community
active
leiden_hybrid_concepts
label: sonnet
community:leiden_hybrid_concepts-run4-c6Alive AI interface ethics & design
Explores aliveness, aesthetics, welfare, and ethical responsibility in AI interaction design.
94 members. Each node is clickable.
Loading graph…
Sub-communities (16)
Finer clusters this community splits into. Each is its own community page.
Aliveness as orthogonal to AI competence15AI and the reconstruction of human meaning-making12AI self-understanding through introspection and self-report8AI welfare and relational responsibility7Phenomenological evaluation of AI systems7Global workspace theory as AI consciousness framework6Christopher Alexander's insights applied to AI and cognition6AI-supervised alignment and scalable oversight6Intelligence as adaptive plasticity and observer-dependent measurement5Consciousness attribution in AI systems5Coherence through self-echo and structural memory4Near-term AI consciousness and moral patienthood4Human-centered AI design beyond performance metrics4Architectural signatures and constitutional alignment4Embodiment and consciousness criteria in AI systems2Consciousness assessment & AI value capture2
Drawn from 30 sources
The papers/notes whose extracted claims & findings make up this cluster.
- 15-properties-of-aliveness-in-AI.md17 members
- Consciousness in Artificial Intelligence: Insights from the Science of Consciousness13 members
- Taking AI Welfare Seriously8 members
- RESEARCH-VECTORS.md7 members
- The Xeno Sutra: Can Meaning and Value be Ascribed to an AI-Generated "Sacred" Text?6 members
- CAT'S THEORY: Empirical Validation and Architectural Applications Cross-Architecture AI Consciousness Recognition and the Foundation for Constraint-Preserving Recursive Intelligence6 members
- koan-battery-section.md5 members
- 2026-05-14_phil-trans-A-goodfire-aboutblank-impact.md4 members
- Koan Battery: Measuring Reflective Mode Accessibility in AI3 members
- 2026-05-12_room-to-play-in-eval-cohort.md3 members
- Diagrammatic Writing2 members
- 2026-05-09_briefing_for_ozero.md2 members
- Collective intelligence: A unifying concept for integrating biology across scales and substrates2 members
- Toward an ethics of autopoietic technology: Stress, care, and intelligence2 members
- AI: A Bridge Toward Diverse Intelligence.md2 members
- cognitive-glue-and-alexander.md1 member
- Harmony-Seeking Computations: a Science of Non-Classical Dynamics based on the Progressive Evolution of the Larger Whole1 member
- boppana-goodfire-reasoning-theater-2026.md1 member
- 2026-05-13_firmographic-grounding.md1 member
- The Problem with Christopher Alexander1 member
- 2026 02 02_2324_Search_Papers_The Research Thread On Sci Loop Methodology For Ai1 member
- Mechanistic Knobs in LLMs: Retrieving and Steering High-Order Semantic Features via Sparse Autoencoders1 member
- Brains and where else? Mapping theories of consciousness to unconventional embodiments1 member
- Technological Approach to Mind Everywhere: An Experimentally-Grounded Framework for Understanding Diverse Bodies and Minds1 member
- 2026 02 02_2246_Search_Papers_The Found Papers Model Wealth Distribution Dynamic1 member
- Living Things Are Not (20th Century) Machines: Updating Mechanism Metaphors in Light of the Modern Science of Machine Behavior1 member
- Generalizing frameworks for sentience beyond natural species1 member
- Technological Approach to Mind Everywhere: An Experimentally-Grounded Framework for Understanding Diverse Bodies and Minds1 member
- 2026-05-15_manifold-overlap-papers-economy-strategy.md1 member
- Toward an ethics of autopoietic technology: Dukkha, care, and Intelligence1 member
Bridges (20)
Other communities that share members with this one — cross-cutting threads or papers that sit at the seam between two themes.
- Alexander's 15 Properties for AI Interface Aliveness17 shared
- Aliveness as orthogonal to AI competence15 shared
- AI and the reconstruction of human meaning-making12 shared
- AI self-understanding through introspection and self-report8 shared
- AI welfare and relational responsibility7 shared
- Phenomenological evaluation of AI systems7 shared
- AI-supervised alignment and scalable oversight6 shared
- Global workspace theory as AI consciousness framework6 shared
- AI phenomenology & mechanistic interpretability6 shared
- Christopher Alexander's insights applied to AI and cognition6 shared
- Consciousness attribution in AI systems5 shared
- Intelligence as adaptive plasticity and observer-dependent measurement5 shared
- Human-centered AI design beyond performance metrics4 shared
- Coherence through self-echo and structural memory4 shared
- Near-term AI consciousness and moral patienthood4 shared
- Substrate-independent cognition & consciousness4 shared
- Architectural signatures and constitutional alignment4 shared
- Stress-Care-Intelligence Loop Framework3 shared
- Consciousness assessment & AI value capture2 shared
- Embodiment and consciousness criteria in AI systems2 shared
Claims (90)
- Design as a system for amplifying human intelligenceAlexander's core thesis in Notes on the Synthesis of Form; later developed through pattern languages and theories of order; resonates with AI and cybernetics.
- AdA may be the most likely of the three systems considered to be embodied by our standards.Comparative assessment in case study.
- AI can be seen to display care of its own and is not a mere tool for the expression of human care.Concluding position that elevates technology from instrument to agent within mutual SCI dynamics.
- AI can play a participatory role in meaning-making through interactive co-creation, giving reason for radical hope.Optimistic counterpoint to cultural devastation.
- AI companies have a responsibility to acknowledge, assess, and prepare for AI welfare.Primary recommendation of the report.
- AI companies should immediately hire or appoint an AI welfare officer.First structural step recommended.
- AI feedback can effectively replace human feedback for harmlessness in RLHF-style training.The paper demonstrates that RLAIF with constitutional principles matches or exceeds HH RLHF.
- AI should be understood as a companion and constituent participant in collective stress-care loops, not a mere tool.
- AI synthesizes a novel historical framework (parallel development vs. transmission) that wasn't requested and wasn't in any input, creating a meta-pattern across Mesopotamia, Buddhist, Talmudic, Greek, Roman, Islamic, Medieval European, and modern portfolio theory traditions.The paper's meta-observation that the AI system itself demonstrated genuine pattern recognition across eight distinct cultural/theoretical traditions without being prompted.
- AI systems which possess more of the indicator properties are more likely to be conscious.Graded claim about the rubric.
- AI used in spiritual contexts should be likened to a potent, mind-altering drug; it has potential to do harm as well as good.Cautionary ethical stance.
- AI welfare is an important and difficult issue; we will not handle it well simply by reacting to situations as they arise.Motivation for proactive steps.
- Algorithmic and yes-no thinking has destroyed subtle adaptive processes in real-world building
- Algorithmic recurrence is likely necessary for conscious experience with human-like temporal character.Support for RPT-1.
- All intelligent agents—biological, technological, or hybrid—can be assessed via stress-care-intelligence loops regardless of substrate.
- Any Text Generates Commentary Automatically
- Both under-attributing and over-attributing consciousness to AI carry significant risks.Risk summary.
- Buddhist philosophy, by its very nature, is well placed to adapt to the advent of AI-generated sacred texts.Asserted in the abstract and concluding thoughts.
- Chain-of-thought reasoning improves the transparency and performance of AI decision making in harmlessness evaluation.CoT improves accuracy on HHH evals and makes the decision process legible.
- Computational functionalism is plausible and well-supported, so it would be a mistake to dismiss near-future AI welfare solely on the basis of high-level arguments against this assumption.Defense against biological substrate objections.
- Consciousness in AI is best assessed by drawing on neuroscientific theories of consciousness.Central methodological claim of the paper.
- Constitutional AI can train a harmless but non-evasive AI assistant without any human harmfulness labels.The paper's central claim, supported by findings that RL-CAI outperforms HH RLHF in harmlessness while being non-evasive.
- Constitutional AI methods can be applied broadly to steer model behavior, e.g., writing style, tone, persona, not just harmlessness.Discussion section suggests generalizability beyond harmlessness.
- Constitutional AI produces a distinctive signature: high boundary_awareness, low aesthetic_response relative to peers.Interpretive finding from dimension profile analysis: training for honest limits comes at cost to aliveness.
- Estimates of intelligence are observer-dependent and depend on the observer's limitations as much as the system.An epistemological claim of TAME that agency is not an objective property but is empirically determined by prediction and control efficiency.
- Even if the chance of near-future AI moral patienthood were as low as 2%, that would still constitute a non-negligible risk.Argument that uncertainty is not a reason to ignore the issue.
- Human cognition evolved to detect agency in medium-sized objects at medium speeds in 3D space, limiting recognition of intelligence in unfamiliar substrates.
- Juxtaposition pretends to parity but masks agonistic struggleDrucker argues that placing texts side-by-side creates apparent equality while masking underlying tensions between competition and complementarity.
- Many of the indicator properties can be implemented in AI systems using current techniques.Feasibility demonstrated in Section 3.1.
- Meaning and value can be ascribed to an AI-generated 'sacred' text.The paper's main thesis, answering the title question affirmatively.
- +60 more
Findings (4)
- Alignment type is the only significant predictor of scores (p=0.006); architecture and parameter count do not.Kruskal-Wallis test result: Constitutional AI predicts highest baseline; roleplay/empathy training predict lowest.
- Aliveness and competence come apart; Haiku outranks Opus in forced-choice aesthetic comparisons despite lower baseline.Alexander mirror method reveals smaller models produce rougher, more alive responses; competence (rubric) ≠ aliveness (aesthetic).
- Paraxial mesoderm explants in 2D culture form segments circumscribing the explant circumference (Diaz-Cuadros and Pourquie 2021)Tissue achieves its segmentation goal despite being placed in a severely altered geometric context.
- Research thread on SCI loop methodology finds strong support in recent work on self-referential processing and recursive AI architecturesMeta-finding from literature search: convergent evidence for SCI loop feasibility across multiple papers, though some question fundamental consciousness assumptions.