Papers
These papers explain my work in a way that focuses on ideas and outcomes rather than technical implementation details. They cover how AI & systems can be made more reliable as well as how structure and geometry may influence physical behavior.
System Paper
AI Papers
- Why AI systems need to have multiple confidence scores per result (for depth of trust)
- How training models to be 'safe' will never work for 100% accuracy esp. in regulated territory
- How to Build AI and Other Systems That Refuse to Act Unless They’re Actually Right
- How AI Can Safely Understand and Modify Software Without Breaking It
- Designing Computer Systems That Stop Safely Instead of Guessing
- Preventing AI From Copying Past Work or Inventing Fake Originality
- Separating Human Intent From AI Memory and Automation
- Why AI Systems Need External Rules Instead of Trusting Their Own Output
- Understanding Why Large Language Models Fail — and How to Design Around It