Jonathan
Forum Replies Created
-
Well, after a couple weeks of going back to the roots of the situation and re-researching, I really think we’ve got a winner with the Oracle. It will be a bit different than previously imagined originally, however it is much more intelligent than I ever thought possible and is basically the digital version of the MIND of Scientology data, both Tech bulletins and lectures from 1950 to 1986, and ALL of the LRH Policy (Admin tech too). As was originally envisioned, everything is cross referenced and cited, references cited are downloadable at a click of your mouse, just as it would be in a College atmosphere. It’s pretty incredible to have such a capability at your fingertips. I’m truly humbled by the capability it has and what it means for study and correct application in the future with Independent Scientology for students of the subject, auditors,Pre-clears, and new public alike.😛
You can actually have an intelligent conversation on any subject with the references and dig as deep as you want in a particular dates, lecture series, or topics. Keep in mind when looking at the images below the data being used is only the Briefing Course lectures data for the sample output images. I have yet to train it on all of the Tech Volumes OR all of the Lectures, this is only the beginning. Just WOW!…
-
The Oracle is in it’s second phase of coding currently, we are adding new features to it and a much larger Pinecone vector database, giving it a lot more capability, as well as a better user interface.
More updates very coming soon!
-
Thank you!
-
Oracle Update:
We have the application installed on AWS at Amazon. It IS working. There are a few little bugs we are working out one at a time, but it DOES work! 🤓Even with just a few PDFs uploaded and digested, it’s incredible what we can do with it and what it means for Independent Scientology.
We’re also looking at getting PDF links in the references it provides that go to the AOGP qual library!
Another update tomorrow!😁
-
Status update on the Oracle coding step.
We will have the installation and delivery of the Oracle code, version 1.0, on Monday afternoon US time.
Uploading PDFs will commence on Tuesday in earnest, with a slight bit of testing prior to ensure functionality.
We had a very brief conversation with the programmer yesterday and everything is going great!
~ Jonathan
-
This next video in the hatting series gives great detail on how the LLM (in this video’s case Chat GPT, ours is actually more intuitive in some ways but they are very similar in most regards) uses the information once it scans the PDF’s by rasterizing (converts an image stored as an outline) into pixels that can be displayed on a screen, printed or interpreted into text in memory.
-
We upload it to the server, it scans the PDF (rasterizes them) and puts the contents in the LLM (large language model) database as vectors or points in a 3D “cloud” of information.
I’ll post a video on this for Wednesday’s video.
Production is under way on our Oracle platform as of yesterday! 😁🤓❤️
-
Here is the second in the series of short videos explaining where we are headed with our model and what it will be trained on.
Which will be the English language (think Key to Life), Dianetics and Scientology (everything LRH mentioned), and a few other things like all definitions that were used by LRH for real English words, metaphors, turn-of-phrase, etc.
Keep that in mind while watching. 🤓
Please give us your thoughts, also! We really want to start the discussion on this!
-
Thank you Richard, I will note that down as pages read.
-
Here is the first in a series of videos I promised last night to bring you up to speed on how all this works. this one is on Large Language Models. This is what we will be using and basically what all the media attention is about with ChatGPT for example:
-
Yes, most definitely! The issue we are currently having is that many of the PDFs we have in the field are images made into a PDF and not truly a PDF text. Which means they have to be OCR’ed (optical character recognition’ed) first by something like say the Adobe Acrobat DC. Which is a necessary evil. If anyone would like to help with the project, (it doesn’t take too long on each document) on these files please (in the photo) please let us know. We have more to add too to the dictionaries but this is a great start, the majority of these along with some of the original books need to be “characterized” from their image state to be ingested into the Oracle and then can be assimilated to then ask questions etc.
-
When I trained at FLAG, that was all that was available at the time, was the 1975 (BTB Vols) Tech volumes.
-
The BTBs (Board Technical Bulletins) were almost all originally HCOBs before becoming BTBs, so we are sticking with the original HCOBs instead. The BTBs were canceled by LRH because they were not authorized by him in the first place per Dan Koon, and they had a lot of alter-is in them also. One of the biggest alter-is’es was in Standardianetic Procedure, when Ron found out about it, that started the evolution that would become New Era Dianetics and its release not long after.
-
Thank you very much, Richard! Very much appreciated!
-
Thank you, keep em’ coming!
