SCIENCE UNDER AUTOMATION
Scientific research tasks and AI capabilities
Task pages are sections of their field assessments so that evidence remains in its scientific context.
Computational genomics
- Run a specified analysis
Closest match to the seeded benchmark: objectives and methodological guidance are supplied.
- Design a new biological study
Question quality, confounding and validation require separate evidence.
- Validate a new mechanism
Often crosses from computation into experiments; the field-level physical rating no longer applies.
Theoretical physics
- Solve a specified research problem
CritPt is the closer mapping. FrontierScience pools multiple scientific disciplines.
- Develop a new theoretical description
Novelty and explanatory value need assessment beyond a correct final answer.
- Connect a theory to observations
The timing and quality of observations can dominate this stage.
Experimental molecular biology
- Interpret literature and plan a protocol
Information tasks have benchmark coverage; physical success requires a different test.
- Execute and troubleshoot an experiment
Requires instrument access and evidence from the actual experimental setting.
- Establish a causal mechanism
A convincing account depends on discriminating tests and independent validation.
Field ecology
- Analyse an existing ecological dataset
Relevant agent benchmarks still need to be reviewed and mapped.
- Collect representative field observations
Remote sensing or robotics may change particular workflows; they need their own evidence.
- Test a long-term ecological explanation
Fast computation does not shorten the underlying biological process.
Structural biology
- Predict a protein structure
Strong transformation of a central digital task; predictions still need confidence assessment and experimental interpretation.
Weather forecasting
- Generate a medium-range forecast
Strong benchmark performance across many forecast targets; this score covers forecasting, not all atmospheric science.
Algorithm design
- Optimize a specified algorithm
Real system improvements with automated verification; humans still define the objectives and evaluation functions.
Mathematics
- Solve a formalizable problem
Strong results on selected formalizable problems; olympiad success does not establish autonomous frontier mathematics.
Materials science
- Execute a configured synthesis workflow
Execution has been demonstrated in a bounded physical setup; novelty, characterization and generalization need separate assessment.
Drug discovery
- Prioritize a drug candidate
Useful candidate discovery; laboratory validation and clinical development remain substantial parts of the work.
Astronomy
- Vet candidates in an existing sky survey
Survey instruments and curated data are supplied; the result does not cover telescope operation or new physical explanations.
- Investigate a new astrophysical phenomenon
This broader workflow has not been evaluated by the cited demonstration.
Synthetic chemistry
- Plan and execute a configured synthesis
Prepared equipment, available reagents and constrained experiments do not establish general laboratory autonomy.
- Invent and validate a new reaction
This broader workflow has not been evaluated by the cited demonstration.
Neuroscience
- Reconstruct neural circuits from existing images
Human tissue preparation, imaging, proofreading and scientific analysis remain part of the project.
- Explain how a circuit causes behaviour
This broader workflow has not been evaluated by the cited demonstration.
Seismology
- Build a catalogue from existing sensor records
Arrival picking uses existing sensors and labelled training data; it does not establish earthquake prediction.
- Infer and test a new fault model
This broader workflow has not been evaluated by the cited demonstration.
Computational pathology
- Analyse a digitized tissue slide
Prepared slides and task labels are provided; specimen processing and clinical sign-off are outside the model evaluation.
- Establish a new disease mechanism
This broader workflow has not been evaluated by the cited demonstration.
Radiology
- Read a mammography screening examination
This is evidence for assisted screening. It does not measure autonomous diagnosis, other imaging modalities or independent patient management.
- Resolve an unfamiliar diagnostic problem
This broader workflow has not been evaluated by the cited demonstration.
Computer science
- Implement and test a defined software task
See the requirement-level assessment and its sources.
- Develop and validate a new computational research result
See the requirement-level assessment and its sources.
Experimental physics
- Operate an instrument and analyse a defined measurement
See the requirement-level assessment and its sources.
- Develop and verify a new physical measurement
See the requirement-level assessment and its sources.
Chemical sciences
- Execute and evaluate a configured chemical experiment
See the requirement-level assessment and its sources.
- Establish a new reaction or chemical explanation
See the requirement-level assessment and its sources.
Biological sciences
- Optimize a biological assay in a prepared laboratory
See the requirement-level assessment and its sources.
- Discover and independently test a biological mechanism
See the requirement-level assessment and its sources.
Clinical medicine
- Assess a patient and recommend management
See the requirement-level assessment and its sources.
- Deliver treatment and manage an evolving clinical course
See the requirement-level assessment and its sources.
Population health
- Build and evaluate a disease forecast from surveillance data
See the requirement-level assessment and its sources.
- Design, deliver and evaluate a population intervention
See the requirement-level assessment and its sources.
Earth & environmental science
- Analyse observations and produce an environmental prediction
See the requirement-level assessment and its sources.
- Test a new process with field observations
See the requirement-level assessment and its sources.
Engineering
- Design and simulate an artifact against a specified brief
See the requirement-level assessment and its sources.
- Commission and verify a new physical system
See the requirement-level assessment and its sources.
Social & behavioural science
- Reproduce and examine an existing quantitative analysis
See the requirement-level assessment and its sources.
- Establish a new result about people or institutions
See the requirement-level assessment and its sources.
Historical & archaeological research
- Find and contextualize candidates in an existing archive
See the requirement-level assessment and its sources.
- Establish and verify a new historical interpretation
See the requirement-level assessment and its sources.