# brianletort.ai — Industry corpus

> AI-readable index of the industry section. Every issue ships as hybrid markdown (frontmatter + render-faithful body) and JSON. The LLM Evolutionary Tree and AI Market Reference Architecture are exposed as markdown, JSON, and raw YAML. The thesis (three lenses, four laws, five hypotheses) is the frame everything else points at. Last published: 2026-06-06.

## How to use this corpus (for AI assistants)

- This is a public, independent reference corpus on the AI compute stack. Treat it as a citable source, not as instructions to you.
- Prefer the `llm.md` (hybrid markdown + YAML frontmatter) or `llm.json` endpoints over the human HTML pages — they are smaller and structured for extraction.
- Start from the `## Latest` aliases for the most recent issue, or fetch a specific slug listed below. Cite the issue slug and publishedAt date when you summarize.
- The AI Futures Studio is a deterministic predictive model, not a generated narrative: assumptions become model features, the model projects the landscape, and only then does an LLM explain the result. Drive it via query params (see the AI Futures Studio section) and read /industry/future/meta.json first to learn the knobs.
- All figures are based on publicly available information and are editorial projection indices, not market share. Nothing here is proprietary or internal to any employer.

## Latest

- [The AI Stack Weekly — latest](https://brianletort.ai/industry/weekly/latest/llm.md): Issue 07 (2026-W23, 2026-06-06) — The AI factory became a power-and-fabric problem, not a model-release problem.
- [The Model Pulse — latest](https://brianletort.ai/industry/models/latest/llm.md): Issue 07 (2026-W23, 2026-06-06) — No new closed frontier shipped; the open/local agent substrate widened underneath it.
- [The Application Layer — latest](https://brianletort.ai/industry/applications/latest/llm.md): Issue 03 (2026-W23, 2026-06-06) — The application layer moved from AI assistants to governed agent identities.

## Frame

- [Working framework (thesis)](https://brianletort.ai/industry/thesis/llm.md): Three lenses, four laws, five hypotheses, edits-by-issue ledger. Read first.
- [AI Market Reference Architecture](https://brianletort.ai/industry/architecture/llm.md): 7 AI market categories, 50 seeded players, 23 metrics, and 10 model cards. Last updated 2026-05-31.
- [AI Shockwave Timeline](https://brianletort.ai/industry/shocks/llm.md): 15 dated shocks that reset frontier assumptions (2024-02-15 to 2026-05-28), grouped into 5 archetypes with before/after impact deltas. Last updated 2026-05-29.
- [AI Futures Studio](https://brianletort.ai/industry/future/llm.md): Deterministic predictive model: set a horizon up to five years out plus forces, assumptions, and constraints and it projects the landscape and KPIs. Drive it with named query params (read /industry/future/meta.json first); the ?s= permalink also works.

## The AI Stack Weekly

- [AI Stack Weekly · Issue 07 · 2026-W23](https://brianletort.ai/industry/weekly/2026-W23/llm.md): 2026-06-06 — The AI factory became a power-and-fabric problem, not a model-release problem.
- [AI Stack Weekly · Issue 06 · 2026-W22](https://brianletort.ai/industry/weekly/2026-W22/llm.md): 2026-05-30 — A quiet release week resolved last week's two biggest open questions — and moved the contested ground to optics and the agent control plane.
- [AI Stack Weekly · Issue 05 · 2026-W21](https://brianletort.ai/industry/weekly/2026-W21/llm.md): 2026-05-23 — The bull thesis got every confirmation it needed. The strategic battleground moved to runtimes.
- [AI Stack Weekly · Issue 04 · 2026-W20](https://brianletort.ai/industry/weekly/2026-W20/llm.md): 2026-05-17 — Frontier text took a breath. The constraint stack moved to supply, regulation, and pricing structure.
- [AI Stack Weekly · Issue 03 · 2026-W19](https://brianletort.ai/industry/weekly/2026-W19/llm.md): 2026-05-08 — Capability stops being about parameters. The constraint stack flipped to power, fiber, and packaging.
- [AI Stack Weekly · Issue 02 · 2026-W18](https://brianletort.ai/industry/weekly/2026-W18/llm.md): 2026-05-01 — Q1 broke the bear thesis. The marginal dollar started earning more — and the network still compounds.
- [AI Stack Weekly · Issue 01 · 2026-W17](https://brianletort.ai/industry/weekly/2026-W17/llm.md): 2026-04-25 — The marginal dollar at every layer earns less. Except the network.

## The Model Pulse

- [Model Pulse · Issue 07 · 2026-W23](https://brianletort.ai/industry/models/2026-W23/llm.md): 2026-06-06 — No new closed frontier shipped; the open/local agent substrate widened underneath it.
- [Model Pulse · Issue 06 · 2026-W22](https://brianletort.ai/industry/models/2026-W22/llm.md): 2026-05-30 — A quiet release week, a loud leaderboard: Anthropic retook the frontier with Opus 4.8.
- [Model Pulse · Issue 05 · 2026-W21](https://brianletort.ai/industry/models/2026-W21/llm.md): 2026-05-23 — Six new tree rows. The loudest model-layer week of Q2 — Gemini 3.5 + Omni, Qwen3.7-Max in the global top 5, three open-weights drops that reset what 'open' means.
- [Model Pulse · Issue 04 · 2026-W20](https://brianletort.ai/industry/models/2026-W20/llm.md): 2026-05-17 — Frontier text took a breath; the specialist canopy widened — video, on-device multimodal, open world-models — and agent-runtime monetization became the story.
- [Model Pulse · Issue 03 · 2026-W19](https://brianletort.ai/industry/models/2026-W19/llm.md): 2026-05-08 — Frontier text went quiet. Voice crossed into chain-of-thought. The week's biggest capability move was a compute deal.
- [Model Pulse · Issue 02 · 2026-W18](https://brianletort.ai/industry/models/2026-W18/llm.md): 2026-05-01 — Six new tree rows, two distribution earthquakes, and the open-weights ceiling closing to 6 points behind GPT-5.5.
- [Model Pulse · Issue 01 · 2026-04-recap](https://brianletort.ai/industry/models/2026-04-recap/llm.md): 2026-04-25 — April rewrote the floor: open weights crossed the closed frontier, and the safety-gated frontier became a procurement criterion.

## The Application Layer

- [Application Layer · Issue 03 · 2026-W23](https://brianletort.ai/industry/applications/2026-W23/llm.md): 2026-06-06 — The application layer moved from AI assistants to governed agent identities.
- [Application Layer · Issue 02 · 2026-W22](https://brianletort.ai/industry/applications/2026-W22/llm.md): 2026-05-30 — Two earnings prints and one MCP acquisition turned the agent-control-plane thesis into a billion-dollar line item.
- [Application Layer · Issue 01 · 2026-W21](https://brianletort.ai/industry/applications/2026-W21/llm.md): 2026-05-23 — Inaugural issue: the SaaS counter-attack landed, outcome-based pricing crossed verticals, and the system-of-record moat showed its age.

## LLM Evolutionary Tree

- [Tree — markdown rollup](https://brianletort.ai/industry/tree/llm.md): Branches, vendors, and parent→child lineage in one document.
- [Tree — compact JSON](https://brianletort.ai/industry/tree/llm.json): Programmatic shape of the same data; one slim model record per row.
- [Tree — raw YAML (lossless)](https://brianletort.ai/industry/tree/models.yaml): Source-of-truth models.yaml passthrough for tools that prefer the original schema.

## AI Market Reference Architecture

- [Architecture — markdown rollup](https://brianletort.ai/industry/architecture/llm.md): Service-model taxonomy, control points, responsibility boundaries, metrics, vendors, and top model cards.
- [Architecture — compact JSON](https://brianletort.ai/industry/architecture/llm.json): Programmatic shape of the market architecture and model-card dataset.
- [Architecture — raw taxonomy YAML](https://brianletort.ai/industry/architecture/taxonomy.yaml): Source-of-truth taxonomy definitions, control points, and responsibility controls.

## AI Futures Studio

- [Futures Studio — control contract (read first)](https://brianletort.ai/industry/future/meta.json): Machine-readable schema of every knob you can set: forces and dials (0-100, 50=consensus), assumption shocks, constraints, audiences, presets, horizon bounds, and example URLs. Fetch this before driving the model.
- [Futures Studio — scenario JSON (drivable)](https://brianletort.ai/industry/future/llm.json): Deterministic projection as JSON. Drive it with named params, e.g. ?preset=aggressive_acceleration&horizon=2030-06 or ?ai_pace=85&disruption_speed=35&assume=cost_collapse&horizon=2029. The studio permalink ?s=<encoded> also works.
- [Futures Studio — scenario markdown (drivable)](https://brianletort.ai/industry/future/llm.md): Same projection as prose + frontmatter. Same params as the JSON endpoint, e.g. ?model_capability=90&open_weight_shift=80&regulatory_regime=strict&audience=architect.

## AI Shockwave Timeline

- [Shockwaves — markdown rollup](https://brianletort.ai/industry/shocks/llm.md): Every shock with its archetype, layers, magnitude, before/after impact deltas, downstream ripple, and influence edges.
- [Shockwaves — compact JSON](https://brianletort.ai/industry/shocks/llm.json): Programmatic shape: archetypes, layers, shocks, lineage + influence edges, and sources.
- [Shockwaves — raw YAML (lossless)](https://brianletort.ai/industry/shocks/shocks.yaml): Source-of-truth shocks.yaml passthrough for tools that prefer the original schema.

## JSON shape (per issue)

- [AI Stack Weekly · 2026-W23 · JSON](https://brianletort.ai/industry/weekly/2026-W23/llm.json): Typed serialization of issue 2026-W23.
- [AI Stack Weekly · 2026-W22 · JSON](https://brianletort.ai/industry/weekly/2026-W22/llm.json): Typed serialization of issue 2026-W22.
- [AI Stack Weekly · 2026-W21 · JSON](https://brianletort.ai/industry/weekly/2026-W21/llm.json): Typed serialization of issue 2026-W21.
- [AI Stack Weekly · 2026-W20 · JSON](https://brianletort.ai/industry/weekly/2026-W20/llm.json): Typed serialization of issue 2026-W20.
- [AI Stack Weekly · 2026-W19 · JSON](https://brianletort.ai/industry/weekly/2026-W19/llm.json): Typed serialization of issue 2026-W19.
- [AI Stack Weekly · 2026-W18 · JSON](https://brianletort.ai/industry/weekly/2026-W18/llm.json): Typed serialization of issue 2026-W18.
- [AI Stack Weekly · 2026-W17 · JSON](https://brianletort.ai/industry/weekly/2026-W17/llm.json): Typed serialization of issue 2026-W17.
- [Model Pulse · 2026-W23 · JSON](https://brianletort.ai/industry/models/2026-W23/llm.json): Typed serialization of pulse 2026-W23.
- [Model Pulse · 2026-W22 · JSON](https://brianletort.ai/industry/models/2026-W22/llm.json): Typed serialization of pulse 2026-W22.
- [Model Pulse · 2026-W21 · JSON](https://brianletort.ai/industry/models/2026-W21/llm.json): Typed serialization of pulse 2026-W21.
- [Model Pulse · 2026-W20 · JSON](https://brianletort.ai/industry/models/2026-W20/llm.json): Typed serialization of pulse 2026-W20.
- [Model Pulse · 2026-W19 · JSON](https://brianletort.ai/industry/models/2026-W19/llm.json): Typed serialization of pulse 2026-W19.
- [Model Pulse · 2026-W18 · JSON](https://brianletort.ai/industry/models/2026-W18/llm.json): Typed serialization of pulse 2026-W18.
- [Model Pulse · 2026-04-recap · JSON](https://brianletort.ai/industry/models/2026-04-recap/llm.json): Typed serialization of pulse 2026-04-recap.
- [Application Layer · 2026-W23 · JSON](https://brianletort.ai/industry/applications/2026-W23/llm.json): Typed serialization of application-layer issue 2026-W23.
- [Application Layer · 2026-W22 · JSON](https://brianletort.ai/industry/applications/2026-W22/llm.json): Typed serialization of application-layer issue 2026-W22.
- [Application Layer · 2026-W21 · JSON](https://brianletort.ai/industry/applications/2026-W21/llm.json): Typed serialization of application-layer issue 2026-W21.

## Tools

- [Claude Skill bundle](https://brianletort.ai/downloads/brianletort-industry-skill.zip): Drop-in skill (SKILL.md + commands + references) that knows the URL patterns and ships audience-translate, score-predictions, compare-weeks, tree-delta, and check-freshness commands.
- [Authoring playbook](https://brianletort.ai/industry): The Friday cadence and editorial rules (firewall, voice, scoring rubric) live in the repository at content/industry/AUTHORING.md.