Ensuring AI safety and developing Safe Autonomous Organizations without humans in the loop to solve humanity's biggest problems
Requirements
- Experience with AI control research
- Knowledge of AI safety protocols
- Familiarity with foundation models and LLM agents
- Programming skills to build and deploy SAOs
- Understanding of AI alignment techniques
- Experience with evaluation frameworks for AI models
- Knowledge of software development for autonomous organizations
Responsibilities
- Build control protocols and/or alignment techniques to make sure our SAOs are safe
- Monitor for misalignment and implement AI control protocols to prevent similar behaviour in the future
- Identify interesting ventures, build integrations and deploy our agents in the real world
- Scale SAOs to be everywhere in society
- Develop and evaluate the capabilities of foundation models and LLM agents
- Create, run, and score custom evaluation tasks to ensure safe, robust and reliable models
- Audit dangerous capabilities in the next generation of models
Other
- US citizen/visa only
- Senior and above school year
- In-person work in San Francisco
- Competitive salary and stock compensation
- Mission critical to ensuring humanity's prosperous future
