Marty Sampson.aiAI, applied
Field notes

Practical AI concepts, kept in proportion.

Source-backed explainers that distinguish documented mechanisms, original interpretation and work that has not been tested.

Visual explainer

Context is not memory

A visual guide to saved history, retrieval and compaction, with a downloadable evaluation checklist.

memorycontextresearch
Read →
Field guide

What should an AI agent remember?

A practical retention test for durable instructions, learned notes, searchable history and ephemera.

memorypermissionsreview
Read →
Explainer

Private notes are not a publishing queue

A synthetic, privacy-first model for transforming private research into public evidence.

privacyprovenancepublication
Read →
Method

A local-model result needs a record

A reproducibility method and blank experiment card—not a benchmark or result.

local modelsevaluationmethod
Read →
Explainer

When a fact changes, history should not disappear

A hand-authored temporal graph separates current facts, history, provenance and contradiction.

knowledge graphstimeretrieval
Read →
Explainer

An instruction is not a permission boundary

A synthetic five-layer trace separates guidance from enforceable capability boundaries.

agentspermissionssecurity
Read →
Systems explainer

A live signal is not yet a model turn

An original sequence separates continuous tracks, timestamps, optional segmentation, model context and interruption.

multimodalrealtimeturns
Read →
Systems explainer

Where Omarchy fits in an AI workstation

A source-backed layer map shows what the distribution integrates—and what remains a separate agent, provider or project concern.

Omarchyworkstationsagents
Read →
Workflow explainer

A media workflow is a chain of custody

Separate generation, editing, transcription and release while preserving rights, consent and provenance.

mediaconsentprovenance
Read →
Methods note

Simulated feedback is a hypothesis, not a user study

Keep phrasing coverage, deterministic system behavior, generated interpretation and real-user evidence in separate records.

evaluationpersonasevidence
Read →
Methods note

Can your PDF extraction support the citation?

Inspect reading order, tables, footnotes, captions, and anchors from the claim backward.

PDFretrievalcitations
Read →
Systems explainer

A research assistant has several jobs

Keep finding, extracting, proposing, executing, and concluding tied to the evidence each stage actually produced.

researchevidenceagents
Read →
Systems explainer

Who owns an email before it is sent?

Trace receipt, storage, agent context, drafting, confirmation, and the outbound send request as different state transitions.

agentsemailstate
Read →
Methods note

A benchmark is a question, not an answer

Match a published evaluation to the workload, method, grader, version, and remaining local evidence gap.

benchmarksevaluationdecisions
Read →