1 | AI scientist / artificial scientific intelligence¶
Bottleneck¶
Scientific validity, evaluation, tool reliability, verification
Backchain¶
- Foundations I: W1 proof, probability, causal, scientific computing, optimization, metrology
- Foundations II: W2 safety + domain substrate
- Specialization: AI and autonomous science
First proof of work¶
Reproduce one agentic-science result and force every claim to executable evidence
Connected study¶
Editorial study connections grounded in the existing technology bottlenecks and subject descriptions. Background is useful preparation, not a validated prerequisite or a complete technology curriculum.
Statistics + causal inference¶
Background: Distinguish causal evidence from correlations when evaluating an automated scientific claim.
Causal Inference: What If
Open resource · Reading details
Metrology + experiments¶
Background: Design measurements and experiments whose uncertainty and failure modes are explicit.
NIST/SEMATECH e-Handbook of Statistical Methods
Open resource · Reading details
Optimization¶
Background: Understand optimization objectives and constraints when choosing or automating experiments.
Convex Optimization
Open resource · Reading details