Decades of AI research inherited a view of mathematics as the primary foundation of reasoning. A quieter school of thought runs through figures such as Gödel and Friedman, whose work raises a possibility often overlooked in modern computing: that mathematics may not be the foundation of reasoning, but a consequence of the logical assumptions beneath it. In that view, mathematics is representational: a language built on top of deeper logical structures rather than the source of those structures themselves. Change the logic, and the mathematics changes with it.

Bhārat (India) always found logic, not mathematics, to be the more foundational. Few civilizations have invested as deeply in the study of logic. Not one school. Many — Kapila Maharishi's samkhya logic of transformations, Gautama Maharishi's nyaya logic of inference, Rishabhanatha's syādvāda logic of context-dependent attribution, Siddhartha Gautama's (aka Buddha) madhyamika approach to catuṣkoṭi logic (true, false, both, and neither), and many many more. Debated and sustained across decades, centuries, millennia, and often ages through intergenerational living traditions and institutions with generations of inquiry and commentaries. Never assuming that questions must collapse cleanly into true or false. Continuous states, uncertainty, contradiction, and context are treated as objects of study rather than problems to be eliminated.

Meanwhile, the modern world with newfound wealth decided to build compute around binary logic, inherited from Boole and operationalized by Shannon. That choice changed the course of human history. But it was a choice, not the only one available.

Today, something unusual is happening.

With AI, we are once again facing questions that do not fit comfortably into binary evaluation. At the same time, computation itself is expanding beyond the assumptions that shaped the digital era. Dynamical systems, continuous representations, stateful computation, and new computational substrates are forcing a reexamination of how intelligence may be represented and implemented. And ancient knowledge traditions contain logical frameworks built precisely for reasoning beyond simple true and false.

We see convergence.

Ancient knowledge. Artificial intelligence. Dynamic computation.

Three fields. Different starting points. Increasingly similar questions.

Modern computing emerged from one logical tradition. As AI and new computational paradigms expose the limits of inherited assumptions, we are interested in what becomes possible when a broader landscape of logic is brought back into the conversation.

The white space between them is research worthy.

We are not presenting a finished theory. We are inviting researchers to an open, multi-horizon research-to-build program.

Our goal is to write in the gaps, connect the a priori to the present, publish what we find, invite criticism, identify missing pieces, and build theses that bridge these domains. Philosophy studies logic. AI studies systems. Computing studies implementation. We are interested in the questions that emerge after those boundaries disappear.

From Bhārat to the world.

3 Tracks

Knowledge Theory

Logic systems practiced in Bhārat since antiquity — Nyāya's inference, Jaina's syādvāda, Buddha's madhyamika — formalize many-valued and graded truth, treating it as a native object of study rather than as noise to resolve. Not as approximations of binary logic, but as alternative logical foundations for reasoning in their own right. Most AI research focuses on improving models, algorithms, and mathematics. Much less attention is paid to the logical assumptions underneath them. We are interested in that layer. In what becomes possible when alternative logical foundations are treated not as philosophy, but as engineering.

Artificial Intelligence

Most AI systems today inherit their foundations from statistical mathematics, and with it the assumptions of binary logic. We investigate whether those foundations can instead begin with logic, treating mathematics as its representation — including whether many-valued frameworks such as catuṣkoṭi and syādvāda decompose into stable evaluative structures, and how embedding those structures into AI architectures changes what AI systems can represent, detect, and optimize.

Dynamic Computation

If logic determines mathematics, and mathematics determines computation, then alternative logical foundations for reasoning may imply alternative forms of computation. We investigate whether continuous evaluation requires computational structures beyond the binary, transistor-based paradigm of modern computing — like many-valued logic gates or continuous-state automata — and what those structures reveal about intelligence itself.

How We Work

We write. In the blank spaces. Publish. Share notes. Speak. Present. Debate. Until we know enough.

Then we build. 🚧

Decades of AI research inherited an assumption: that mathematics is the bedrock of reasoning. A quieter line of thinking pushes back on that. In this view, mathematics is more like a language, built on top of logic rather than the source of it. Change the logic underneath, and the mathematics built on top of it changes too.

Bhārat (India) always found logic, not mathematics, to be the more foundational. Few civilizations have studied logic as deeply, or for as long. Not one school. Many. Debated and sustained across decades, centuries, millennia, and often ages. Carried forward through living traditions and institutions of teachers and commentary, generation after generation. Continuous states, uncertainty, contradiction, and context were treated as things worth studying, not problems to be eliminated.

Meanwhile, the modern world, flush with newfound wealth and industrial power, built its computers around a single switch: on or off, true or false, one or zero. That choice changed the course of human history. But it was a choice, not the only one available.

Today, something unusual is happening.

AI keeps running into questions that resist a simple yes or no — judgment, nuance, partial truths. At the same time, computing itself is moving past the assumptions that shaped the last seventy years of machines: new kinds of computation built around continuous change, memory, and context rather than switches. And some of the most ancient knowledge traditions in the world built logical frameworks made precisely for reasoning beyond true and false.

We see convergence.

Ancient knowledge. Artificial intelligence. Dynamic computation.

Three fields. Different starting points. Increasingly similar questions.

Modern computing emerged from one logical tradition. As AI and new computational paradigms expose the limits of inherited assumptions, we are interested in what becomes possible when a broader landscape of logic is brought back into the conversation.

The white space between them is research worthy.

We are not presenting a finished theory. We are inviting researchers to an open, multi-horizon research-to-build program.

Our goal is to write in the gaps: connect ancient ideas to present-day problems, publish what we find, invite criticism, identify what's missing, and build the arguments that bridge these fields. Philosophy studies logic. AI studies systems. Computing studies how things get built. We are interested in the questions that emerge once those boundaries disappear.

From Bhārat to the world.

3 Tracks

Knowledge Theory

Logic systems practiced in Bhārat since antiquity allow for shades of truth between true and false—not as rough sketches of yes/no logic, but as complete systems of reasoning in their own right. Most AI research focuses on better models, better algorithms, and better mathematics. Far less attention goes to the logical assumptions underneath them. We are interested in that layer—in what becomes possible when alternative logical foundations are treated not as philosophy, but as engineering.

Artificial Intelligence

Most AI systems today inherit their foundations from statistical mathematics, and with it the assumptions of binary logic. We investigate whether those foundations can instead begin with logic, treating mathematics as its representation. This includes exploring whether many-valued frameworks such as catuṣkoṭi and syādvāda give rise to stable evaluative structures, and how embedding those structures into AI architectures changes what AI systems can represent, detect, and optimize.

Dynamic Computation

If logic shapes mathematics, and mathematics shapes computation, then a different logic may point to a different kind of computer. We investigate whether reasoning in shades of truth requires computing built differently from the switches and circuits we inherited—and what those new computational structures might reveal about intelligence itself.

How We Work

We write. In the gaps no one else is filling. Publish. Share notes. Talk. Debate. Until we understand enough.

Then we build. 🚧