I’m interested in how new tools expand human capability.
- General vs. specialized intelligence
- Agent harnesses: context engineering, memory, multi-agent orchestration
- Simulated work environments and RL environments
- Expert human data
- Human-AI interaction data and capability evals
- Long-horizon agent reliability
- Coordination costs as a constraint on scale
- Applied mathematics and computational economics
- Investing in early-stage AI, robotics, quantum etc.
- Based in San Francisco

