Evidence from late‑2023 Treasury data shows that while the U.S. Treasury did issue roughly the amount Chamath was talking about, the demand characteristics were the opposite of “the easiest $2 trillion” with a “line out the door.”

1. The ~$2T in extra borrowing did occur over the relevant window.
Treasury’s own borrowing estimates show:

• July–September 2023: Treasury borrowed $1.010 trillion in privately‑held net marketable debt.
• October–December 2023: Treasury borrowed $776 billion in privately‑held net marketable debt. That is about $1.786T in H2 2023 alone, and Treasury expected another $760–816B in January–March 2024, pushing new borrowing well over $2T across the two following quarters he was discussing. (home.treasury.gov) All of this debt was indeed sold; there was no failed auction.

2. But demand was not “very strong” – it deteriorated sharply as supply rose.
The Federal Reserve’s post‑mortem on the “Treasury tantrum of 2023” documents that in the second half of 2023 the 10‑year yield jumped from below 4% to above 5%, with the rise driven largely by a higher term premium caused by greater Treasury issuance, quantitative tightening, and heightened uncertainty—i.e., investors demanded more compensation to hold the growing supply of long‑term Treasuries. (federalreserve.gov) That is inconsistent with an environment where demand is so strong that extra supply is absorbed easily at prevailing yields.

An analytical review of this period notes that by late 2023 U.S. bond auctions were “flashing warning signs,” with investors “showing less enthusiasm for U.S. debt issuance, especially at longer maturities.” It highlights that a November 2023 30‑year Treasury auction had weak demand and helped push the 30‑year yield briefly above 5%, reviving fears of a UK‑style gilt shock. (benmaccorquodale.com)

3. Key long‑bond auctions in the period were notably weak, not over‑subscribed.
The November 9, 2023 30‑year auction was widely described as one of the ugliest on record:

• $24B of 30‑year bonds were sold at a 4.769% yield, versus a 4.716% when‑issued level, a 5.3 bp tail – the largest on record at the time.
• Bid‑to‑cover was 2.24 vs a 2.44 six‑auction average, and
• Primary dealers were forced to take 24.7% of the issue vs a 12.7% six‑auction average and 18.2% in the prior auction, indicating that end‑investor demand was unusually weak and dealers had to absorb far more supply. (forexfactory.com)

A contemporaneous Q&A on Treasury auctions describes the November 9, 2023 30‑year sale as having “weak” demand, with primary dealers “tak[ing] on more than they wanted” and both long‑term Treasuries and stocks falling more than usual that day. It frames this as a warning sign that long‑term Treasuries “may be harder to sell” amid persistent deficits and elevated inflation. (finregrag.com)

These are exactly the opposite of “line out the door” conditions.

4. Bottom line for the prediction.

• On the narrow question “Will the bonds sell?”, Chamath was effectively right: Treasury successfully borrowed around the planned amounts; there was no outright failure to place the debt. (mondovisione.com)
• But his actual claim was much stronger: that this extra ~$2T would be the “easiest” debt to sell with very strong demand.

In reality, the period was characterized by a Treasury tantrum, historically large jumps in yields driven by higher term premia tied to increased issuance, and at least one record‑bad 30‑year auction with very weak demand and heavy dealer takedown. (federalreserve.gov) That contradicts the idea of exceptionally strong, effortless demand.

Given that the core of the prediction was about how strong and easy demand would be—not just that auctions would technically clear—the evidence points to the prediction being wrong.