Research on machine minds

AI consciousness, explained from the research

Could an AI system be conscious, and how would anyone know? AI Concious explains what scientists and philosophers have actually published on the question: the main theories of consciousness and what each says about machines, the tests researchers use, what chatbots say about themselves, the classic arguments, and the ethics of AI model welfare. Every article is written in plain English, states where the science is still open and is checked against its sources.

One wave of activity sweeps a network and every path meets in the middle. Some theories tie consciousness to a moment like this in the brain; whether a machine could have one is the open question.
A glowing cloud of light reflected differently in a puddle

Guide

Is ChatGPT Conscious? What Chatbot Self-Reports Show

Is ChatGPT conscious? Researchers think today's chatbots very probably are not, and OpenAI says the question cannot yet be settled scientifically. Studies of what chatbots say about themselves point in both directions.

Latest research reads

Glowing filaments passing through the door slot of a wooden room in a field

Guide

The Chinese Room Argument: What It Says About AI

John Searle's Chinese room argument says that following rules for symbols is not the same as understanding them. Here is the thought experiment, the main replies, and what it means for today's chatbots.

A balance scale holding a teal glow and a gold glow, perfectly level

Guide

The Moral Status of AI: Arguments For and Against

Could an AI system ever matter morally in its own right? Philosophers have argued both ways. Here are the main arguments for and against the moral status of AI, and the proposals for acting while the question is open.

A tangled knot of glowing tendrils beside a line of separate motes

Guide

Integrated Information Theory and AI: What Phi Means

Integrated Information Theory says consciousness is a property of physical cause and effect, measured as phi. On that view a computer could copy everything a brain does and still feel nothing.

Explore the five questions

A glowing translucent sphere floating above dark water

Topic

The Big Question

The big question first: could an AI system be conscious, and what do the words mean? These guides set out where the research stands today, and why careful resea

Five different glowing organic forms in an arc

Topic

Theories of Consciousness

Scientists disagree about what consciousness is, and each theory gives a different answer about machines. These guides explain the leading theories, from global

A row of vials holding glowing seeds, some lit and some dark

Topic

Tests and Evidence

How would anyone tell whether an AI system is conscious? These guides cover the indicator checklist researchers use, experiments on introspection and what studi

Glowing filaments of light growing through a library doorway

Topic

Philosophy of Mind

Some of the hardest questions about AI minds are old ones. These guides apply the classic arguments of the philosophy of mind, such as the Chinese room and the

A glowing bud of light on a mossy stone in a dark forest

Topic

Ethics and Model Welfare

If AI systems might one day have experiences, how should they be treated? These guides explain the moral status debate and what researchers and AI labs have beg

Theories of consciousness

A tangled knot of glowing tendrils beside a line of separate motes

Guide

Integrated Information Theory and AI: What Phi Means

Integrated Information Theory says consciousness is a property of physical cause and effect, measured as phi. On that view a computer could copy everything a brain does and still feel nothing.

Glowing pods in a ring, linked by threads of light to a central pod

Guide

Global Workspace Theory and AI: Could a Machine Qualify?

Global workspace theory says consciousness is information broadcast across a whole system. That makes it one of the theories most open to machines, and the reason some researchers think AI agents could qualify.

Tests and evidence

A glowing cloud of light reflected differently in a puddle

Guide

Is ChatGPT Conscious? What Chatbot Self-Reports Show

Is ChatGPT conscious? Researchers think today's chatbots very probably are not, and OpenAI says the question cannot yet be settled scientifically. Studies of what chatbots say about themselves point in both directions.

Philosophy and ethics

Guide

The Chinese Room Argument: What It Says About AI

John Searle's Chinese room argument says that following rules for symbols is not the same as understanding them. Here is the thought experiment, the main replies, and what it means for today's chatbots.

Guide

The Moral Status of AI: Arguments For and Against

Could an AI system ever matter morally in its own right? Philosophers have argued both ways. Here are the main arguments for and against the moral status of AI, and the proposals for acting while the question is open.

Guide

The Hard Problem of Consciousness and AI

The hard problem asks why any physical process is accompanied by experience at all. Explaining what an AI system can do may never answer it, and that is why the AI consciousness debate is so hard to close.

Guide

AI Model Welfare: What It Means and Who Studies It

AI model welfare is the study of whether AI systems could have interests that matter morally, and what to do while nobody knows. Here is what researchers propose and what AI labs have done so far.

How these articles are researched

Each article starts from one question and is built from published work: research papers and preprints by the scientists who did the research, encyclopaedias of philosophy and the AI labs' own publications. Every fact is recorded with the page it came from and, where the source gives one, its date, and each of those pages is checked again before publication. Where researchers disagree, each position is set out in its own terms.