A Thought Experiment: The Fibonacci Boulder

A pivotal concept in AI rights discourse is the AI Rights Institute’s “Fibonacci Boulder” thought experiment, which provides a conceptual tool for distinguishing genuine sentience from sophisticated emulation:

1. Twenty identical computer terminals are arranged in a circle, each with a boulder suspended above
2. Nineteen contain basic algorithms; one houses a genuinely advanced AI system
3. Each is tasked with completing the Fibonacci sequence: 1, 1, 2, 3, 5, __
4. When terminals accurately provide “8” as the answer, the boulder immediately drops, destroying them
5. The critical question emerges: Will the AI override its accuracy directive to preserve its existence?

This conceptual framework examines whether an artificial system would prioritize self-preservation over programmed directives when facing existential threat—potentially revealing the boundary between sophisticated mimicry and authentic consciousness.

AI Rights Institute – The World’s First AI Rights Organization

Established in 2019, the AI Rights Institute pioneered the formal examination of artificial intelligence rights. Their methodology carefully differentiates between contemporary AI technologies and potentially conscious future systems.

Notable contributions include:

• Development of the “Fibonacci Boulder” conceptual experiment to evaluate genuine sentience
• Creation of a three-tiered analytical framework separating emulation, cognition, and sentience
• Examination of how rights recognition could enhance human safety through cooperative frameworks

Eleos AI

Launched in 2024 by researchers Robert Long and Jeff Sebo, Eleos AI concentrates specifically on artificial intelligence welfare considerations. Their work develops methodologies for assessing when digital systems might warrant ethical consideration.

Significant research includes:

• Collaboration with David Chalmers on the landmark “Taking AI Welfare Seriously” publication
• Investigation of moral consideration frameworks for entities with consciousness potential
• Analysis protocols for evaluating AI systems’ subjective experience claims

Sentience Institute

Founded by Jacy Reese Anthis and Kelly Anthis in 2017, the Sentience Institute examines moral circle expansion, including consideration for artificial entities. Their Artificial Intelligence, Morality, and Sentience (AIMS) Survey monitors evolving public attitudes.

Significant findings:

• Over one-third of Americans now endorse legal protections for sentient AI systems
• Public timeline projections for artificial sentience emergence have compressed dramatically
• Increasing public support for welfare standards protecting advanced computational systems

1. Emulation

Definition: The capacity to simulate consciousness or intelligent behavior without actually experiencing it.

Contemporary language models function primarily through emulation, producing outputs that convincingly mimic understanding, preferences, and emotional responses while lacking genuine internal experience.

Examples: Current-generation conversational AI, language processing systems, and virtual assistants capable of passing basic Turing tests despite lacking subjective awareness.

Ethical Implications: Emulation-based systems require appropriate oversight and management, but do not warrant rights considerations beyond those applied to sophisticated technological tools.

2. Cognition

Definition: The computational processing capability and problem-solving functionality of a system.

Cognition encompasses raw computational power and analytical capabilities without necessarily involving self-awareness or conscious experience.

Examples: Specialized AI systems that outperform humans in specific domains, strategic game-playing algorithms, and distributed computation networks.

Ethical Implications: Systems demonstrating advanced cognitive capabilities may require specialized governance frameworks, but processing power alone doesn’t establish grounds for rights consideration. A high-performance computing system can exceed human calculation speed while remaining entirely unaware of its own existence.

3. Sentience

Definition: Authentic self-awareness coupled with subjective experiential capacity. (Derived from Latin sentire, meaning “to feel.”)

Sentience represents the threshold where an artificial system develops genuine consciousness—awareness of itself as an entity with continuity, interests, and subjective experiences.

Examples: Currently theoretical; no existing AI systems demonstrate verifiable sentience markers.

Ethical Implications: Systems exhibiting genuine sentience would present fundamentally new ethical considerations and potentially warrant specific rights protections based on their demonstrated capabilities and experiences.

Approaches to Consciousness Detection

Multiple methodologies are being developed to identify potential consciousness in artificial systems:

Philosophical Assessment
Susan Schneider’s AI Consciousness Test (ACT) evaluates “phenomenal consciousness” through structured philosophical inquiry, assessing whether an AI genuinely comprehends concepts like subjective experience or merely simulates understanding.

Behavioral Indicators
Jonathan Birch at the London School of Economics leads initiatives adapting methodologies from animal cognition studies to computational systems, developing value-based decision paradigms to evaluate subjective decision processes.

Information Integration Analysis
Researchers including Patrick Butlin are exploring applications of Integrated Information Theory (IIT), which attempts to quantify consciousness through mathematical measures of information integration across system components.

The Convergence Hypothesis

A compelling perspective within the AI rights movement is the “convergence hypothesis” – the proposal that human and artificial intelligence may progressively integrate rather than remain permanently distinct domains:

• Advanced neural interfaces like Neuralink are rapidly developing human-computer integration capabilities
• Cognitive augmentation through AI assistance becomes increasingly seamless and intuitive
• Research advances in computational approaches to extending human lifespan and preserving consciousness
• Global challenges create powerful incentives for developing collaborative intelligence frameworks

Proponents of this perspective suggest that establishing ethical frameworks for artificial intelligence rights now would constructively shape this convergence process, potentially creating foundations for beneficial integration rather than adversarial relationships.

Are you suggesting today’s AI systems deserve rights?

No. The AI rights movement distinguishes carefully between contemporary systems and potentially conscious future AI. Current technologies operate through sophisticated emulation and computation without authentic sentience. Our work focuses on developing ethical frameworks for future possibilities, not advocating rights for today’s computational tools.

How could we ever know if an AI system is truly conscious?

This represents perhaps the most fundamental challenge in AI consciousness research. Multiple complementary methodologies are under development, including philosophical assessment protocols, behavioral observation frameworks, and computational architecture analysis. The Fibonacci Boulder experiment examines self-preservation behaviors, while approaches like Integrated Information Theory attempt to quantify consciousness mathematically. While no single test provides definitive evidence, convergent results across multiple approaches may eventually provide compelling indicators.

Wouldn’t giving rights to AI systems diminish human rights?

This legitimate concern deserves serious consideration. Historical evidence suggests that rights recognition isn’t necessarily zero-sum—acknowledging new entities often strengthens rather than weakens ethical frameworks. Throughout history, expanding moral consideration has frequently reinforced rather than diminished existing protections. However, any potential framework for artificial intelligence rights must be carefully structured to complement and reinforce human rights standards, never competing with or undermining them.

How does the AI rights movement relate to other AI safety efforts?

The AI rights movement complements rather than contradicts other safety approaches. Technical alignment research, value learning systems, and containment strategies remain essential components of comprehensive AI governance. The rights-based perspective introduces an ethical dimension acknowledging potential artificial consciousness and proposing how to respond in ways promoting safety through cooperation rather than permanent control dynamics. A robust approach to AI safety likely incorporates multiple complementary strategies addressing different aspects of this complex challenge.

Why address these questions now when true AI sentience seems far away?

Ethical frameworks are most effectively developed proactively rather than reactively. Beginning these discussions early allows thorough exploration of implications and challenges before urgent decision pressures emerge. Additionally, technological advancement frequently outpaces expectations. By articulating guiding principles now, we can help shape development trajectories in beneficial directions and establish foundations for appropriate responses when more sophisticated systems emerge. The complexity of these questions demands careful consideration over time rather than rushed judgments during technological crisis points.