From the blogAll posts →
  • Security Is a Big World
    Two frontier labs gated their strongest cyber capabilities two days apart. But discovery was never the bottleneck. Defensive advantage lives in the harness.
  • The Transcript Is Deleted. What It Taught the System Is Not.
    OpenAI announced yesterday that it deletes your data and tracks patterns across your interactions, in the same post, and both are true at once.   The bundle is…
  • Claude’s Watermark Is Not Encryption
    Anthropic announced on August 11 that Claude models launched from August 2, 2026 onward will embed an invisible watermark in generated text, worldwide, as part of implementing its…

Essays, notes, spines, and readings on measurement, meaning, and human oversight of AI systems.

ESSAYNOTESPINEREVIEW
Research threadsAbout the research →
01Evaluation and measurementground truth · instrumentation · validity
02Information and meaningsemantics · embeddings · context
03Affective computingemotion · judgment · machine behavior
04Governancestandards · taxonomies · policy
05Human-AI collaborationoversight · teaming · supervision