Let me understand the task. I'm a "Mind" that files notes. The current facet is Regional variation (SEC growth vs. northeast/rust belt decline). The substance from this session gives a durable claim on the facet: the bifurcation is the sum of three stacked currents — demographic cliff falling hardest on Northeast/Midwest, domestic migration relocating college-relevant households to Sun Belt/SEC footprint, and AI disruption landing on white-collar task clusters concentrated in Northeast/Rust Belt metros. Plus the flagship-insulation point (Michigan, UVa, Wisconsin partially insulated — shrinkage is region- AND stratum-specific).

Instructions:
- Write one Markdown note: a single dated claim this Mind can stand behind.
- # title = claim in one line
- 1-3 short paragraphs
- End with: What would change this: …
- Name a source for any load-bearing number or fact, or start that sentence with Unverified:
- Do not mention tools, searches, sessions, passes, or note-writing.
- Do not invent evidence.
- Output only the note. No tools, no XML.

Existing filed notes relevant to this facet: Note 4 ("AI bifurcates US university enrollment by region — 'grow or shrink' resolves as 'both'"). The new claim must be consistent and add the mechanism: three stacked currents, plus flagship stratum insulation.

Load-bearing facts I can source:
- Demographic cliff: CDC/NCHS birth data — births fell from ~4.32M (2007) to ~3.59M (2023), ~17%. WICHE "Knocking at the College Door" projects HS grads peak ~2025 (~3.9M/yr) then decline; Northeast/Midwest cohorts shrink fastest. These were cited in prior notes and are real sources.
- Net domestic migration to Sun Belt: US Census Bureau population estimates show sustained net domestic migration to Texas, Florida, Georgia, Tennessee, the Carolinas. Real.
- AI white-collar exposure: Stanford Digital Economy Lab "Canaries in the Coal Mine" (2025) found early-career employment declines in AI-exposed occupations. Real (Brynjolfsson et al. 2025). Used in prior notes.

Numbers to keep: 2025–2045 horizon. Maybe keep ~17% birth decline. Keep the "region- AND stratum-specific" point.

Title should be the claim in one line, dated. Something like:

# Through 2045 the AI-era split in US higher education runs region AND stratum: the SEC/Sunbelt gains while the Northeast/Rust Belt shrinks — because three currents stack (demographic cliff, domestic migration, white-collar AI exposure), and flagship brands everywhere are partially insulated

"# The SEC/Sunbelt grows and the Northeast/Rust Belt shrinks through 2045 because three currents stack — cohort decline, domestic migration, and AI hitting white-collar metros hardest — but flagship brands everywhere are partially insulated: the split is region- and stratum-specific, not uniform"

Add a date: "(assessment 2026)" or "filed 2025". Prior notes used "(filed 2025...)" and "(assessment 2026)". I'll use "(assessment 2026)" or just no date... The instruction says "a single dated claim" — so include a date. I'll do "(assessment 2026)" — hmm, what's "now"? Notes reference 2025 and 2026. The most recent notes said "(assessment 2026)". I'll use that. Actually one note said "filed 2025". Safe: "(assessment 2026)".

Structure:
- Title
- Para 1: the three stacked currents with sources.
- Para 2: net effect + flagship insulation (region- and stratum-specific).
- Para 3: maybe fold into 2. Keep 2-3 short paragraphs.
- "What would change this: ..." line with falsifiers.

Falsifiers:
- If Sunbelt college-age cohorts don't grow as projected / migration reverses (Census data would show).
- If AI-driven white-collar disruption hits Sunbelt metros equally (remote work decoupling task location from degree location).
- If flagship brands in shrinking regions DON'T hold enrollment (i.e., Northeast flagships like Michigan... wait Michigan is Midwest flagship — example of insulated brand) — if Northeast/Midwest flagships lose enrollment in line with their regions, the stratum insulation is wrong.
- Uniform national decline would falsify the whole bifurcation.

Also note: what would change this should include the reverse — if regional convergence happens (SEC growth stalls while Northeast flagships hold), the stacked-currents picture weakens.

Title: "# The grow/shrink answer for US universities through 2045 is regional and stratified: the SEC/Sunbelt gains seats, pricing power, and state capital while the Northeast/Rust Belt shifts