sequences

Week of 2026-05-09

• Agent capability as context, harnesses, memory, process control, and workflow data.
• Alignment as eval design, constitutions, personas, and institutional formation.
• Business gravity: durable AI value moves toward owned context, actuation, and feedback loops.
• Inference economics: long-context agents turn latency, cache locality, and tool-call placement into product architecture.

• Constitutions / Monastic Alignment
  Sequence planning shell for AI constitutions, moral training, and monastic alignment.
• Cybernetic Economics / Business Gravity
  Sequence planning shell for AI business gravity, cybernetic economics, context moats, and outcome loops.
• Forecasting / Policy / Progress
  Sequence planning shell for timelines, progress forecasts, benchmark interpretation, and policy context.
• Context / Harness
  Sequence planning shell for context engineering, harnesses, memory, and agent control loops.
• Inference Economics
  Sequence planning shell for agentic inference costs, token factories, managed runtimes, tool-call latency, and model-specific serving adaptation.
• Long-Horizon RL / Goal-Conditioned Control
  Sequence planning shell for long-horizon RL, goal-conditioned control, horizon reduction, and planning.
• Research Alignment / Evals
  Sequence planning shell for AI alignment research, evals, personas, control, and benchmark interpretation.
• Vertical Tasks / Data
  Sequence planning shell for vertical AI, workflow data, legal benchmarks, and domain-specific task economics.
• Visual Artifacts / Source Presentation
  Sequence planning shell for visual research presentation, browsable artifacts, and source-forward sequence design.