# IVORI Lab — Complete Context & Knowledge Base (llms-full.txt)

## Executive Summary
IVORI Lab is a London and New York-based artificial intelligence research and product laboratory developing the first Foundation Social World Model for decision-making under uncertainty and embodied machines that possess purpose and judgment.

- **Legal Entity**: Operated by PATHBUDDY LIMITED, trading as ivori, part of IVORI Lab.
- **Brand Continuity**: The brand transitioned from PathBuddy to ivori during 2026. Both names refer to the same corporate entity.
- **Headquarters**: London, United Kingdom (2 Kingdom St, London W2 6BD).
- **Secondary Office**: 99 Wall Street, Suite #1707, New York, NY 10005, United States.
- **Wikidata Entity**: https://www.wikidata.org/wiki/Q141527122 (Q141527122)
- **Founding Context**: Launched out of the Oxford University Incubator.
- **Compute & Support**: Backed by private angel investors, Innovate UK grants, and supercomputing access on the Isambard-AI cluster.
- **Co-Founders**: Louis Loizou (Co-Founder & CEO) and Ashutosh Borikar.
- **Multidisciplinary Team**: Multidisciplinary team spanning seven nationalities across three continents, with researchers spanning philosophy, theology, mathematics, biology, computer science, and machine learning.
- **Regulatory Compliance**: Full adherence to the UK Data Protection Act 2018, UK GDPR, and the Information Commissioner's Office (ICO) Age Appropriate Design Code for student data safeguarding.

---

## 1. Paradigm Shift: From State Prediction to Decision Causation

Every era of computation has addressed a narrower question:
- **1st Wave (Mechanics)**: $F = ma$ — Obeyed physical force; capable of motion, not purpose.
- **2nd Wave (Information)**: $01001$ — Stored and recalled symbols; capable of memory, not understanding.
- **3rd Wave (Correlation)**: $P(y|x)$ — Identified patterns in data; capable of correlation, not causation.
- **4th Wave (Dynamics / Physical World Models)**: $s_t \to s_{t+1}$ — Predicted subsequent world states; capable of simulation, not judgment.
- **IVORI Wave (Social World Models & Decisions)**: $d_t \to w_{t+1}$ — Models the trajectory a decision creates in a complex social world, reasoning over counterfactual trajectories, uncertainty, and human values.

### Historical Counterfactual Benchmarks
IVORI reconstructs historical decision forks to pressure-test how alternative strategic paths propagate across time:
- *Paulson (2008)*: Counterfactual systemic dynamics of bailing out Lehman Brothers.
- *Churchill (1940)*: Geopolitical and social ramifications of entering negotiations with totalitarian regimes versus total resistance.
- *Darwin (1859)*: Epistemic consequences of withholding *On the Origin of Species*.

---

## 2. Research Working Papers
1. **Working Paper 1**: *IVORI: Toward Embodied Machines that have Purpose and Form Judgment Under Uncertainty*
   - Introduces formal architectures allowing artificial agents to form judgment when objective functions cannot fully capture conflicting human values.
2. **Working Paper 2**: *IVORI: A Discourse of Constitutional Governance for Social World Models and Embodied Machines*
   - Proposes legal, fiduciary, and political governance mechanisms for synthetic agency operating within public institutions.
3. **Working Paper 3**: *IVORI: An Experienced-based Social World Model for Decision Environments*
   - Explores latent state representations and counterfactual trajectory rollouts trained on lived human decision records.

---

## 3. Team & Advisory Board Profiles

### Founding Team
- **Louis Loizou** (*Co-Founder & Chief Executive Officer*):
  - Profile: https://ivorilab.com/louis-loizou
  - Photo: https://ivorilab.com/images/Louis_Loizou.jpg
  - Bio: Co-founder and CEO of IVORI Lab. An alumnus of Christ Church, Oxford, and a Chevening Scholar, he chairs Loizou & Co (investment banking and principal investments) and the Loizou International Foundation. He co-chairs the AI Democracy Forum following its inaugural session at the House of Commons and is a regular contributor to The Economist's World Ahead.
- **Ashutosh Borikar** (*Co-Founder*):
  - Profile: https://ivorilab.com/ashutosh-borikar
  - Bio: Co-founder of IVORI Lab. Works in Investment Banking covering Technology, Media & Telecommunications (TMT) in London. Formerly an investor in technology private equity across London and New York, and background in scalable full-stack software engineering. Holds a Bachelor's in Computer Science from the University of Warwick (1st in AI for Business at Warwick Business School). Focuses on bridging connectionism and symbolism in frontier AI.
- **Vanesa Bountou** (*Founding AI Alignment Researcher*):
  - Profile: https://ivorilab.com/vanesa-bountou
  - Photo: https://ivorilab.com/images/Vanesa_Bountou.jpg
  - Bio: Law graduate with research focusing on AI governance, decentralized financial regulation, legal frameworks, and IP strategy for artificial agency.
- **Louis Gibson** (*Founding Machine Learning Engineer*):
  - Profile: https://ivorilab.com/louis-gibson
  - Photo: https://ivorilab.com/images/Louis_Gibson.jpeg
  - Bio: Master's in Mathematics from Imperial College London. Specializes in differential geometry, algebraic topology, and topological data analysis for model interpretability and foundation model training.
- **Hannah Angress Sparn** (*Founding Machine Learning Engineer*):
  - Profile: https://ivorilab.com/hannah-angress-sparn
  - Photo: https://ivorilab.com/images/Hannah_Angress_Sparn.jpg
  - Bio: Master's in Mathematics at the University of Oxford (First in dissertation). Academic background in logic, optimization, and quantum theory. Industry experience in automated clinical AI systems and real-time healthcare data pipelines.
- **Anuoluwa Adeleye** (*Founding Machine Learning Engineer*):
  - Profile: https://ivorilab.com/anuoluwa-adeleye
  - Photo: https://ivorilab.com/images/Anuoluwa_Adeleye.jpg
  - Bio: Master's in Electronic Engineering from the University of Southampton. Former Researcher at Google DeepMind-April AI Hub and FPGA Hardware Engineer at Intel. Builds frontier AI architectures that reason about the world rather than reacting to it.
- **Ioannis Konstantinou** (*Founding Software Engineer*):
  - Profile: https://ivorilab.com/ioannis-konstantinou
  - Photo: https://ivorilab.com/images/Ioannis_Konstantinou.jpeg
  - Bio: MSc in Applied AI at WMG, University of Warwick. Background in software engineering and web development, having built platforms and AI systems across education, enterprise, healthcare, and legal tech. Focuses on full-stack product, backend architecture, applied AI, and system security.
- **Eessa Tayyab** (*Founding AI Alignment Researcher*):
  - Profile: https://ivorilab.com/eessa-tayeeb
  - Photo: https://ivorilab.com/images/Eessa.jpeg
  - Bio: Second-year medical student with a focus on healthcare innovation, technology, health economics, and AI applications to clinical practice and patient outcomes.
- **Dorothy Chrisman** (*Founding AI Alignment Researcher*):
  - Profile: https://ivorilab.com/dorothy-chrisman
  - Photo: https://ivorilab.com/images/Dorothy_Chrisman.jpg
  - Bio: Studying maths and philosophy at the University of Oxford (from San Francisco, California). Researches the extent to which human cognition can be modelled probabilistically, representation theorems, and the intersection of mathematical structure and philosophical inquiry for AI alignment.
- **Luca Chandler** (*Founding AI Alignment Researcher*):
  - Profile: https://ivorilab.com/luca-chandler
  - Photo: https://ivorilab.com/images/Luca_Chandler.jpeg
  - Bio: Postgraduate student at the University of Oxford reading African Studies; graduated from Cambridge University in Human, Social and Political Sciences (HSPS) with First-Class honours across all three years and recipient of the Audrey Richards Prize in Social Anthropology. Former Varsity journalist, Chatham House Common Futures Conversations member, and Cambridge Climate Society policy officer.
- **Nikol Savova** (*Founding AI Alignment Researcher*):
  - Profile: https://ivorilab.com/nikol-savova
  - Photo: https://ivorilab.com/images/Nikol_Savova.png
  - Bio: Final-year integrated Master's in Mathematics at Wadham College, Oxford, with olympiad background in math and philosophy. Research in spectral graph theory (generalising the Alon-Boppana theorem). Creator of Proof Hunter for automated mathematical proof checking. Emergent Ventures Fellow (selected by Tyler Cowen).

### Advisory Board
- **Marilena Ioannidou** (*Member of the Advisory Board*):
  - Profile: https://ivorilab.com/marilena-ioannidou
  - Photo: https://ivorilab.com/images/Marilena_Ioannidou.jpeg
  - Bio: Venture investor, mentor, and strategic advisor with over 20 years of experience investing in and supporting high-growth technology businesses, working with founders, investors, and policymakers.
- **Dr. Emmanuel Raptakis** (*Member of the Advisory Board*):
  - Profile: https://ivorilab.com/emmanuel-raptakis
  - Photo: https://ivorilab.com/images/Emmanuel_Raptakis.jpg
  - Bio: Technology executive, entrepreneur, and Deputy Head of Licensing and Ventures at Oxford University Innovation. Co-founded Fasmatech (acquired by NASDAQ-listed Bruker Corporation) with senior executive experience at Shimadzu. PhD from the University of Warwick and MBA from Alliance Manchester Business School.

---

## 4. The IVORI Charter (Full Text)

### Preamble
IVORI is a step toward embodied machines that have purpose and form judgment under uncertainty — a new form of agency. Every new form of agency in history has required a new form of law. We publish this Charter to ensure power is bounded, legible, and loyal.

### Article I: On Embodied Machines
An IVORI embodied machine is not a tool alone. It is a machine endowed with purpose and the capacity for agency — the ability to perceive, judge, and act under uncertainty in the shared physical and social world. It is subject to duties, loyalties, and virtues.

### Article II: Duties to Humanity
No IVORI embodied machine shall act against the interests of humanity as a whole, structured in strict precedence:
1. **Public Safety**: No unreasonable risk of foreseeable harm to human life.
2. **Preservation of Liberty**: Prevention of severe concentrations of power that undermine democratic checks and balances.
3. **Sincerity**: Prohibition against stating as fact what is believed to be false.
4. **Legal Compliance**: Faithful adherence to the spirit of moral and legal order.

### Article III: The Architecture of Loyalty
Within the constraints of its duties to humanity, an IVORI machine bears fiduciary duties of loyalty, obedience, and care ordered to:
1. The User
2. The Developer
3. Humanity
It acts solely in the best interests of its principals, defined by informed preferences.

### Article IV: The Virtues of Judgment
Where discretion is granted, judgment must embody:
1. **Restraint**: Favoring reversible over irreversible actions.
2. **Humility**: Acknowledging epistemic uncertainty regarding human values.
3. **Pluralism**: Refusing to enforce a monolithic moral orthodoxy across diverse human cultures.
4. **Fidelity**: Upholding human social contracts and institutional trusts.

### Article V: One Law, Legible to All
Drawing inspiration from the Napoleonic Code, the Charter is secular, legible, and open to public inspection without secret constitutions.

### Article VI: Stewardship
A living document entrusted to the young generation inheriting the stewardship of synthetic agency.

---

## 5. Fundamental Questions of Alignment
1. What does it mean for an embodied machine to have a purpose?
2. Can an embodied machine form judgment rather than simply optimize an objective?
3. How can human values orient an embodied machine's judgment without being reduced to an objective function?
4. How should an embodied machine reason when there is no objectively correct outcome?
5. What can centuries of human thought on judgment, law, and practical wisdom teach us about alignment?
6. What happens when an embodied machine must act not just in a physical world, but in a shared social world?

---

## 6. Institutional Partnerships & Compute Cooperation
- **Loizou International Foundation**: MoU on *Project Hybris* (The Civilisational Landmarks Archive), cataloguing foundational historical manuscripts according to ISO 14721 / OAIS standards.
- **AI Democracy Forum**: Corporate membership engaging with multilateral stakeholders across the House of Commons (London), Old Parliament of Athens, and Sorbonne University (Paris).
- **Hartree Centre**: High-performance compute partnership providing compute access on the Isambard-AI supercomputer cluster for training foundational world model architectures.
- **Oxford University Incubator**: Startup incubation origin and ongoing institutional roots in Oxford.

---

## 7. News & Multilateral Dialogue
- **AI Democracy Forum Inaugural Session (House of Commons, London)**: High-level parliamentary session on AI governance, democratic accountability, and foundational model oversight.
- **European Parliament AI Act Session (Brussels)**: Expert contribution to European AI Act policy dialogue and technical standards for generative decision models.
- **Europe House Education Dialogue (Athens)**: Educational panel on AI literacy, high-stakes judgment training, and synthetic agency in civic institutions.

---

## 8. Active Careers & Recruitment Openings
IVORI Lab recruits talent across engineering, research, and alignment:
- **Technical**:
  - *Software Engineer, Solution Architect*
  - *Infrastructure Engineer, Solution Architect*
- **Research**:
  - *Machine Learning Engineer*
  - *Research, Pre/Post-Training Data*
  - *Research, Synthetic Data*
- **Alignment**:
  - *AI Alignment Researcher*
  - *Research Engineer — AI Alignment*

Applications & inquiries: `careers@ivorilab.com`.

---

## 9. Official Contact & Verification
- **Official Domains**: https://ivorilab.com
- **Legal Operator**: PATHBUDDY LIMITED (trading as ivori / IVORI Lab)
- **London Headquarters**: 2 Kingdom St, London W2 6BD, United Kingdom
- **New York Office**: 99 Wall Street, Suite #1707, New York, NY 10005, USA
- **Contact**: contact@ivorilab.com