Freelance AI evaluation
The financial anchor, and the live root of everything else. Since 2024 I've done human-in-the-loop AI-evaluation work for US-based AI quality-assurance platforms - judging model output for honesty, safety, and subtle failure, and debugging AI-generated code. It is steady, detailed judgment work, it is the skill the research is built on, and it is what pays for the research. It's the engine, not a service I'm selling you here.