The Freedom of ThoughtResearch Initiative
Research programme of The Freedom of Thought

AI & human intellect.

A research programme of The Freedom of Thought investigating artificial intelligence, human intellect, and the conditions of intellectual freedom.

Through the Lens of Freedom of Thought

Ongoing research

Artificial intelligence is becoming capable of performing increasingly complex forms of reasoning, synthesis, language, memory, problem-solving, and decision-making. This raises a question deeper than whether AI will make humans more productive: what happens to the conditions under which human beings can think freely when intelligence is no longer uniquely human?

Explore the researchResearch questions
§ 01Overview

A new condition for an old question.

Human beings have historically occupied a unique position as the most capable known form of general intelligence available to them. Artificial intelligence introduces the possibility of a qualitatively new intellectual condition, in which non-human systems can perform increasingly sophisticated cognitive functions.

This programme is not a technology section of the initiative. It is the same philosophical question — what it means to live an intellectually free life — pursued under a new condition.

Research philosophy

This programme does not begin with the assumption that AI will replace human intelligence, nor with the assumption that AI will necessarily enhance it. It investigates competing possibilities through philosophical analysis, technical research, empirical experimentation, and critical reflection.

Research problem

The programme's research problem is not the capability of machines but the condition of human thought. As cognitive operations migrate into artificial systems, the environment in which intellectual autonomy is formed, exercised and lost is itself changing. That changing environment — not the technology in isolation — is the object of research.

Core research question
“How does the emergence of artificial intelligence transform the conditions under which human beings can think freely?”
Secondary question

Can human beings remain intellectually autonomous while increasingly relying on systems whose cognitive capabilities may exceed their own in important domains?

Research hypothesis — not an established finding

Could advanced artificial intelligence eventually create a functional cognitive asymmetry between humans and machines comparable, in significance rather than biology, to the asymmetry humans experience between themselves and other animals?

§ 02The Freedom of Thought lens

Why artificial intelligence is a freedom of thought problem.

The methodological bridge between the initiative's general framework and this particular field of research.

Artificial intelligence is studied here not as a technology sector but as a new condition of human thought. Systems that draft, summarise, retrieve, argue and decide do not sit outside intellectual life; they increasingly constitute the environment within which it takes place.

The moment a cognitive operation is delegated, the question of who is thinking — and who is answerable for the conclusion — becomes a live one. That question is the classical problem of intellectual autonomy, arriving in a new form.

The programme therefore treats AI as one field through which the problem of freedom of thought is investigated, alongside the initiative's other programmes.

What the lens makes the programme attend to
  • What human beings still do for themselves
  • What is delegated, and under what conditions
  • Which knowledge becomes cognitively available
  • Who holds epistemic authority
  • How language and concepts are formed
  • Where intellectual responsibility rests
Conceptual relationship
  1. Artificial intelligence
  2. Condition of human thought
  3. Cognitive delegation & intellectual agency
  4. Intellectual autonomy
  5. Freedom of thought
Framework, not conclusion

The lens is the programme's analytical framework, not its conclusion. Nothing on this page reports a finding.

§ 03Competing hypotheses

Four possibilities, none assumed.

These are research hypotheses, not predetermined conclusions. The programme is designed to test them against one another.

01

Augmentation

AI may expand human intellectual capacity by extending memory, reasoning, creativity, synthesis, and problem-solving.

02

Substitution

AI may progressively assume cognitive functions that humans traditionally perform themselves.

03

Transformation

AI may not simply increase or decrease human intelligence, but transform what human intelligence itself means and consists of.

04

Cognitive asymmetry

Advanced AI may eventually create a functional asymmetry in which humans remain intellectually autonomous but are no longer the most cognitively capable agents within their intellectual environment.

§ 04Conceptual model

From tool to intellectual agent.

  1. 01AI as tool
  2. 02AI as assistant
  3. 03AI as cognitive partner
  4. 04AI as cognitive substitute
  5. 05AI as autonomous intellectual agent

The central research problem is not simply whether AI becomes more capable. It is whether the human relationship with intelligence changes from using an external tool to participating in an increasingly distributed intellectual system.

Each position on this progression describes a different relationship between the thinker and the system — and therefore a different set of conditions for intellectual freedom. The programme treats the progression as an analytical model to be tested, not as an inevitable sequence.

§ 05Core concept

Cognitive delegation.

Cognitive delegation occurs when a human transfers part of a cognitive task to an artificial system.

  1. 01

    Assistance

  2. 02

    Augmentation

  3. 03

    Delegation

  4. 04

    Substitution

  5. 05

    Intellectual displacement

The research investigates where and how the transition along this continuum occurs: which tasks move, under what conditions, and with what consequences for the person doing — or no longer doing — the thinking.

Delegation is not presumed to be harmful. The programme investigates cognitive enhancement and cognitive dependency as equally serious possibilities, including the possibility that both occur simultaneously in different domains.

Under investigation

No claim is made here about the point at which delegation becomes displacement. Locating that threshold — if it exists as a threshold at all — is one of the programme's research objectives.

§ 06Function map

What happens to human intellectual functions?

For each function: human capacity, AI capability, and the research question the programme takes up. Functions marked as unresolved involve philosophical questions this programme does not claim to have settled.

Memory

Human

Biologically limited and reconstructive.

AI

Potentially massive external memory and retrieval capability.

Research question

What happens to human intellectual development when memory becomes increasingly externalized?

Calculation

Human

Slow, effortful, and error-prone beyond small magnitudes.

AI

Reliable and effectively unbounded within formal domains.

Research question

When a capacity is fully delegated, does the underlying human competence matter intellectually?

Language

Human

Acquired socially; bound to embodiment, culture, and intention.

AI

Fluent statistical generation across registers and languages.

Research question

Does fluent machine language production change how humans form and test their own expression?

Reasoning

Human

Deliberate, revisable, and accountable to reasons.

AI

Increasingly capable within structured tasks; reliability contested.

Research question

Which forms of reasoning can be delegated without loss of intellectual responsibility?

Synthesis

Human

Slow integration across sources, disciplines, and time.

AI

Rapid aggregation and summarisation at scale.

Research question

Does rapid synthesis deepen understanding, or substitute for the labour that produces it?

Creativity

Human

Situated in experience, purpose, and risk.

AI

Recombination and generation across large learned distributions.

Research question

What distinguishes generation from origination?

Judgment

Human

Formed over time through practice, error, and correction.

AI

Contested; systems can rank options without bearing consequences.

Research question

Can judgment be delegated at all, or only its inputs?

Concept formation

Unresolved philosophical question
Human

Developed through use, dialogue, and revision.

AI

Internal representations exist; their conceptual status is disputed.

Research question

Does using machine-produced concepts alter how humans form their own?

Critical thinking

Human

Requires effort, scepticism, and willingness to be wrong.

AI

Can produce critique on request; independence of critique unclear.

Research question

Does frequent access to machine critique strengthen or replace the disposition to criticise?

Decision-making

Human

Bound to values, stakes, and accountability.

AI

Optimisation over specified objectives.

Research question

Where should the boundary of automated decision support lie in intellectual work?

Wisdom

Unresolved philosophical question
Human

Traditionally understood as the integration of knowledge, character, and life.

AI

No established basis for attribution.

Research question

Is wisdom in principle non-delegable? An open philosophical question.

Self-reflection

Unresolved philosophical question
Human

Thought turning upon its own conditions.

AI

Self-report exists; its status as reflection is unresolved.

Research question

Can reflection be assisted without being outsourced?

Consciousness

Unresolved philosophical question
Human

Presupposed in ordinary intellectual life; philosophically contested.

AI

No established evidence; claims remain speculative.

Research question

This programme treats machine consciousness as an unresolved philosophical question, not a premise.

Responsibility

Unresolved philosophical question
Human

Attaches to persons who can answer for their conclusions.

AI

No established basis for attribution of responsibility.

Research question

Who is intellectually responsible for a conclusion reached through a distributed system?

§ 07Research areas

Nine areas, one problem.

The programme's subject matter organised through the freedom of thought framework. Each area names why it matters to freedom of thought, its core question, and the research streams where the work is carried out.

04

AI and epistemic authority

Why it matters to freedom of thought

What counts as authoritative determines which interpretations are available for thinking at all.

Core question

Who or what determines what becomes cognitively available, and on what basis is machine output trusted?

06

AI and human intelligence

Why it matters to freedom of thought

The comparison being made requires clarity about what human intelligence and intellect actually are.

Core question

What constitutes human intelligence, and what distinguishes intelligence from intellect?

07

AI and intellectual dependence

Why it matters to freedom of thought

Dependence is not assumed harmful; it is the empirical question at the centre of the programme.

Core question

Does sustained AI assistance strengthen independent intellectual capacity, leave it unchanged, or weaken it?

08

AI and human intellectual becoming

Why it matters to freedom of thought

Freedom of thought here is understood as the capacity of thought to reflect upon and transform its own conditions.

Core question

Does AI accelerate intellectual becoming, or increasingly perform it on the human's behalf?

09

AI and the future of intellectual freedom

Why it matters to freedom of thought

The synthesis: what intellectual freedom means when the thinker may no longer be the most capable intelligence present.

Core question

Which conditions of intellectual freedom are newly at stake, and which are unchanged?

§ 08Research streams

Six streams of inquiry.

§ 09Research questions

Questions the programme takes up.

These are research questions, not established conclusions.

Core research question

How does the emergence of artificial intelligence transform the conditions under which human beings can think freely?

Secondary questions
  1. 01

    What happens to intellectual autonomy when cognitive tasks are increasingly delegated to AI?

  2. 02

    Does dependence on AI strengthen or weaken independent intellectual judgment?

  3. 03

    How does AI alter the production and circulation of knowledge?

  4. 04

    Who or what determines what becomes cognitively available to human beings?

  5. 05

    How does AI-mediated language affect conceptual formation?

  6. 06

    Can human beings remain intellectually autonomous while increasingly relying on machine intelligence?

  7. 07

    Does AI expand the possibility of thought or narrow it?

  8. 08

    What happens to human intellectual becoming in an environment increasingly mediated by AI?

Further questions across the streams
  1. 01

    What constitutes human intelligence?

  2. 02

    What distinguishes intelligence from intellect?

  3. 03

    Which human intellectual functions can AI reproduce?

  4. 04

    Which functions can AI augment?

  5. 05

    Which functions can AI substitute?

  6. 06

    What happens when humans systematically delegate cognitive work?

  7. 07

    Does AI strengthen or weaken critical thinking?

  8. 08

    Does AI alter human intellectual autonomy?

  9. 09

    Does AI change the process of intellectual becoming?

  10. 10

    Can humans remain intellectually autonomous while depending heavily on AI?

  11. 11

    What happens when AI becomes cognitively superior to humans in increasingly general domains?

  12. 12

    What does freedom of thought mean when the thinker is no longer necessarily the most capable intelligence in the intellectual environment?

§ 10Philosophical connection

AI and intellectual becoming.

The Freedom of Thought project understands intellectual freedom not simply as freedom from external constraint, but as the capacity of thought to reflect upon and transform its own conditions.

Artificial intelligence may become not merely a tool used by thought, but part of the conditions through which human thought occurs. AI may therefore become a condition of human intellectual becoming.

Research question

Does AI accelerate human intellectual becoming, or does it increasingly perform intellectual becoming on behalf of the human?

The becoming sequence
  1. Experience
  2. Reflection
  3. Dialogue
  4. AI-mediated cognition
  5. Judgment
  6. Transformation
§ 11Emerging concept

Distributed intellect.

Human intellectual activity may increasingly become distributed across biological cognition, artificial intelligence, external information, tools, and networks.

If cognition becomes distributed, does intellectual autonomy also become distributed?

Who is thinking?

Research concept under investigation

Distributed Intellect is a concept the programme is developing and testing. It is not presented as an established theory, and it may be revised or abandoned in light of what the research finds.

§ 12Proposed empirical work

The AI withdrawal experiment.

A proposed experimental programme. No findings are reported here, and no study has yet been conducted.

Phase I

Human alone

Baseline intellectual performance without AI assistance.

Phase II

Human + AI

Sustained work with AI assistance across the same task families.

Phase III

AI removed

Return to unassisted work to observe what independent capacity remains.

Proposed measures
  • Reasoning
  • Memory
  • Critical evaluation
  • Writing
  • Conceptual understanding
  • Creativity
  • Judgment
  • Confidence
  • Independent performance
Research question

Does sustained AI assistance strengthen independent intellectual capacity, leave it unchanged, or weaken it?

Instrument in development

HAICB — Human–AI Intellectual Capability Battery

A proposed assessment instrument to be developed as part of the programme. The dimensions below are candidates, not a validated instrument.

  • Reasoning
  • Memory
  • Critical evaluation
  • Concept formation
  • Creativity
  • Argumentation
  • Judgment
  • Epistemic humility
  • Independent problem-solving
§ 13Publications

Research in development.

Research publications will appear here as the programme develops. The titles below are planned works; no dates, venues, authors, or findings are claimed.

Flagship prospective paper

When Intelligence Is No Longer Uniquely Human: Freedom of Thought in the Age of Artificial Intelligence

  1. 01

    What Is Human Intelligence? A Philosophical Reconsideration in the Age of Artificial Intelligence

    Research in development
  2. 02

    From Cognitive Assistance to Intellectual Substitution: A Theory of Human–AI Cognitive Delegation

    Research in development
  3. 03

    Does Artificial Intelligence Augment or Displace Human Intellectual Agency?

    Research in development
  4. 04

    Cognitive Offloading and the Transformation of Intellectual Autonomy

    Research in development
  5. 05

    Human–AI Distributed Intelligence: Rethinking the Location of the Intellect

    Research in development
  6. 06

    Intellectual Becoming in the Age of Artificial Intelligence

    Research in development
  7. 07

    Freedom of Thought After Human Cognitive Supremacy

    Research in development

All initiative publications

§ 14Research leadership & team

Who carries the programme.

The programme follows the initiative's standard research structure: leadership, core research team, advisors, and open roles.

Research leadership

Masud Zakaria

  • Principal Investigator
  • First & Corresponding Author

Holds intellectual direction and accountability for the programme, and leads its philosophical synthesis on freedom of thought.

Core research team

Sadiq M Alam

  • Co-author
  • AI & Human Intelligence Research Lead

Leads the programme's work on artificial intelligence and human intelligence, and is a co-author of the programme's research.

Research advisors

Gautam Barua

  • Intellectual Communication & Language Research Advisor

Advises on intellectual communication and language as conditions of conceptual formation and independent thought.

Sadiq M Alam

  • AI and Human Intelligence Research Advisor

Also serves the programme in an advisory capacity on artificial intelligence and human intelligence.

Open roles within the programme
Appointment in progress

Principal Researcher — Artificial Intelligence and Human Intellectual Development

To lead the technical and empirical investigation of how increasingly capable artificial intelligence affects human cognitive capacities, intellectual agency, cognitive autonomy, and the evolving relationship between human and machine intelligence.

To be announced

Philosophical Research Lead

To lead the conceptual work on intelligence, intellect, intellectual becoming, and freedom of thought, and to hold the programme's philosophical synthesis.

Open

Research Assistants and Collaborators

Doctoral researchers, empirical collaborators, and institutional partners will be invited as the streams open.

Conceptual architecture
  1. Freedom of thought
  2. The human intellect
  3. Intellectual becoming
  4. Artificial intelligence
  5. Cognitive delegation
  6. Intellectual autonomy
  7. The future of human thought

The question is not only what AI will become. It is also what human intellect will become in response.

Explore the research into intelligence, intellectual autonomy, cognitive delegation, and the future of freedom of thought.

Explore researchCollaborateContact the research team