University Student Implementation

University Student Implementation of Dharmic AI

This page presents a proposal with suggestions for how to implement a project like Susiddha AI by using university students who are working on an advanced degree in Computer Science (CS) or Business Administration.

Since the project involves Sanskrit and the Vedic literature, the students interested in working on it would probably be (but not limited to) people with Hindu background. There are many universities that have such a population of students, including in the USA where students with an F-1 visa can pursue a degree with on-campus study followed by several years of internship and continued education.

A university could permit some of these students to work on Dharmic AI as part of their master’s thesis/project. To begin with, the Susiddha AI project would only need a relatively small team, with three or four students from Computer Science and an equal number from the Business school. Also, two faculty members (one from CS and one from Business) would be involved as mentors and evaluators of the student’s work.

The students will set up the AI project to be a scaleable, long-term, open-source project, which ultimately will accept contributions from thousands of programmers and Vedic scholars. It should be noted that working on an open-source project is a valuable experience for students. Microsoft owns Github, the largest open-source repository, and developed the “Copilot” AI-assisted programming tool using all the code on GitHub. (Microsoft is now rolling out a more general version of Copilot in Windows.)

Note regarding “contributor”: this term designates anyone who contributes code (or text, service, etc.) to an open-source project. An example project is Chromium, an open-source web browser; both Google Chrome and Microsoft Edge are based on Chromium. This project has almost 3000 contributors (as of July 2023). It’s located on Github.[1]

Note regarding “GPT”: this term (as used in this document) indicates any Large Language Model (LLM) software, not necessarily the OpenAI version. LLMs appeared in 2012, and several open-source LLMs exist (and OpenAI’s GPT is based on earlier open-source LLMs). However, LLMs are NOT the “holy grail” of AI. Much better and more human-like AI is still being developed, and a Dharmic AI project will evolve as newer AI technologies appear.

Below is a list of tasks to be performed by the student team, roughly divided between the computer science and business sides. Members of both sides can collaborate on tasks, and they will make big decisions as a single team.

Computer Science team tasks

The computer science side would perform such tasks as:

  • Set up code repositories, and establish procedures for managing it, including approving contributions, etc.
  • List programming and other tasks that can be done by potential contributors, so that when people ask how they can contribute, there will be a list of discrete tasks they can start on.
  • Set up servers, probably in the Microsoft cloud which will use the GPT API. Note, Google and Amazon clouds will soon have similar API offerings.
  • Upload Vedic texts for GPT to learn. Initially it may be necessary to upload English translations, until the team figures out how to make GPT process Sanskrit.
  • Study the AI technologies and consider how they can be used to process the Vedic texts and make their knowledge available for decision making.
  • Create crowd-sourcing projects, e.g. to annotate Sanskrit texts, and create word for word translations. This can be done via platforms such as Zooniverse.

Business team tasks

The business side would perform such tasks as:

  • Set up a non-profit organization/foundation for the project; the students
  • should think big, something along the lines of the Linux Foundation[2] and the Python Software Foundation.[3]
  • Set up a structure to maintain the project’s knowledge. This will be used to make knowledge (about the workings of the project) available to contributors and to the next cohort of the student team. “Wiki” collaborative software may suffice for this purpose.
  • Solicit donations, apply for grants, brainstorm, etc.
  • Create a website worthy of such an important project. The “susiddha.org” website (which you are reading) is just a plain and simple website intended to present the vision of the project.
  • Name the project. “Susiddha” is the code name for the project to build Dharmic AI (and hopefully create an Avatar), but the team can decide on what the official public name should be.
  • Create social media presence.
  • Design and run marketing campaigns, both to get donations and to attract contributors to the project.
  • Set up business banking accounts, to facilitate accepting donations, credit card and PayPal transactions, etc.
  • Disperse money, e.g. for cloud server rental, website design and hosting, advertising campaigns, etc.
  • Set up communication channels. The marketing campaign will quickly result in thousands of people wanting to contribute from around the world. So, there needs to be a system to manage all the contactors and keep them informed. Also, a forum could be set up to allow contributors to meet and collaborate with each other.
  • Create a database of subject matter experts. One subject needing expert help would be the annotation and markup of Sanskrit texts for computer processing.

In order for the initial team to get started, some “seed” money will be necessary to pay the costs of: creating a non-profit foundation (including legal fees), building an attractive website, renting servers in the cloud, accessing GPT APIs, running online marketing and advertising campaigns, etc. This money will need to come from donors and/or the university. The student team (via successful marketing) will quickly get donations flowing in from a global audience, so that the project will sustain itself.

Going forward, the project needs a youthful, enthusiastic leader. The initial student team can search for a leader, and with luck, one of the students of the team will emerge as the obvious choice. Then, the university can give that student an internship position to work there as long as their visa permits. Having such a leader will give the project continuity between student cohorts and changing members of the project team.

The university will want to have a stake in the project, since it could bring attention and donations to the university. The university can house the open-source foundation, and designate faculty and staff to serve as officers and members of the corporate board.

If the initial team does well, they will have set up an infrastructure that will attract thousands of contributors from around the world, and let those people start on the list of tasks. Initially, it’s not necessary that progress on usable code be made, only that the project be “well begun”. The AI technologies are evolving quickly and getting more powerful, such that it’s hard to predict what will happen when this project gets thousands of bright contributors thinking about how AI and the Vedic literature can be combined to transform the world.

Thus ends the proposal for University Student Implementation of Dharmic AI.

Notes and References

  1. https://github.com/chromium/chromium
  2. https://en.wikipedia.org/wiki/Linux_Foundation
  3. https://en.wikipedia.org/wiki/Python_Software_Foundation