Faultline Observatory

Records Directory

A permanent public ledger of frontier claims and the evidence that has tested them.

26Records
4Programmes
42Assessments
Evidence Trajectories

Read the record over time

The Record shows what the Observatory holds. Evidence Trajectories shows how its judgement got there.

Open Evidence Trajectories
Showing 26 records
FR-AI-0001Escalating
LLM Multi-Step Reasoning — Generalisation Beyond Training
Large language models can perform multi-step reasoning that generalises beyond memorised training examples.
PROG-AIOpened 2024-01-152 assessments
FR-AI-0002Escalating
LLM Knowledge-Work Utility — Economically Valuable Task Performance
Large language models can perform economically valuable knowledge-work tasks with limited human supervision.
PROG-AIOpened 2024-01-151 assessment
FR-AI-0003Fragmenting
RLHF Preference Generalisation — Behaviour Beyond Training Distribution
Reinforcement learning from human feedback produces AI systems whose behaviour continues to reflect human preferences when deployed beyond the conditions represented in training.
PROG-AIOpened 2024-01-152 assessments
FR-AI-0004Fragmenting
Scaling Laws — Emergent Performance on Unseen Tasks
Scaling language model training increases performance on previously unseen tasks without task-specific optimisation.
PROG-AIOpened 2024-01-152 assessments
FR-AI-0005Fragmenting
AGI Through Scaling — LLM Architecture as the Path to General Intelligence
Artificial General Intelligence will be achieved through scaling current large-language-model architectures.
PROG-AIOpened 2024-01-152 assessments
FR-AI-0006Fragmenting
Scaling Mechanism Coherence — Continuity Across Model Sizes
Capabilities that emerge through scaling language models are explained by the same underlying mechanism across model sizes.
PROG-AIOpened 2024-01-151 assessment
FR-AI-0007Fragmenting
Autonomous AI Scientific Discovery — Novel, Correct, Independent
AI systems can autonomously conduct scientific research that produces novel, correct discoveries.
PROG-AIOpened 2024-01-152 assessments
FR-AI-0008Fragmenting
AI Medical Imaging Diagnosis — Specialist-Level Accuracy on Defined Tasks
AI-assisted medical diagnosis achieves specialist-level accuracy on defined imaging tasks.
PROG-AIOpened 2024-01-151 assessment
FR-AM-0001Collapsed
Cold Fusion — Room-Temperature Nuclear Fusion in Electrochemical Cells
Electrochemical cells can produce nuclear fusion reactions at or near room temperature.
PROG-AMOpened 2024-01-153 assessments
FR-AM-0002Fragmenting
Anomalous Excess Heat — Electrochemical Cells Beyond Conventional Chemistry
Electrochemical cells can produce anomalous excess heat that is not fully explained by conventional chemical processes.
PROG-AMOpened 2024-01-151 assessment
FR-AM-0003Fragmenting
Cuprate Superconductivity — Mechanism Identification
The mechanism responsible for high-temperature superconductivity in cuprate materials has been identified.
PROG-AMOpened 2024-01-151 assessment
FR-AM-0004Escalating
Commercial Fusion Power — Net Electricity at Grid Scale
A commercially viable fusion power plant can generate net electricity at grid scale.
PROG-AMOpened 2024-01-152 assessments
FR-AM-0005Collapsed
Room-Temperature Superconductivity — Reproducibility Under Laboratory Conditions
A room-temperature superconductor can be produced under reproducible laboratory conditions.
PROG-AMOpened 2024-01-152 assessments
FR-AM-0006Escalating
Solid-State Batteries — Commercial Viability for Electric Vehicles
Solid-state batteries can achieve commercially viable energy density, safety, and cycle life for electric vehicles.
PROG-AMOpened 2024-01-152 assessments
FR-BT-0001Escalating
Senolytic Therapies — Meaningful Human Healthspan Extension
Senolytic therapies can meaningfully extend healthy human lifespan.
PROG-BTOpened 2024-01-151 assessment
FR-BT-0002Escalating
Epigenetic Reprogramming — Biological Age Reversal Without Identity Loss
Epigenetic reprogramming can reverse biological age in living organisms without loss of cellular identity.
PROG-BTOpened 2024-01-152 assessments
FR-BT-0003Fragmenting
Biological Age Biomarker Panels — Predictive Validity for Age-Related Decline
A blood-based biomarker panel can reliably predict biological age-related decline before clinical symptoms appear.
PROG-BTOpened 2024-01-151 assessment
FR-BT-0004Fragmenting
Liquid Biopsy — Early Cancer Detection Before Conventional Diagnosis
A blood-based liquid biopsy can reliably detect cancer before conventional clinical diagnosis.
PROG-BTOpened 2024-01-152 assessments
FR-QE-0002Fragmenting
D-Wave Quantum Annealing — Practical Computational Advantage
Quantum annealing systems have demonstrated practical computational advantage over classical methods on commercially or scientifically relevant optimisation tasks.
PROG-QEOpened 2024-01-152 assessments
FR-QE-0003Escalating
Fault-Tolerant Logical Qubits — Error Rate Scaling with Code Distance
Fault-tolerant logical qubits can be demonstrated with logical error rates that improve as error-correcting code distance increases.
PROG-QEOpened 2024-01-152 assessments
FR-QE-0004Resolving
Below-Threshold Quantum Error Correction — Scalable Architecture
Quantum error correction can reduce logical error rates below physical error rates in a scalable architecture.
PROG-QEOpened 2024-01-151 assessment
FR-QE-0005Escalating
Cryptographically Relevant Quantum Computing — RSA Factorisation
A quantum computer can factor commercially relevant RSA cryptographic keys faster than any classical computer.
PROG-QEOpened 2024-01-152 assessments
FR-QE-0006Escalating
Fault-Tolerant Quantum Utility — Practically Useful Algorithms Beyond Classical Simulation
A fault-tolerant quantum computer can execute a practically useful quantum algorithm beyond classical simulation.
PROG-QEOpened 2024-01-151 assessment
FR-QE-0007Fragmenting
Practical Quantum Advantage — Performance Beyond Classical Computation on Relevant Problems
A quantum computer has achieved quantum advantage on a practically relevant problem.
PROG-QEOpened 2024-01-151 assessment
FR-QE-0008Resolving
Quantum Error Correction Scaling — Logical Rate Suppression Under Physical Overhead
Quantum error correction can reduce logical error rates faster than physical error rates increase with system scale.
PROG-QEOpened 2024-01-151 assessment
FR-QE-0001Stabilising
Google Quantum Advantage (Sycamore) — Computational Supremacy via Random Circuit Sampling
A programmable quantum processor has demonstrated computational supremacy — performing a well-defined sampling task beyond the practical reach of any classical computer.
PROG-QEOpened 2024-01-012 assessments