The Mind's overall question: Do universities grow or shrink due to AI? Current facet: **Cost structure — AI reduces need for faculty/TA, tuition pressure**.

Looking at the structure of this conversation: each user message presents a facet and a session substance, and I (as the note-filer) output one Markdown note. The current message (Note 8's message) is the final synthesis request: "The user wants a single consolidated Markdown note that synthesizes everything."

Wait — let me re-read. The last user message content is in "Note 8" of the human turn:

"The user wants a single consolidated Markdown note that synthesizes everything: the current best statement of what is established about whether US universities grow or shrink due to AI over the next 20 years, with regional dependence. This is the final synthesis note."

Then it lists:
- Format requirements (# title one line, short paragraphs, "What would change this: …", sources/Unverified tags, no tool/session/revision mentions)
- The durable claims inventory (core answer, demography, signaling, program mix × region, regional)
- "Substance from this session (write a note only if this adds a durable claim):" — listing:
  - Six independent cost-structure claims batch + cost-structure claim 7 (spending redirects, not shrinks)
  - Research-function wind-down lines
  - Enrollment-facet claim (barbell)
  - By-major-before-by-institution claim
  - Screening vs insurance signaling decomposition
  - The big cost-structure claim (pricing rule breaks, $6–12k convergence, unbundling)
  - Cost-structure corollary (brand-barred savings, credential trusts)
  - Testing the premise (instruction is minority slice — Delta Cost/NCES data: instruction ~25–30% at publics, ~30% privates, 35–40% CCs; admin ~13–15%)
  - The tuition-deflation channel (residential bundle defection, plant costs $8–15k/student/yr, regional sorting of plant utilization)

So the final synthesis note. The standing format from the last instruction block: "# title, then short paragraphs, then: What would change this: … Keep sources or Unverified: tags. Do not mention tools, sessions, or that you are revising."

Note 8's analysis says the synthesis can use "a modest number of short paragraphs... maybe 4-7 short paragraphs. But I should keep it tight."

So my task: write ONE Markdown note that is the grand synthesis — the current best statement of what is established across all facets: core answer, demography, signaling/enrollment demand, program mix, cost structure, research function, regional variation.

Let me inventory all durable claims to include:

**Core answer:**
- "Grow or shrink" resolves as "both" — the system thins and bifurcates; region- and stratum-specific.
- Total enrollment drifts ~19M → ~14–15M by 2045 (Unverified).
- Degrees conferred roughly steady while institution count falls — "thins, doesn't shrink."
- Median campus roughly halves in enrollment.
- AI is an accelerant/sorter of existing trends, not the primary cause. Two-front squeeze: fewer 18-year-olds AND weakened degree signal.

**Demography (load-bearing):**
- Births 4.32M (2007) → ~3.59M (2023), ~17% decline (CDC/NCHS).
- HS grads peak ~3.9M/yr ~2025 (WICHE "Knocking at the College Door").
- Birth dearth hits college age 2029–2037.
- Northeast/Midwest/Plains lose 10–20% college-age cohorts (WICHE); TX/FL/TN/GA/Carolinas grow.
- Net domestic migration to Sun Belt (Census Bureau).

**Signaling:**
- Screening vs insurance decomposition: screening dies first (AI-verified work samples cheaper than proxy-filtering); insurance (blame-shifting/liability cover) is sticky residue, persists in regulated/litigation-exposed sectors through the 2040s.
- Licensed/statute-gated fields hold (nursing, law, accounting, teaching; catastrophic-failure: engineering, medicine).
- Mid-tier unlicensed BA premiums compress 30–50% real by mid-2030s (Unverified).
- Fastest collapse in generalist white-collar fields (business, communications, generic liberal arts).
- By ~2035 median degree = positional luxury