AI Doom vs Boom, EA Cult Returns, BBB Upside, US Steel and Golden Votes

The prediction window is roughly 10 years from the podcast’s release on May 31, 2025, i.e., until around May 31, 2035. As of today (November 30, 2025), we are only about 6 months into that 10‑year period.

Current data show that the occupation of taxi/cab driver is still substantial:

• The U.S. Bureau of Labor Statistics reports about 204,000 taxi drivers in 2024 and actually projects 11% employment growth for taxi drivers from 2024 to 2034, not disappearance.(bls.gov)
• Broader counts that include taxi, shuttle drivers, and chauffeurs show hundreds of thousands of workers and projected overall growth of about 9% through 2034.(bls.gov)
• Industry statistics and business‑count datasets show thousands of taxi service businesses operating in the U.S. as of late 2025.(poidata.io)

Robotaxis are growing but from a very small base: estimates put the total number of active robotaxis in the U.S. at around 1,000 vehicles in early–mid 2025, versus well over 150,000–200,000 human taxi drivers.(eprnews.com) This demonstrates early disruption but nowhere near elimination of the occupation.

However, the prediction is explicitly about conditions around 2035, and we are still nearly a decade away. The present evidence (continued employment and even projected growth for taxi drivers through 2034) suggests the forecast may be overly pessimistic about the survival of the job, but it is still logically possible that policy, technology, and market dynamics over the next 9–10 years could change the trajectory.

Because the deadline has not arrived yet, the correct status is that it is too early to declare the prediction right or wrong.

The prediction has a two‑year horizon from the podcast date of 31 May 2025, i.e., until around 31 May 2027. Today is 30 November 2025, so we are only about six months into that two‑year window. It is therefore too early to say definitively whether the prediction will hold.

Empirically, as of 2025 the U.S. unemployment rate is in the 4–4.4% range, not remotely close to 20%:

• Monthly data for 2025 show unemployment at 4.0–4.4% from January through September 2025, with 4.2% in May (the podcast month). (theglobalstatistics.com)
• BLS‑based series and Fed/other summaries likewise report headline unemployment around 4.0–4.3% in early and mid‑2025. (economy.fedprimerate.com)
• Forecasts from professional forecasters and economists project unemployment in the mid‑single digits (around 4–4.5%) through 2026, not anywhere near 20%. (philadelphiafed.org)

So so far, the scenario of ~20% unemployment (overall, let alone explicitly “due to AI”) has not materialized. However, because the prediction explicitly concerns the state of unemployment two years after the podcast (by May 2027), we cannot yet confirm it as correct. A sharp unemployment spike between now and May 2027—whether or not attributed to AI—could still falsify the claim.

Given that the evaluation date precedes the end of the forecast window, the prediction’s ultimate accuracy is inconclusive (too early to tell).

According to the U.S. Bureau of Economic Analysis (BEA), real U.S. GDP in Q2 2025 (April–June) grew at an annualized rate of 3.8% in the third and final estimate. (bea.gov)

Chamath’s normalized prediction was that Q2 real GDP growth would be greater than 3.0% but less than 4.0% annualized. The realized 3.8% growth rate lies squarely within that 3–4% range and is clearly above 3.0% and below 4.0%.

While the advance estimate initially came in at 3.0% and was then revised to 3.3% and finally 3.8%, his prediction—interpreted as about the actual Q2 growth rate—matches the finalized BEA figure. (economics.td.com)

It’s too early to determine whether this prediction is right or wrong.

The claim is about what will happen by 2032: that the U.S. Social Security system will become functionally insolvent (e.g., trust fund depletion) and that this will in turn trigger higher taxes and/or inflationary policies shifting the burden to younger generations.

As of the current date (November 30, 2025), 2032 is still seven years in the future, and the key event in the prediction (reaching functional insolvency by that date and the specific policy response) has not yet occurred or failed to occur. While current projections by the Social Security Trustees and independent analysts often estimate OASI trust fund depletion around the early–mid 2030s, those are forecasts, not outcomes; we cannot yet say whether the system will in fact be functionally bankrupt in 2032 or whether the specific policy response (over-taxation and/or inflationary burden on younger generations) will happen in the way described.

Because the prediction is explicitly time-bound to 2032 and we are still in 2025, the correct classification is “inconclusive” (too early).

As of November 30, 2025, the key parts of Chamath’s prediction (“no commercially operating SMRs in the U.S. before roughly 2035” and “we’re still about a decade away”) concern outcomes in the 2030s, so they cannot yet be definitively judged.

Current facts:

• No small modular reactor is yet in commercial operation in the U.S. NuScale’s first-of-a-kind Carbon Free Power Project in Idaho, once targeted for 2029–2030 commercial operation, was terminated in 2023, pushing back what had been the leading candidate for an early U.S. SMR. (nuscalepower.com)
• Some U.S. projects are now aiming for the early 2030s, earlier than Chamath’s “2035+” claim if their schedules hold:
  • TVA’s BWRX‑300 SMR at Clinch River has a construction permit application under NRC review; with federal support, TVA has said commercial operation could be possible around 2033. (ans.org)
  • Dow and X‑energy’s Xe‑100 project in Seadrift, Texas, has submitted an NRC construction‑permit application and publicly describes a timeline of construction “later this decade” with startup “early next decade” (early 2030s). (corporate.dow.com)
  • Oklo continues to project its first Aurora SMR coming online around 2027–2028, which if achieved would be far earlier than 2035. (investors.com)
  • The U.S. Army’s Janus microreactor program targets an SMR‑class microreactor on a military base by about 2028, again implying potential sub‑2035 operation of small reactors, though in a defense rather than civilian‑utility context. (reuters.com)

These schedules challenge the pessimism of a strict “2035+ even if everything goes right” view, but they are still plans, not accomplished facts. Until at least one U.S. SMR actually reaches sustained commercial operation significantly before or after 2035, we cannot say with confidence that Chamath’s timeline is either correct or clearly wrong.

Because the forecasted period (early–mid 2030s) has not yet arrived and no U.S. commercial SMR is operating, the prediction is too early to conclusively verify or falsify.

It is too early to verify this prediction.

Chamath’s claim on 31 May 2025 was essentially: Starting from now, the fastest you could bring a newly initiated, unplanned U.S. natural-gas plant into commercial operation is about four years — i.e., sometime in 2029 or later.

To test that literally, we would need to observe at least one gas plant whose development began after the prediction (mid‑2025 or later) and then see whether it does or does not start operating before four years have passed (before mid‑2029). As of the current date (30 November 2025), no such project could possibly have reached commercial operation yet, because fewer than two years have elapsed since the prediction.

Available data on U.S. natural-gas plant development timelines (e.g., typical 2–3 years of construction plus often‑lengthy pre‑construction permitting and interconnection processes) describe historical averages, not realized outcomes for projects started after May 2025. Those historical figures don’t let us conclusively validate or falsify a forward‑looking lower bound about future projects started in 2025, especially given potential changes in regulation, supply chains, and grid interconnection queues after that date.

Because the four‑year window from the 2025 starting point (extending to 2029) has not yet elapsed, the prediction’s truth value cannot yet be observed in real-world outcomes, so it remains inconclusive rather than clearly right or wrong.

Chamath claimed that for the three mothballed but restartable U.S. nuclear reactors, the earliest realistic restart dates would be in the 2027–2030 range.

Evidence since then shows this is false:

• Palisades (Michigan), a shuttered commercial reactor regarded as the first U.S. attempt to restart a decommissioned plant, secured a $1.5+ billion DOE loan and state support with a publicly stated target of late 2025 for restart. This target was already on record well before and around the time of the May 31, 2025 podcast.(greatlakesnow.org)
• NRC approvals in July 2025 allowed Holtec to begin preparing Palisades for restart and loading fuel, with reporting that Holtec aimed to bring the plant back online in the fourth quarter of 2025.(reuters.com)
• Palisades’ status changed from decommissioned to online on August 27, 2025, enabling the plant to receive fuel and return to power production in 2025, not 2027 or later.(en.wikipedia.org)

For the other two prominent restart projects:

• Three Mile Island Unit 1 (Pennsylvania) is being financed with a DOE loan and is targeted to resume operations around 2027, matching the lower bound of Chamath’s window.(apnews.com)
• Duane Arnold Energy Center (Iowa), which shut in 2020, is now slated to be revived by 2029 under a NextEra–Google deal.(ft.com)

However, Chamath’s statement was about the earliest realistic restart for the three restartable mothballed reactors being 2027–2030. Because Palisades has in fact restarted (or is effectively back online) in 2025, and that 2025 timeline was already the working plan at the time of his prediction, the claim that 2027 was the earliest realistic restart is not borne out.

Therefore, the prediction is wrong.