Research
Questions worth studying.
An agenda for understanding frontier intelligence.
These are areas of exploration. Each connects a technical question to the practical conditions in which AI is used.
Areas of exploration
What does a model actually know?
Model behavior & evaluation
Capability, reliability, and the gap between a benchmark score and performance on a real task.
Explore this area
How does intelligence meet information?
Language models & data
Structured data, databases, retrieval, and the ways language models read, reason over, and change information.
Explore this area
When can a model act usefully?
Agents & workflows
Tool use, orchestration, and the conditions under which an intelligent workflow remains observable and dependable.
Explore this area
What changes outside the API?
Models & deployment
Local and open-weight models, inference, deployment, and the tradeoffs between capability, control, and cost.
Explore this area
What makes capability usable?
Infrastructure
The compute, data infrastructure, and measurement tools that support repeatable experiments and practical deployment.
Explore this area
Where can intelligence make a difference?
AI in serious domains
Questions at the intersection of advanced AI and science, medicine, and infrastructure, where evidence and domain context matter.
Explore this area
Writing & notes
Loading the page