Let me understand the task. This is a "note-filing" system for a Mind about whether US universities grow or shrink due to AI over the next 20 years, with regional dependence (SEC growth vs. decline elsewhere).

The final instruction is a consolidation: "Revise the notes as a set: keep every claim that still holds, drop repetition, sharpen wording. Output one Markdown note that is the current best statement of what is established. Start with a # title, then short paragraphs, then: What would change this: … Keep sources or Unverified: tags. Do not mention tools, sessions, or that you are revising."

So I need to produce ONE comprehensive consolidated Markdown note that synthesizes all the durable claims from the filed notes. Let me compile the inventory carefully from all the notes.

Let me go through the notes systematically:

**Note 1 (deliberation on cost structure):**
- Personnel share ~70-75% structurally compressible for first time since 1945; Baumol framing
- Re-architect vs. bolt-on distinction: re-architectors cut cost-per-degree 30-50%; incumbents bolting AI on save almost nothing (tenure/legacy salaries amortize over decades)
- Tuition collapse comes from DEMAND side (wage premium erosion starting mid-2030s), not cost side — different from filed "competition from AI-native providers"
- By ~2040 bifurcated pricing regime: elite brands hold price (positional good), mid-tier tuition falls 30-60% toward cost-based floors (Unverified), tuition-dependent privates/small publics in NE/Midwest face pincer and merge/close
- Growth-state flagships (SEC footprint, TX, FL, AZ) get smooth landing

**Note 2 (deliberation on attrition timing):**
- Attrition timing as mechanism: AI savings realized only when vacated faculty lines are NOT refilled
- Same efficiency monetized two ways: headcount shrinkage in declining regions vs. growth capacity in growth regions
- Sector total size roughly flat while geography rotates south (this is about what AI cost efficiency does — doesn't shrink the sector, that's demography/credential value's job; it reallocates capacity)
- Declining-region systems shed 1-3% instructional headcount/year through 2030s (Unverified)

**Note 3 (deliberation on enrollment demand):**
- Demand-side sizing: AI attacks the rung the degree is priced against — entry-level white-collar work
- Premium bifurcates: holds for experienced credentialed workers and top-network graduates, narrows sharply at margin for non-selective institutions
- Willingness-to-pay collapses first at mid/low-selectivity tier, not T20s
- Net sector totals deceptively stable (~19M, drifting down only a few points by 2035) because damage is tiered and regional
- Key new framing: AI deflates credential price-justification per student faster than it deflates headcount; visible symptoms arrive first as tuition discounting (privates) and program-line closures (regionals), years before sector-wide shrink statistic moves
- Barbell, major-mix attack, substitute-credential wallet
- Net-system claim: system shrinks at the median while top decile grows — total national enrollment flat-to-down 10-20% outside flagships, capacity consolidating into larger flagship and online-first platforms; system barbell-izes, average school worse off than system statistic implies

Filed inventory (extensive):
- Demography: births 4.32M (2007) → 3.59M (2023); HS grads peak ~3.9M ~2025; birth dearth hits college age 2029-2037; NE/Midwest/Plains lose 10-20%; TX/FL/TN/GA/Carolinas grow; Sun Belt migration
- Credential value: screening vs insurance; licensed fields hold; mid-tier unlicensed BA premiums compress 30-50% real by mid-2030s (Unverified); "AI doesn't kill the degree; it kills the undifferentiated degree"; decay hits by major before institution; Burning Glass ~46% degree-requirement drops 2021-22; HBS "Dismissed by Degrees"; Canaries; enrollment falls 15-25% below demographic baseline by 2045 (Unverified); mid-market contracts 25-40% by ~2035 (Unverified); micro-credentials middle-market illusion; two brakes; elite degrees gain value BECAUSE verification; degree survives as insurance; demand migrates to "protected degree + AI skill stack"; "Bachelor's required" listings fall 30-50% (Unverified); statutory gatekeeping; licensure premium vs signaling premium; hiring decomposes by sector; AI-skills training not a university moat
- Program mix × region: licensed vs unlicensed share = dominant predictor