Dr Elija Perrier

About

Welcome to my website. I'm Elija Perrier. I'm an interdisciplinary mathematical physicist and researcher specialising in quantum–classical AI systems integration, multi‑agent modelling and AI alignment using optimal control theory. A bit more about me:

• I publish research on quantum machine learning, AI risk measurement and governance, and multi‑agent systems.
• I completed my PhD in quantum machine learning at the University of Technology, Sydney and my other Bachelor degrees at the University of Sydney, NSW, and Murdoch University in Perth, Western Australia.
• I have a multi-disciplinary background spanning physics, computer science, mathematics, economics, law and philosophy.
• I've been a lawyer for over 20 years as a practitioner and legal academic across Australia, Europe and the UK. I currently work in investment banking.
• I'm also a software engineer proficient in a range of frontend and backend languages (Python mostly), stack developer (AWS) and HPC/Singularity.

Current activities

I am currently an honorary fellow at the Centre for Quantum Software and Information at UTS, Sydney where I completed my PhD. In quantum, my focus is on classical and quantum AGI, especially quantum agents. I am also an a research affiliate with the Gradient Institute based out of the University of Sydney. There I work on mostly AI alignment and control/measurement. I was previously a visiting scholar at Cambridge Law School and a fellow at the Stanford Center for Responsible Quantum Technology.

Education

• PhD, Quantum Machine Learning/AI, University of Technology Sydney (2018–2024). Thesis: “Quantum Geometric Machine Learning”.
• BSc Physics/Mathematics (First Class Honours) & BA Economics/Philosophy, University of Sydney (2011–2017).
• LLB & BA Communications/English (First Class Honours), Murdoch University (1999–2004). Admitted as solicitor and barrister of the High Court of Australia and Supreme Court of NSW (2005).
• Graduate Diploma in Legal Practice College of Law, Sydney (2005)

Select Research & Professional Positions

• Research Affiliate, Gradient Institute.
• Visiting Fellow, Centre for Quantum Software & Information (UTS) (2024–present).
• Visiting Scholar, Cambridge Law School, University of Cambridge (June 2024).
• Fellow, Stanford Center for Responsible Quantum Technology (Dec 2023–Jan 2025).
• Founder, Measure Theory Analytics (2018–present) – developed multi‑agent simulations combining LLMs with microsimulation techniques for electoral modelling and campaign strategy.
• Research Affiliate, Project Q, University of Sydney (2021–2023); Humanising Machine Intelligence & MINT Lab, Australian National University (2020–2023).
• Data Analyst, NSW Labor (2017–2021) – built machine‑learning and agent‑based models for electoral campaigns.

Selected Publications

Recent Articles (2024–2025)

• Quantum AIXI: Universal Intelligence via Quantum Information – accepted for publication at AGI‑25 Conference (2025).
• Quantum AGI: Ontological Foundations – accepted for publication and talk at AGI‑25 Conference (2025).
• Hamiltonian Formalism for Comparing Quantum and Classical Intelligence – accepted for AGI‑25 (2025).
• Watts‑Per‑Intelligence: Part I (Energy Efficiency) – accepted for AGI‑25 (2025).
• Infrastructure for AI Agents – accepted in Transactions on Machine Learning Research (2025).
• KP Quantum Neural Networks – accepted for the Geometric Science of Information Conference (2025).
• Threshold Crossings as Tail Events for Catastrophic AI Risk – SAFE AI workshop (2025).
• Towards Measurement Theory for Artificial Intelligence – under review (2025).
• Agent Identity Evals: Measuring Agentic Identity - submitted for ICLR 2026
• Out of Control--Why Alignment Needs Formal Control Theory (and an Alignment Control Stack) - a position paper under review at Neurips 2025
• Statistical Scenario Modelling and Lookalike Distributions for Multi-Variate AI Risk - under review for Statistics (Springer)
• Position: Stop Acting Like Language Model Agents Are Normal Agents - under review at Neurips 2025

Earlier Articles

• QDataSet: Quantum Datasets for Machine Learning. Nature Scientific Data 8, 582 (2022).
• The Quantum Governance Stack: Models of Governance for Quantum Information Technologies. Digital Society 1, 22 (2022).
• Quantum Fair Machine Learning. Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society (2021).
• Quantum Geometric Machine Learning for Quantum Circuits and Control. New Journal of Physics 22 (10), 103056 (2020).
• Deep hierarchical graph convolution for election prediction from geospatial census data AAAI 2019 33 (01), 647-654
• World Economic Forum Quantum Computing Governance Principles: Insight Report (2021)
• Ethical Quantum Computing: A roadmap arXiv:2102.00759

Book Chapters

• "Quantum Information Technology and the Law" - forthcoming chapter in Quantum Technology Governance: Law, Policy and Ethics in the Quantum Era (Springer, 2025)
• "Quantum Governance and the Life Sciences" - forthcoming chapter in Quantum Technology Governance: Law, Policy and Ethics in the Quantum Era (Springer, 2025)
• "Quantum Information Technologies and Public International Law" - forthcoming chapter in Quantum Technology Governance: Law, Policy and Ethics in the Quantum Era (Springer, 2025)

Key Projects

• Quantum AGI: Ontological Foundations: Developing operational theory of quantum agency in light of quantum foundations' results
• AI Agent Infrastructure & Alignment: Developing foundational frameworks for deploying AI agents at scale, including evaluation methods and control‑theory‑based models for AI alignment.
• Measurement Science for AI Risk: Applying statistical risk analysis and measurement theory to quantify AI‑specific risk and mechanistic interpretability.
• Quantum AGI: Integrating quantum information theory to develop a comprehensive quantum and classical theory of agency.
• Multi‑Agent Political Simulation Platform: Designed a production‑scale LLM‑powered microsimulation system modelling national population dynamics and real‑time electoral prediction.
• Quantum Geometric Machine Learning: Developed new techniques combining differential geometry with quantum computing, including novel quantum neural network architectures and large‑scale quantum datasets.

Grants

• DeepFunding 2025 Grant for Quantum AGI: Ontological Foundations (see above)

Select Professional Experience

Data Science

Data Analyst / Staffer, Australian Labor Party / MPs (2017–2021). Led data science across campaign activities, building full‑stack applications, multi‑agent simulations and voter modelling.

Legal & Finance

Associate Director, ANZ Banking Group (2021–present). Senior manager responsible for executing complex financial transactions (USD $1B+), risk analysis and project management across resources, healthcare and financial institutions.

Senior Legal Counsel, Challenger Limited (2019–2021). Advised on multi‑jurisdictional finance transactions, regulatory frameworks and quantitative risk assessment.

Associate, Freshfields Bruckhaus Deringer (London) / Allen & Overy (Frankfurt) (2007–2011). Worked on cross‑border syndicated financing and financial regulation.

Teaching & Lecturing

Served as tutor and lecturer in constitutional, administrative and jurisprudential law (Macquarie University), deep learning and public economics (University of Sydney), physics (UTS) and tutor in law (Murdoch University).

Contact

If you would like to get in touch, please reach out via email or connect through social media.

• Email: eper2139 at uni.sydney.edu.au
• GitHub: github.com/eperrier
• LinkedIn: linkedin.com/in/elija-perrier-63695810/