Curriculum overview
Mag.AI-S (Magister of Artificial Intelligence in Science) is 18 courses in three terms. You progress from world models, measurement, and reproducibility to experiments, inference, and literature-scale systems, then to competition, allocation, governance, and a capstone thesis. Work is artifact-driven: you build executable articles (Jupyter notebooks in each course’s Jupyter Book, run in iNQspace), not slide-only summaries. Policy: Executable articles.
Each course is published as its own book (separate site). The table below links to those books; the full index is on Course sites.
Program arc¶
| Term | Theme | What you’re building toward |
|---|---|---|
| I — World models & integrity | Hypotheses, uncertainty, reproducibility, and ethics as executable systems | Worlds you can test before you scale claims |
| II — Evidence & systems at scale | Experiments, statistical learning, simulation, literature intelligence, instruments | Coordinated pipelines with delays, noise, and honest failure modes |
| III — Strategic science | Comparison, policy, allocation, autonomy, and governance | Strategic depth, governed autonomy, and a defense-ready thesis |
Typical path: complete Term I (AINS-S6001 → AINS-S6006), then Term II (AINS-S6101 → AINS-S6106), then Term III (AINS-S6201 → AINS-S6206). Entry: AINS-S6001 is the recommended start.
Technical spine: Across all terms, the program weaves in AI literacy (evaluation, grounding, retrieval, generative workflows, scoped tooling). See the full curriculum under AI literacy spine.
Term I — World models & integrity¶
| Code | Course | Core focus |
|---|---|---|
| AINS-S6001 | Scientific World Models & Executable Hypotheses | Assumptions, observables, and tests as the smallest honest world |
| AINS-S6002 | Measurement, Uncertainty & Calibration | Error, calibration, and limits as first-class objects |
| AINS-S6003 | Computational Reproducibility & Provenance | Environments, seeds, and lineage that survive reruns |
| AINS-S6004 | Data Systems, Documentation & Lineage | Schemas, metadata, and audit trails |
| AINS-S6005 | Literate Computation & Experiment Logs | Notebooks and logs as inspectable science |
| AINS-S6006 | Ethics, Integrity & Responsible Conduct | Integrity, authorship, and misuse boundaries |
Term II — Evidence & systems at scale¶
| Code | Course | Core focus |
|---|---|---|
| AINS-S6101 | Experiment Design & Causal Reasoning | Design and identification for strong claims |
| AINS-S6102 | Statistical Learning & Model Evaluation | Generalization, evaluation, and honest metrics |
| AINS-S6103 | Simulation, Generative Models & Synthetic Data | Synthetic data and simulators under constraints |
| AINS-S6104 | Literature Intelligence, RAG & Evidence Synthesis | Retrieval, grounding, and synthesis at scale |
| AINS-S6105 | Instruments, Signals & ML for Measurement | Signals from instruments and learned measurement |
| AINS-S6106 | Open Science, Collaboration & Review Stress Tests | Adversarial review and collaboration fragility |
Term III — Strategic science¶
| Code | Course | Core focus |
|---|---|---|
| AINS-S6201 | Competing Hypotheses & Model Comparison | Model comparison and strategic interaction |
| AINS-S6202 | Research Strategy as Policy Under Uncertainty | Policy mapping states to actions |
| AINS-S6203 | Portfolio & Resource Allocation for Research Programs | Budget and risk across initiatives |
| AINS-S6204 | Autonomous Lab & Instrumentation Systems | Autonomy with human-in-the-loop boundaries |
| AINS-S6205 | Safety, Governance & Control in Scientific AI | Governance and oversight for scientific AI |
| AINS-S6206 | Magisterium Thesis (Science) | Capstone — deploy, validate, document, defend |
Where to go next¶
All course books (index): Course sites
Course-by-course detail: Full curriculum
How you’re assessed: Assessment framework
Credential: Certificate
Magisterium deployment: magisterium.md
Legal¶
The Castalia Institute Magisterium confers proprietary credentials based on demonstrated work and evaluation. These credentials are not accredited academic degrees and do not confer professional licensure.