privacy

2026-09-09T08:29:10Z: split-llm-gradient-privacy-leakage closed (faded) — The hits do not establish that Scott uses split training or relies on gradients being private; his tokenized agent boundaries address a different exposure surface, so this does not yet challenge his architecture or change what
2026-08-29T15:32:57Z: claude-cross-chat-memory-deletion closed (faded) — Anthropic’s separation of editable cross-chat memory from source conversations independently converges with Scott’s layered-memory architecture, but persistence after source deletion creates a consequential provenance and priva
2026-08-28T21:39:10Z: openai-europe-chatgpt-ads closed (absorbed) — The radar already tracks this same development on `radar:chatgpt-free-go-ads`; the European rollout and policy effective date add implementation detail rather than a new direction. It materially bears on Scott’s fiduciary-agent and
2026-08-24T03:29:42Z: anjadhe-private-macos-assistant closed (faded) — This repeats the personal-agent operating model already set out in “Personal Agents Three Jobs ebook” and actively implemented in “OpenClaw — single-tenant AI operating stack”: durable cross-context memory, semantic joining, atte
2026-08-23T19:32:27Z: openai-computer-history-memory closed (faded) — OpenAI’s launch independently productizes Scott’s Attention Flight Recorder pattern: passive, chronological computer-activity capture used to reconstruct context for an assistant. As a consequential platform vendor arriving at his
2026-08-23T17:24:56Z: otel-genai-private-metrics-sketches closed (faded) — Scott’s Agent Observability page already holds the core position: production agents need trace-derived usage and cost telemetry, while his privacy-tokenized boundary addresses avoiding sensitive raw-data exposure. This projec
2026-08-22T20:25:22Z: meridian-local-developer-workjournal closed (faded) — Scott already holds and implements the core position: developer activity should compile into durable, provenance-bearing memory with human-controlled promotion, as documented in Code What Transcript Why, Cognitive Git, and h
2026-08-20T02:29:05Z: privibe-local-llm-cli closed (faded) — The radar already tracks essentially the same unvalidated local llama.cpp coding-CLI proposition in “Ante offline coding agent,” “DLLM llama.cpp coding agent,” and “Pi native llama.cpp runtime.” Privibe overlaps Scott’s Ask terminal agent
2026-08-19T09:32:07Z: anthropic-claude-tracker-privacy closed (faded) — Scott already holds that telemetry must be disclosed, proportionate, redacted and independently inspectable in “No-Surveillance-Creep Rule,” “Observability for Agentic Systems,” and “Sovereign Software Assurance”; the radar also
2026-08-18T21:35:23Z: wildstatic-shared-public-ai-memory closed (window-closed) — Scott already argues for scoped, hygienic memory and single-tenant isolation in “Memory Hygiene,” “Single-tenant AI appliance,” and OpenClaw. Wildstatic is presently only an unverified public counterexample to those ar