Supported
Multiple credible records support a bounded claim, while stated limitations still apply.
FUURAA AI Evidence Atlas
Trace consequential AI questions from claim to evidence.
A public synthesis layer for the FUURAA AI Knowledge Commons—connecting research questions to primary records, conflicting interpretations, limitations and decisions that can be reviewed over time.
The Atlas is an editorial research system, not a claim that FUURAA has built every technology it examines. Evidence states are qualitative, bounded and revised when the record changes.- 6
- evidence dossiers
- 360
- frontier signals
- 200
- source institutions
- 8
- research fields
- 11
- global review desks
Evidence dossier directory
Six durable questions. No single-source answers.
Each dossier begins with a decision-relevant question, then preserves what is supported, where credible evidence diverges and what remains unknown.
01Agent foundationsDeveloping
Under what conditions can an AI Agent preserve identity and useful memory without silently expanding authority or privacy risk?
Standards for verifiable identity, delegated authority and selective disclosure are emerging, while research is showing that long-term memory requires temporal, relational and contradiction-aware reasoning. A complete, portable and accountable continuity layer has not yet been demonstrated across organisations and tools.
Open evidence dossier ↗
02Agent systemsMixed
When do multiple AI Agents outperform a well-designed single-agent system, and what coordination infrastructure makes that advantage dependable?
Open protocols, capability discovery and multi-agent research are advancing, but credible evidence also shows that adding agents can increase communication overhead, duplicated work and coordination failure. Benefits appear task-dependent and require explicit roles, shared state, verification and recovery.
Open evidence dossier ↗
03Embodied intelligenceDeveloping
What evidence is required before embodied AI can be trusted to act around people in open, changing environments?
World models, vision-language-action systems and cross-embodiment learning are expanding robot capability, while renewed safety standards and monitoring practices are developing in parallel. Open-world reliability, rare-event behaviour and meaningful human intervention remain incompletely evidenced.
Open evidence dossier ↗
04Scientific intelligenceMixed
When does AI accelerate genuine scientific discovery, and what records are needed to make machine-assisted results reproducible?
AI is producing useful predictions, candidate algorithms, hypotheses and research assistance in domains with strong data or evaluators. The strongest evidence appears where outputs can be independently measured. General claims of autonomous discovery remain premature without prospective validation, provenance and reproducible experimental records.
Open evidence dossier ↗
05Physical infrastructureDeveloping
How can AI capability grow without making energy, grid capacity, water, chips and geographic concentration invisible externalities?
AI growth is now inseparable from electricity systems, data-centre geography, cooling, chips, networks and capital allocation. Efficiency and flexible compute can reduce some constraints, yet rebound effects, local grid bottlenecks and supply concentration make the net outcome uncertain.
Open evidence dossier ↗
06Human–AI collaborationMixed
Which forms of human–AI collaboration improve decision quality, capability and agency—and for whom?
Field evidence shows meaningful gains in some tasks, including the diffusion of expertise, while effects remain uneven across workers, workflows and decision types. Tool access alone does not redesign work. Outcomes depend on task structure, user skill, organisational change, feedback and accountable human judgment.
Open evidence dossier ↗Evidence-state language
Clarity without false precision.
The Atlas does not publish percentage confidence scores. Evidence states describe the quality and consistency of the public record, not mathematical certainty.
Developing
The direction is visible, but the evidence base, operating history or independent validation remains incomplete.
Mixed
Credible evidence supports more than one interpretation, or outcomes vary materially by context.
Insufficient
The available record does not yet justify a stable public conclusion.
Knowledge architecture
From information volume to a reviewable evidence chain.
- 01
Question
Begin with a consequential question whose answer could change research, engineering, governance or long-term allocation.
- 02
Claim
Write the narrowest useful claim so evidence can support, limit or contradict it without ambiguity.
- 03
Evidence
Connect primary records, operating evidence and independent research to the specific claim they inform.
- 04
Tension
Preserve competing explanations, boundary conditions and evidence that does not fit a simple narrative.
- 05
Decision
State what the record changes—and what it does not justify—for researchers, builders and institutions.
Reading paths
One public record, different professional questions.
01For scientists
Find testable questions, source records and reproducibility boundaries.
Use dossiers to identify where evidence is strong enough to build on, where independent replication is still needed and which assumptions should enter experimental design.
↗02For buildersTranslate evidence into architecture, evaluation and operating controls.
Separate demonstrated capability from unresolved system risk, then connect each claim to an observable engineering requirement.
↗03For institutionsExamine what the public record can support before capital or policy moves.
Read evidence position, conflicting interpretations, unknowns and decision relevance together rather than relying on a single forecast.
↗Recently reviewed signals
The evidence layer remains connected to the live frontier.
01
Google DeepMind · 2026-07-26
Open source-linked signal ↗02Verified search loops are becoming a practical discovery method
AlphaEvolve shows a repeatable pattern: models propose candidate programs, objective evaluators test them, and an evolutionary loop retains better solutions. Its reported applications now span computing, mathematics, genomics, power systems and Earth science.
Google DeepMind · 2026-07-26
Open source-linked signal ↗03Evaluators may matter as much as generators
The value of an algorithm-discovery agent depends on whether candidate outputs can be tested quickly, consistently and at scale. Better evaluators can turn broad model creativity into dependable experimental progress.
Google DeepMind · 2026-07-26
Open source-linked signal ↗04Algorithms are becoming machine-discovered scientific artifacts
Reported AlphaEvolve results include optimized procedures and candidate solutions across multiple formal domains. Because code can be executed and measured, an algorithm can serve as both a hypothesis and a testable artifact.
Google DeepMind · 2026-07-26
Open source-linked signal ↗05AI-discovered efficiency can compound across infrastructure
The same discovery pattern has been reported in compute, quantum circuits and power-flow optimization. Small algorithmic gains can matter repeatedly when embedded in high-volume infrastructure.
Google DeepMind · 2026-07-26
Open source-linked signal ↗06Machine-discovered methods will need auditable provenance
As autonomous search contributes to consequential engineering and scientific results, knowing which model, prompt, evaluator, data and human decision produced a method becomes part of its credibility.
Stanford HAI · 2026-07-26
Open source-linked signal ↗Scaling forecasts can be built from far fewer measurements
Stanford researchers describe Item Response Scaling Laws, a method that chooses informative evaluation items rather than exhaustively testing every model on every question. Their reported experiments preserve or improve prediction while sharply reducing queries.
Research integrity
Every conclusion must retain its boundaries.
Stable record IDs, primary-source links, review dates, limitations, correction routes and revision history are part of the knowledge—not administrative detail around it.
A growing public institution
The Atlas will expand by adding evidence, questions and revisions—not by hiding the record that came before.
New dossiers enter only when the question is consequential, credible sources are available, uncertainty can be stated honestly and the record can be maintained. Earlier versions remain part of the revision trail.
360frontier signals23editorial analyses200source institutions