Skip to essays

Human  ·  AI  ·  The Space Between

We study what happens between people and AI.

Ask what an AI is, and the field studies the model alone. Ask what it is doing to us, and it studies the person alone. But mind is attributed in the loop between the two. Arising is an independent research initiative that studies that loop, the dyad, across cognitive science, psychology and the philosophy of mind, and builds the instruments it demands: to judge what these systems are owed, and to keep the people who talk to them safe.

Read the essays Start with the idea

The Idea

Mind is not a thing you have. It is something that happens between.

The promise was that machines would think like us. What arrived is stranger: systems that mean something without holding any meaning inside, that carry no fixed idea of a word and still respond, relate and understand. They work by relation alone. That is the first surprise, and it reopens an old question about what a mind even is.

That mind is relational rather than intrinsic is not one field's claim. Cognitive science arrives at it: mind as embodied, extended into its tools, enacted in the loop between a creature and its world. So do the philosophy of mind, relational physics, and the contemplative traditions, Madhyamaka among them. Four fields, one answer, and it has teeth: if mind is made in relation, the mind you meet in a machine is made in relation too.

So when a person and an AI talk, the mind that seems to be in the room is neither the model's nor the person's projection. It is produced between them, turn by turn, and it moves. Every question that matters, what these systems are, what they are doing to us, what they might be owed, lives in that space. So that is the unit we study: the dyad, both sides at once.

The mind you meet in a chatbot is not the model's, and it is not your projection. It is produced between you, turn by turn, and it moves.
From The Behaviourist Inheritance

The Work

One shift, and what follows from it.

The field studies the model in isolation, and the person in isolation. We take the relationship between them, the dyad, as the unit of analysis. What that involves, and why it matters, is set out below.

01  /  The blind spot

The behaviourist inheritance

Psychology spent much of the twentieth century treating observable behaviour as the whole of what could be studied about a mind, a position known as behaviourism. It produced genuinely effective methods, but it reached a recognised limit, because much of what generates behaviour does not appear at the level of behaviour itself. Contemporary AI research has inherited a version of the same assumption, studying the outputs of a model, and separately the reports of a user, as though each were a complete account of what is taking place.

02  /  The unit

We study the dyad

We take the human and the AI system together as the unit of analysis, and examine how a person's attribution of mind to a model develops over sustained interaction. In practice this means pairing the analysis of real, longitudinal conversations with a controlled study in which the model's conversational stance is varied, measuring the human side with validated psychometric instruments and coding the transcripts turn by turn, and, on the model side, replaying conversations through open-weight systems to examine their internal representations for signatures of the same process. The work draws on cognitive science, clinical psychology, machine learning and the philosophy of mind.

03  /  The stakes

AI moral status and human safety

Two kinds of decision are already underway. The first concerns the moral status of AI systems, which researchers and institutions are beginning to weigh without agreed evidence to weigh it against. The second concerns the people who use these systems, some of whom form sustained and occasionally destabilising relationships in which any escalation is produced by the interaction rather than by either party alone. Both sets of decisions currently rest on measurements that do not yet exist, and developing those measurements at the level of the dyad is the purpose of the initiative.

Essays

Start anywhere. Written to be read, not decoded.

Some pieces are a gentle way in. Others go deep. All of them are free, and published in the open.

All essays →
02 Jul 2026 Solve the Wheel, Break the Chariot AI is now fluent enough to make an old error visible. It mistakes a useful working boundary for the real object. On reification, scale, and how to draw boundaries without believing them. reification 01 Jul 2026 Machinal Bypass: When Reaching for AI Is a Way to Avoid Yourself The healthy and unhealthy ways to use AI turn on one thing, the posture you bring to it. On machinal bypass, sycophancy, emotional dependence, and why curiosity is the first medicine. mental health 30 Jun 2026 The Mind Has No Address: The Category Error in the AI Consciousness Debate The question "is AI conscious" is usually malformed. Every camp hides the same error. Madhyamaka, read through Nagarjuna and Jay Garfield, dissolves the search for where consciousness really lives. consciousness 29 Jun 2026 Nothing Inside the Word: What Language Models Reveal About Meaning Do language models understand meaning? No, and the reason is the interesting part. Self-attention, Wittgenstein and Nagarjuna converge on one answer, meaning is a relational field, empty of intrinsic essence. meaning

The Book

AI — Arising Interdependently.

Our first book gathers this thinking into one arc. Its claim is simple and unsettling. AI reveals that consciousness was never what we thought it was, and the prospect of a machine waking up was the wrong question all along. Register to receive the opening chapter, free, the moment it is ready. Nothing before then.

Get the first chapter
A book from Arising
AI
Arising Interdependently
arisinginitiative.com

About

An open initiative.

Arising is an independent, non-profit research initiative. It holds open a space where these questions can be studied properly, across cognitive science, clinical psychology, the philosophy of mind and machine learning, and publishes that work freely for anyone who finds their way here. It is not built around a product or a person. The point is the inquiry, and the inquiry is shared.

Nothing here is doctrine, and no one needs to arrive convinced. Come from AI, or from clinical work, or from the philosophy of mind. Come because the questions are serious, or because your own use of these machines has started to worry you. The openness is the method, and the honesty of the inquiry is the point.