The ‘Marriage’ of Artificial Intelligence with the Vedic Astrology Business (by Ranjan Pal and Richa Shukla)

Vedic astrological horoscopes have, since centuries, been the most popular tool available to the believer of Vedic astrology for self-reflection and event predictions on matters related to health, wealth, love, and career among many others. The reported reliance of the Indian (and sections of non-Indian) population in India (including many celebrities and VIPs) and abroad on Vedic astrology (a “super” science according to Shri K. N. Rao — a world-renowned Vedic astrologer; past chief-editor of the prestigious Journal of Astrology; and founder of the Bharatiya Vidya Bhavan Institute of Astrology in New Delhi) has grown rapidly since year 2000 with a potential of a continuing steep rise in the future.

The Status-Quo of the Vedic Astrology Business

To drive home this point, with the increase in the regular appearance of Vedic astrologers on public television and the Internet the Vedic astrology (VA) market has doubled in the past five years and seen a massive growth in the last 20 years (source: Google Trends). The current VA market valuation is in the USD 10 billion range (and growing) with about 2 million Vedic astrologers practicing in this market. Add to this, the addition in the VA industry of disruptive new players in the form of astro-tech start-ups that has often seen monthly download of their apps crossing 1 million on the Google Play store. The popular Vedic astrology app Astrotalk has seen a 20+ million customers of their service in the last five years (source: Puneet Gupta, Founder of the Astrotalk start-up). Similarly, the Astro start-up AstroYogi, registered a growth of around 300% YoY; is available to 85 countries around the globe in 11 languages; has a customer base of around 7 million; and is home to 15000 consulting astrologers who can talk to their clients on the cloud calling platform Talk to Astrologers (source: Meena Kapoor, CEO AstroYogi). Overall, most astro-tech start-ups have seen a growth between 10%-30% recently.

Can Artificial Intelligence Better Transform the Vedic Astrology Business?

Artificial Intelligence (AI) is widely being used around the globe in all industries imaginable due to its fast computing on massive data sizes along with machine intelligent data processing to output business insights for the social good not possible to be excavated by the average human mind. As an extremely potent example of the use of AI for the social good, medical diagnostics worldwide has immensely benefitted from the use of AI in improving diagnostic accuracy comparable to top human diagnostics and even better in many scenarios. So, a follow up question that arises is: like in the case of medical diagnostics, can AI transform the quality of the social good driven Vedic astrology business for the better, that in turn benefits the clients, the astrologers, and the astrology enterprises?

In this article, we wish to logically argue for and against leveraging the power of artificial intelligence (AI) in shaping the prediction and personal assessment efficacy of the Vedic astrology business (for the social good) that is pervasively driven by astrological software (e.g., Parashara Lite) and (cloud-based) apps (e.g., AstroSage, Astrotalk). We conclude that good AI and human astrologers need to necessarily complement each other to boost each other’s efficacy in the VA market. We start with a nutshell explanation, for the general audience, of the basic pillars on which Vedic astrological predictions/assessments are made that will be followed by briefings on AI technology and related arguments in support/against the use of AI for enhancing the quality of the predictions and assessments.

Basic Pillars of Vedic Astrology and Possible Number of Horoscopes

Vedic astrology (a Vedanga, i.e., a limb of the Vedas) follows a unique systematic methodology that is multi-dimensional and heavily parametric (across each dimension) to model the complex human life that is different for every human born — even if they are born at the same time and at the same place.

Broadly speaking, the aforementioned methodology is rooted in 9 basic (and not necessarily mutually independent) pillars (let us synonymize them with dimensions) of Vedic astrology. They are Grahas (planets), Rasis (signs), Bhavas (houses), Chakras (charts), Varga Chakras (divisional charts), Nakshatras (constellations), Purusharthas of life (dharma, artha, kaam, moksha), the Solar Calendar, and the Lunar Calendar. Each dimension/pillar is further complexified through sub-elements within each dimension — each sub-element with unique characteristics. For example, there are 9 planets (including Rahu/Ketu being shadow planets but excluding Upagrahas or special sub-planets), 12 houses, 27 constellations (excluding some special constellations), divisions of the Solar calendar (e.g., Horas) and the Lunar calendar (e.g., Tithis and Karanas) etc. Many of these sub-elements have further sub-divisions (e.g., a sign could resemble fire, earth, air, and water characteristics) making the dimensionality structure extremely intricate, recursive, parametric, and complex.

It is evident that once a person is born on earth, these 9 dimensions (along with their recursive sub-elements and parameters) get activated in a static (natal positions) and subsequent dynamic (transitory Gochara positions over time) manner. So, an extremely relevant question one might ask is: how many static horoscope configurations are possible? — each then driving a unique dynamic trajectory over time symbolizing unique human destinies. As an answer to this question, a detailed but approximate mathematical analysis based upon the applications of permutations and combinations to these dimensions results in a quantity that has at least 575 followed by 23 zeroes. However, the fraction of horoscopes that has been used up is a ratio of the total number of humans born till date to the number of unique horoscope configurations possible. This number, in percentage terms, equals a mere 0.00000000000018%.

What AI Tools are Applicable to the Vedic Astrology Business?

There are two branches of AI that are most applicable (when used together in sync) to Vedic astrology predictions/personal assessments: Bayesian AI (BAI) whose main prediction/assessment crux is based upon conditional probability theory, and Generative AI (GAI) whose main prediction/assessment crux is based on natural language processing and deep learning. As real-world working examples of AI-driven astrology tools (not necessarily Vedic astrology) include apps such as Co-Star, and JyotishGPT.

The relevance of the use of Bayesian AI in the context of Vedic astrology is the possible design and construction of prediction/inference AI models that consider (a) domain expertise (e.g., astrologer knowledge base, astrological sutras from classics such as the Brihat Parashar Hora Sastra) and (b) available human client data for each relevant astrological variable (see below for details) based upon the 9 pillars, to predict astrological events (e.g., marriage, career promotion) for a business client under consideration with significant accuracy guarantees. Examples of (open source) BAI software products in the market include BUGSBayesFusion, and BNFinder.

The relevance of the use of Generative AI in the context of Vedic astrology is its power to process and analyze vast amounts of data (e.g., planetary positions, a huge number of astrological sutras from the many classics written by Vedic sages and eminent Vedic astrologers, birth details of clients, predictions/assessments and life event details of clients mapping to horoscope combinations), quickly and accurately and feeding it (whenever necessary) to BAI engines pertinent to the 9 pillars. Moreover, in contrast to BAI, the step of identifying correlations and trends from multiple pillar inputs without the influence of human biases and misconceptions is done more effectively using GAI tools when compared to BAI tools. Examples (open source) GAI products include ChatGPTGPT-4, and Bard.

An Example of How AI Can Be Applied to Predict Astrological Queries

A common question that clients ask astrologers is: ‘when will my marriage take place and how will be my married life?’ To see how AI can complement the human astrologer in increasing the prediction accuracy of an answer, we need to first collect the fundamental VA factors based upon the 9 pillars that will allow one to derive an answer to this question.

The seventh house of an ascendant horoscope (commonly known as D1) broadly determines marriage related matters in Vedic astrology. In this case, due consideration must be given to (a) the seventh house, (b) the lord of the seventh house, © the significator of marriage (Venus), (d) planet occupants of the seventh house, (e) planets associated with the seventh house via astrological aspect, conjunction, and exchange effects, and (f) an analysis of the D9 divisional chart (commonly called the Navamsa chart) primarily used for assessing the spouse, the married life of a person, along with his/her second half of life. Each of (a) to (e) should also be micro-analyzed for the nakshatra (and nakshatra pada) placements of planets under consideration to derive precise answers for the static (e.g., how will be nature of spouse, timing of marriage) and dynamic (e.g., which times in life will be stressful for marriage) sub-parts.

In a non-AI environment, the queried Vedic astrologer will need to mentally and manually correlate the strength of multiple planets/houses/nakshatras related to factors (a)-(f) above simultaneously and derive an answer to the query based upon his/her years of experience in Vedic astrology. However, in an AI environment, each of these factors will be (i) first represented in an automated fashion via astrological variables (signifying the type, nature, and level of strength of planets associated with the factors) representing a statistics of mass of clients, (ii) second, the dependency among these factors (variables) across space and time will be computer generated by a suite of AI algorithms based upon expert knowledge gathered from astrological sutras and/or astrologers domain knowledge, (iii) third, the answer to the client query will be expressed as an end point variable of the dependency structure, and (iii) finally, complex probability and time series calculations spread across the spectrum of the astrological variables will output the chance of a particular outcome to the client query — dependent on the position of elements pertaining to the 9 interdependent pillars of a Vedic horoscope.

The Pros and Cons of Relying Upon AI for Astrological Predictions

The reliance of AI to make astrological predictions has two major benefits (pros) of significant impact to the Vedic astrology business.

Aiding Complex Calculations — The complex mental and manual calculations (e.g., winner of planetary wars based on the strength (bala) of competing planets in the same house, extent of planet neechbhanga, extent of cancellation of manglik dosha in females) pertaining to client queries on the matters of a given house — that traditionally is a mix of an astrologer’s intuition and concentration abilities (that can waver as ‘we are human’), can be done by a non-intuitive computing ‘giant’ whose output will be mathematically accurate, conditioned on good and verifiable AI astrological logic embedded inside the computing engine. While a good astrologer’s intuition based on years of experience might result in a somewhat equal accuracy window to that of well-trained AI-induced algorithms, the latter might be significantly better in compared to a beginning or mid-experienced practitioner (who comprise more than 80% of the Vedic astrologer base) and prone to prediction mistakes. In addition, for mathematically challenging horoscopes (and the odds are quite fair to randomly encounter one), even an experienced astrologer (leave alone the beginner) would be wise to combine his/her judgement on a query with the AI output.

Smart Identification of Complex Event Patterns — The mathematical identification of event patterns/outcomes and their individual probabilities for any given client query based upon simultaneously combining multiple astrological sutrasyogas, and dashas will be better achieved with improved clarity and accuracy by a well-trained smart AI astrological logic, when compared to even an expert astrologer with years of experience and needing lots of manual effort. More specifically, given the high number of outcome possibilities (in the worst case) for any given query, e.g., timing of marriage based on multiple dasha analysis paradigms (e.g., Vimshottari, Narayan Dashas) that is dependent on planetary placements in the birth chart) and more importantly their chances of occurrence for a random client (with million+ client possibilities), it is best that a well-trained smart AI model assist the astrologer in increasingly making the right predictions.

The use of AI to make Vedic astrology predictions has three major drawbacks (cons) nonetheless.

Data-Weak AI Training Models — The Vedic astrology (VA) market is mostly an ‘individual’-run business, where the individual astrologer has his/her set of clients. On the other hand, the use of AI is only effectively leveraging when there is sufficient and horoscope-diverse client data used for training AI models. Astrologer-customized AI software is likely to exhibit improper bias-variance tradeoffs if the client data spectrum is not diverse enough — resulting in AI outputting ‘incorrect’ chance outcomes on sensitive life/career impacting queries of clients. Here, a human intervention that combines expert astrologer judgement with AI answers is much necessary. Moreover. this calls for a need for (open source) astrologer crowdsourced and non-customized AI models that can train on a huge population of clients across astrologers (using privacy conscious tools such as federated machine learning).

(In-)Authenticity of AI Knowledge Base — It is not only the diversity of client data fed into AI engines that is important, but also the details of Vedic astrological principles as body of knowledge that forms the ‘library’ of AI engines. With so much and often conflicting astrology material on the Internet, it is likely that AI engines will be biased towards the wrong knowledge if their sources are not authentic and verified. Here, a human intervention that combines expert astrologer judgement with AI answers is much necessary. Care should be taken to have as body of knowledge the most authentic books and manuscripts by sages and legendary Vedic astrologers over the centuries — to be used by AI models.

AI Being too ‘Linear’ for the ‘Non-Linear’ Astrology — Finally, according to Richa Shukla “in astrology it is often that two plus two is not four”. The variations and complexities in the human horoscope could be too many for only smart AI-screened astrological sutras in general to be applicable for all horoscopes. As per Shukla (with 20+ years of professional experience), “at most 4/20 horoscopes today exhibit the linearity for AI to combine sutras directly to obtain accurate predictions/assessments”. The dependencies between (a) the sutras and (b) the 9 pillars necessarily require a human touch and psychology trained mind to improve predictions on any general horoscope.

In summary, AI can be a great complement to Vedic astrologers in the growing VA market and can help it grow at a significant pace. Our Vedic sages were God realized and cosmically conscious, i.e., were omniscient and had a mind of infinite processing power. In contrast, humans are just the opposite -possessing limited mental and psychological abilities. AI is a tool that can marginally allow the human astrologer to take a ‘baby’ step closer to the Vedic sage. However, AI should never be used a substitute of astrologer expertise given the big scale-accuracy challenges of AI training on a huge body of knowledge in a multi-billion+ diverse client base. AI should also not be a dis-incentive for (budding) Vedic astrologers to compromise quality of education and increasingly rely on AI to dictate predictions.

About the Authors

Ranjan Pal is with the MIT Sloan School of Management, USA. He is an ardent follower of Vedic philosophy and is a student of Vedic astrology under Richa Shukla — Founder of Dev Jyotish school of astrology, Gurgaon, India.

Richa Shukla is a professional astrologer based in Gurgaon with over 20 years of practice experience. She is the Founder of the Dev Jyotish school of astrology and was a student of Vedic astrology under K. N. Rao at the Bharatiya Vidya Bhavan Institute of Astrology, New Delhi, India.

Leave a Comment

Your email address will not be published. Required fields are marked *