The reading room
Explore the articles.
Start with a question. Follow a topic. Read at your own pace.
Article archive
6 articles- AI × Data Security
A tool timeout is an unknown outcome—not proof of failure
Before an AI assistant retries a write, it must find out what happened. Separate the request, the durable result and the evidence a user can actually inspect.
- AI Foundations
When switching models mid-session pays off
A cheaper model can make the next call more expensive when it starts with a cold cache. Compare the remaining workflow, including the cold start and the handoff.
- AI × Data Security
GPT-6 Astra: from clever answers to sustained work
Astra improves long-context retrieval and task execution. Its new Codex memory design and better handling of boundaries point toward a more useful kind of autonomy.
- AI × Data Security
Your AI knows the code. Who knows the plan?
Self-awareness, disappearing context and the surprising impact of the harness: why capable agents still need orchestration, and where T3 Code fits.
- Data Security
Training, inference and storage: trace every copy
Map how a system uses, processes and retains data. Check the purpose, access and deletion rules for each copy.
- AI Foundations
Models, tools, agents: put each control in the right place
A model proposes output. A tool performs an operation. An agent loop decides what happens next. Map each safety rule to the component that enforces it.
New ideas. No inbox required.
Follow every published article in your own RSS reader.