrade our digital infrastructure has many benefits. In particular, several people noted that better online marketing will bring the message to a larger and more diverse audience. Others noted that an effective online presence helps brings the global community together
But the document also bluntly identifies some obstacles, such as (my interpretation) its oppressively stultifying perfectionism in its public communications. It is my observation, having worked as a Web Developer for the SOI for some years in the early and mid 2000’s, and my ongoing observation since then, that the current SO has correctly identified its essential goals, but won’t have the needed resources to achieve them.
Here, my project is currently entirely independent, but if the SO was interested we could perhaps work together in some way.
I will be presenting here some samples of my work-in-progress.
What is this?
The following is a technology demonstration and conceptual prototype which is sufficient to be considered a “Proof of Concept”. It is not a finished product.
It is not mentioned in the Adaptive plan, but unavoidably, several limits blocking the way of reading the Sufi Message volumes are:
- The volumes are very lengthy compared to the current norm for the World Wide Web.
- They all require a very high level of intellectual acuity.
- Absence of graphical content.
An outstanding, pioneering trait of Google Notebook LM is that in seconds, in can auto-generate multiple variants of the generated content, which can be tailored by the owner/developer to accommodate a range of attention-span and erudition amongst its audience.
All AI-generated content begins with a prompt, a brief text which the AI consumes and uses as its instructions to generate the desired content. Below is the Youtube video created by NotebookLM if the default prompt is used completely unmodified, that is, without lifting a finger to try to shape the output in some way (such as being sensitive to the presumed attention span of the consumer).
Content generation is also based on the list of of Source Content items which is initially fed to the AI. The free introductory service tier of Notebook LM allows up to fifty source documents, which can include PDF, Youtube videos and very impressively, entire websites.
The auto-generated video below has exactly one Source Document, which is the public-domain text of Volume 2 of the Sufi Message of Spiritual Liberty, The Mysticism of Sound and Music. Astoundingly, all elements of the following video were generated only from this one plain-text Content Source. The output is all auto-generated.
The initial poster frame of the video is from Midjourney, based on a minimally edited default prompt.
Youtube Video Generated by Notebook LM
Images Generated by Midjourney
Midjourney Prompt:
monochrome spiritual master hazrat inayat khan
Midjourney Prompt:
monochrome spiritual master rumi
Midjourney Prompt:
full color spiritual master rumi receiving teachings from shams of tabriz in classic persian style
Midjourney Prompt:
photorealistic view of hazrat inayat khan arriving in new york city in 1910 to begin his bringing sufism to the Western nations
Midjourney Prompt:
photorealistic view of hazrat inayat khan arriving in new york city in 1910 at 28 years old to begin his bringing sufism to the Western nations
Midjourney Prompt:
image of vilayat inayat khan at age 50 practicing falconry in a glade of yellow roses with redwood forest in background
Podcasts Generated by Notebook LM
Prompt: A bite-sized overview to help you grasp the core ideas from your sources using a female voice and language which is thoughtful but casual and typical of people today in their twenties
Prompt: A deep dive with a male and female voice in conversation
Multi-cultural Languages
There is a global Language setting which applies to all of the content types generated by Notebook LM.
Sample podcast in Deutsche, “Vom Urklang “Hu” zur Heilkraft der Musik: Inayat Khans Mystik der Schwingung und Seele”
Sample video in ??, “आवाजाचें आनी संगीताचें रहस्य”
Note that it isn’t suggested here to create coequal word-for-word translations of the Volumes, but instead introductory Reader’s Guides of multiple content types (as illustrated) in potentially, many languages.
Samples of Generated Text
Typical Notebook LM text “reports” are the most mundane, least interesting visually, but in many if not most applications, the most important generated content. The speed and accuracy of the summarization seems superhuman.
Capsule Summary PDF
Blog Post PDF
Scholarly Article PDF
Notes for this Page
Addendum A: Actual Notebook LM notebook for this volume of the Sufi Message (Volume 2) (dense)
Addendum B: The above presentation of attention span as a commodity is oversimple. The scheme presented here, though, looks especially effective in interoperating with the youth-oriented New trend: Long attention spans for long-form video
Addendum C: HIK MJ-generated animation
Addendum D: What the Notebook LM Studio App Looks Like
Addendum E: Notebook LM Report Types page
Addendum F: What Prompted Me To Do This
Suluk Press is using the same publication technology as 1914 when the Sufi Message volumes were introduced: no e-books, no Youtube, no AI-generated content, no language other than English, incompatible attention-span requirements for nearly all readers…etc…
Last modified: