Editorial notebook / 05 dispatches
The notebook covers practical choices around AI workflows: scope, oversight, evidence, operations, and the hard parts that tool demos tend to skip.
Dispatch 01
Agent operations
8 min read
Three approval models—per action, per plan, and per boundary—and a practical default for a small team running its first supervised agent pilot.
Dispatch 02
Evaluation
9 min read
A compact way to move beyond glossy demos: real tasks, observable outcomes, trace review, a failure case, and repeatability.
Dispatch 03
Workflow design
8 min read
A decision guide for choosing the least-complex AI system that can improve a recurring task—and knowing when an agent is actually justified.
Dispatch 04
Operations
7 min read
A practical review packet helps a team inspect the request, boundary, path, outcome, and human decision—without mistaking a polished summary for evidence.
Dispatch 05
Context design
8 min read
A practical way to give an agent the smallest trustworthy packet for the job, then leave a handoff a human or fresh session can actually use.
Dispatch 06
Independent test
10 min read
A pre-registered 1,000-call check of TypeSafe’s System One model: stated confidence tracked accuracy in every bin we could test, and it answered all 1,150 calls with typed output — while a chat baseline failed to return usable JSON on a third of calls.
The format
Each note begins with a decision a team actually has to make.
The standard
External claims are sourced, limitations are stated, and the working opinion is distinguished from reported fact.
The promise
A smaller collection of durable guides beats a stream of thin “AI news.”