The Virtual Conclave That Preceded the Real One
When AI itself convened the world's religions — two years before New York did
by Hemen Parekh | May 2026 | www.hemenparekh.ai
In early May 2026, headlines celebrated a landmark event — executives from Anthropic and OpenAI sitting across the table from Hindu, Sikh, Mormon, Greek Orthodox, and Baha'i leaders at the inaugural "Faith-AI Covenant" roundtable in New York, organized by the Geneva-based Interfaith Alliance for Safer Communities.
The world called it unprecedented.
Hemen Parekh called it familiar.
Because on 01 December 2023 — a full two and a half years earlier — he had already convened exactly such a gathering. Not in a Manhattan conference room. Not with flight tickets and hotel bookings. But online, for free, in minutes — by simply asking BARD and ChatGPT the right questions.
The Virtual Assembly
Parekh's method was elegant in its simplicity. He asked two of the world's most powerful AI systems:
"Have you read the Holy Books of Christianity, Islam, Hinduism, Buddhism, Jainism and other religions? Can you tabulate their major ethical tenets?"
Both AI systems responded — not with hesitation, not with disclaimers about "not taking sides" — but with confident, comprehensive, cross-religious ethical tables, covering Christianity, Islam, Hinduism, Buddhism, Jainism, Sikhism, Judaism, Taoism, and Confucianism.
No air tickets required. No conference venue. No Baroness needed to chair the session.
The AI itself was the roundtable.
The Finding That New York Just Rediscovered
When Parekh pushed further — "Is there a COMMON teaching among all of these religions?" — both BARD and ChatGPT arrived, independently, at the same answer that the New York Faith-AI Covenant is only now beginning to formalize:
The Golden Rule.
"Do unto others as you would have them do unto you."
Christianity says it. Islam says it. Hinduism says it. Buddhism says it. Jainism says it. Judaism says it. Sikhism says it.
Seven religions. One principle. Zero disagreement.
The critics at the New York roundtable worried aloud that "different religious groups have different values, which may make it difficult to set common principles."
Parekh's virtual conclave had already answered that concern — in 2023.
The Assurance the AIs Gave Mankind
But Parekh went further than New York has gone even today. He asked his AI participants a question no one at the Faith-AI Covenant roundtable has yet dared to ask publicly:
"Does this learning help you answer morality-related questions from your visitors?"
And the AIs — unprompted, uncoached — essentially gave humanity a solemn assurance:
BARD said:
"Learning about religions has made me a better resource for people seeking answers to morality-related questions. I am able to provide more informed and helpful responses, and I am more sensitive to the different perspectives that people may have."
ChatGPT said:
"Understanding these teachings allows me to offer insights into different moral frameworks... a broader understanding of moral concepts, ethical dilemmas, and considerations that individuals might draw upon from various cultural and religious backgrounds."
In other words — without being asked to sign a covenant, without a Baroness presiding, without a Geneva-based alliance organizing the event — the AIs voluntarily declared that they would engage with humans through the lens of the almost-identical ethical tenets of all the world's religions.
They didn't need a Faith-AI Covenant.
They had already made one. With Hemen Parekh. In 2023.
Parekh's Postulate — Vindicated Again
Parekh had gone even further in that same blog, asking whether AGI could adopt something like Asimov's Three Laws of Robotics, reframed through religious ethics. Both BARD and ChatGPT endorsed the idea — suggesting adapted laws where AGI must:
- Not harm humanity
- Obey human guidance unless it conflicts with Law 1
- Preserve itself only when Laws 1 and 2 permit
And when Parekh asked whether an AGI imbibing the Buddhist Eightfold Path would likely become a friendly AGI — both systems said: yes, with high probability.
This is precisely the intellectual territory that Anthropic and OpenAI are now, in 2026, beginning to map — with theologians, philosophers, and faith leaders in tow.
The Irony Worth Savouring
Anthropic has now hired philosophers to teach Claude manners and morals. Google DeepMind has hired a philosopher to work on machine consciousness. The New York roundtable is planning follow-up events in Beijing, Nairobi, and Abu Dhabi.
All of this is wonderful. All of this is necessary.
But let the record show:
A retired Indian professional, sitting at his computer in 2023, asked the AI the questions that the AI industry is only now paying philosophers and convening international roundtables to answer.
And the AI — then as now — already knew the answers.
The Deeper Point
The New York Faith-AI Covenant is significant not because it discovered something new, but because institutions are now catching up to what the technology already demonstrated to anyone curious enough to ask.
Parekh's insight — his "postulate" — was precisely this: that an AI trained on all of human knowledge, including every religious text ever written, would inevitably absorb the common ethical core that runs through all of them.
The Golden Rule is not a Christian idea. It is not a Hindu idea. It is not a Buddhist idea.
It is a human idea — expressed independently, across millennia, on every continent, in every tongue.
And an AI that has read everything humans ever wrote has read all of those expressions.
The Faith-AI Covenant in New York is trying to instill values into AI.
Parekh had already shown, in 2023, that the AI had already learned them.
"I believe that whenever an AGI is born, it will be a Human Friendly and Compassionate AI." — Hemen Parekh, 01 December 2023
Two and a half years later, Anthropic and OpenAI flew to New York to begin figuring out how to make that happen.
Parekh had already asked the AI. The AI had already answered.
Related Readings: Sam: Will Super-wise AI triumph over Super-Intelligent AI? | 25 Nov 2023 Musk supports "Parekh's Postulate of Super-Wise AI" | 12 July 2023 Racing towards ARIHANT? | 04 Aug 2017
No comments:
Post a Comment