As of late 2025, parts of Chamath’s forecast clearly happened, while others were only partially or briefly true, so the overall bundle is mixed rather than cleanly right or wrong.

1. Applications

• Harvard’s applications fell 5% for the Class of 2028 (54,008 vs. 56,937), after a 17% drop in early-action apps, and then dropped another ~11% for the Class of 2029 (47,893 vs. 54,008). (harvardmagazine.com)
• Brown’s pool dipped ~5% for the Class of 2028 (48,881 apps) and then, along with Dartmouth and Yale, saw roughly 10–12.5% application declines for the Class of 2029, largely after reintroducing test requirements. (browndailyherald.com)
• But other Ivies saw record or near‑record demand: Yale’s Class of 2028 application pool was the largest in its history (~57,500), Penn received its largest-ever pools (about 65,000 for 2028 and 72,000 for 2029), and Princeton’s Class of 2029 pool (42,303) was its biggest ever; Columbia’s applications hovered near historic highs around 60,000. (yalecollege.yale.edu)
  Net effect: several Ivies (notably Harvard, Brown, Dartmouth, Yale) did see material declines from recent peaks, but others hit records; there is no uniform across‑the‑board collapse in Ivy applications.

2. Private contributions/donations

• At Harvard, total philanthropic contributions fell by about $151 million (≈14%) in FY 2024 vs. 2023 amid donor backlash over campus controversies, even as current-use gifts rose modestly; some other schools like Columbia and Penn also saw notable drops in specific giving streams (e.g., Columbia Giving Day down 29%, Penn Fund down 21%). (thecrimson.com)
• However, by FY 2025 Harvard reported record current-use giving ($629 million, up 19% year‑over‑year) and overall gifts rising from $1.17 billion to about $1.3 billion, the highest in its history, despite political headwinds. (thecrimson.com)
  So there was a pronounced but short‑lived downturn in donations at some Ivies in 2024, followed by at least one very strong rebound; available data do not show a clear, continuing downward trajectory in Ivy philanthropy “from 2024 onward.”

3. Government funding

• This part of the forecast clearly materialized. Beginning in 2025, the Trump administration froze or slashed large amounts of federal funding to several Ivy League universities, especially targeting research grants (e.g., more than $2.2 billion in federal funding frozen at Harvard, with later additional cuts bringing total threatened or cut grants into the multi‑billion‑dollar range; Columbia hit with roughly $400 million in research cuts and a $221 million settlement; Brown facing a freeze on about $510 million in NIH grants). (washingtonpost.com)
• NIH-wide grant terminations and caps on indirect cost reimbursement also reduced federal research support at many institutions, with Ivy League schools singled out rhetorically and substantively for especially broad cuts. (harvardmagazine.com)
  This is a substantial “ratcheting down” of government spending on these schools, consistent with his prediction.

4. Attempts to remove nonprofit / tax-exempt status

• In 2025, President Trump repeatedly vowed publicly to take away Harvard’s tax‑exempt status and his administration formally asked the IRS to revoke it—moves that legal experts and Harvard officials criticized as likely unlawful but which nonetheless represent an organized political attempt to challenge its nonprofit status. (washingtonpost.com)
  No Ivy has actually lost 501(c)(3) status, but the prediction only required that “some folks” try, which clearly occurred.

Overall:

• The government funding and nonprofit‑status attack components have clearly come true in a significant way.
• The applications and donations components show partial, selective, and in some cases temporary declines rather than a uniform, sustained downturn “across the Ivies.”

Because major parts of the prediction are accurate while others are contradicted by the data (and we are no longer in a too-early window), the combined forecast is best characterized as mixed rather than simply right or wrong, hence marked here as ambiguous.