E148: McCarthy ousted, border chaos, Cruise's robotaxi "accident" & more

Airtable has not gone public as of November 30, 2025, so the prediction about it having a “nice IPO” has not yet been testable.

Multiple up‑to‑date sources describe Airtable as still privately held and explicitly note that it is not listed on major exchanges:

• Forge’s FAQ: “No, Airtable is a privately held company and is currently not publicly traded on any stock markets including NYSE or NASDAQ.” (forgeglobal.com)
• Hiive, a secondary marketplace, likewise states that Airtable “remains a privately held company and cannot be accessed on major public exchanges such as NASDAQ or NYSE.” (hiive.com)
• PremierAlts’ 2025 valuation overview says, in its IPO FAQ, that Airtable has not announced plans for an initial public offering. (premieralts.com)
• Forge’s dedicated Airtable IPO page notes that Airtable “has not officially endorsed a plan to participate in an IPO” and that there is no IPO price yet. (forgeglobal.com)
• A 2024 overview from Cantech Letter similarly remarks that there is speculation but “no concrete evidence” of an Airtable IPO timeline, attributing some of the delay to a weak IPO window. (cantechletter.com)

Because (1) no IPO has occurred and (2) the original quote did not specify a hard deadline, we can’t yet say whether the claim that Airtable will have a well‑received IPO is right or wrong. It remains possible that Airtable could go public in the future and that the offering is either strong or disappointing. Therefore, the only defensible status today is inconclusive (too early).

The prediction is conditional: Chamath claims that if Congress stopped using omnibus/continuing resolutions and instead passed the 12 regular appropriations bills through the normal process, total federal spending would fall by roughly one‑third to one‑half.

As of late 2025, that prerequisite has not occurred:

• For FY2025, Congress funded the government with a full‑year continuing resolution (the Full-Year Continuing Appropriations and Extensions Act, 2025) that continued FY2024 funding levels for all 12 regular appropriations bills, rather than completing those 12 bills through the standard process. (congress.gov)
• CRS data show that omnibus and multi‑bill vehicles have remained a dominant feature of the appropriations process through at least FY2024; FY2006 was the last year when all regular appropriations were enacted individually. (everycrsreport.com)
• For FY2026, Congress again relied on continuing resolutions to keep the government open, rather than finishing all 12 standalone appropriations bills. (congress.gov)

Because the triggering reform (ending omnibus/CRs and moving fully to 12 regular bills) has not been implemented, we have no real‑world observation of what would happen to spending under that scenario. Available evidence only tells us what has happened under continued omnibus/CR use, not under the alternative process Chamath described. The scale of the asserted effect (a 33–50% reduction in total federal spending) is also not something that mainstream budget analyses have quantified for such a narrow process change, so we cannot reliably infer correctness from existing research.

Since the key condition has never been met and the claim is a large, untested counterfactual about what would happen, its truth or falsity cannot be empirically determined from post‑podcast data.

Chamath’s statement is explicitly conditional: “If it stops the CR process, it’ll be effective to the tune of about $500 billion.” He is not predicting that Congress will end omnibus/continuing‐resolution (CR) budgeting; he is asserting what the fiscal effect would be if that structural change occurred.

Since the podcast (Oct 2023), Congress has not ended the use of CRs or omnibus‑style catch‑all measures:

• For FY 2024, the government operated for months on a series of CRs before Congress finally finished funding via two large appropriations packages in March 2024.(en.wikipedia.org)
• For FY 2025, Congress went even further and pursued a full‑year continuing resolution (the Full-Year Continuing Appropriations and Extensions Act, 2025), extending prior funding levels with only modest adjustments.(en.wikipedia.org)
• Policy analysis and reporting in 2025 describe Washington as increasingly reliant on CRs—sometimes for an entire fiscal year—and warn that Congress may lean exclusively on CRs and omnibus/rescission packages rather than completing all 12 regular appropriations bills.(washingtonpost.com)
• Appropriations‑committee leaders repeatedly talk about working to return to regular order but simultaneously push new short‑term CRs such as the Continuing Appropriations and Extensions Act, 2026, indicating the CR process has clearly not been “stopped.”(appropriations.house.gov)

Because the antecedent of his conditional ("stops the CR process" and a true return to regular‑order appropriations) has not occurred, the world he is describing has never been observed. Available scored budget changes tied to actual legislation (for example, the full‑year CR for FY 2025 or the later “One Big Beautiful Bill” package) are in the tens of billions of dollars either way, and are driven by many policy changes beyond the procedural shift he is hypothesizing; none provide a clean empirical test of “ending CRs/omnibuses saves about $500 billion.”(crfb.org)

Given that:

• The key condition (ending the CR/omnibus process) has not been met; and
• There is no direct empirical or official scoring of the specific counterfactual he posits (regular‑order appropriations vs. CR/omnibus status quo) showing a ≈$500B effect,

we cannot say whether his conditional $500B figure is right or wrong. The claim’s truth value is therefore not determinable from real‑world outcomes, so the appropriate classification is “ambiguous.”

The prediction was made in October 2023 and is explicitly long‑horizon (“over the coming years”) and partly about a hard‑to‑measure outcome (long‑term innovation), so by late 2025 it is too early to score definitively.

1. Conditional setup (deficits, inflation, Fed cuts)
• The U.S. has not adopted major deficit‑reduction measures. CBO’s 2024–25 budget outlooks still project very large, persistent deficits (around 6% of GDP) and rising debt‑to‑GDP over the next decade, even after the 2023 Fiscal Responsibility Act.(crfb.org) A large GOP “megabill” enacted in July 2025 is estimated by CBO to increase deficits by about $4.1 trillion over 10 years.(politico.com)
• Inflation has come down substantially: average CPI inflation over 2025 is just under 3%, vs. much higher rates in 2022.(officialdata.org) Core PCE inflation is running in the high‑2% range as of late 2025.(wsj.com)
• The Fed has cut policy rates from a 5.50% peak in 2023 to a 3.75–4.00% target range by October 2025.(ycharts.com)
So the antecedent of his conditional prediction (no serious deficit action and later lower inflation and Fed cuts) is largely satisfied.

2. Long‑term interest rates and cost of capital
• The 10‑year Treasury yield is about 4.0% in late November 2025—down from ~5% highs in 2023, but still well above pre‑2020 norms and described in CBO‑based analysis as “elevated,” with an expected ~3.9–4.1% average 10‑year yield over 2025–34.(ycharts.com)
• Long‑term borrowing costs like the 30‑year mortgage rate remain around 6.2%, higher than typical 2010s levels, indicating a still‑high overall cost of capital compared with the prior decade.(apnews.com)
This is broadly consistent with his claim that long‑term rates would “remain high” even after inflation fell and the Fed began cutting, although rates have eased somewhat from their 2023 peaks.

3. Innovation and investment effects
The crucial final part of the prediction is that persistently high long‑term rates “are going to reduce long term innovation in the economy.” That is very difficult to verify by 2025:

• "Innovation" has no single, decisive metric; proxies (R&D, patents, startup formation, productivity) evolve slowly and are influenced by many non‑rate factors.
• Long‑term budget analyses (e.g., CBO and related commentary) emphasize that high debt and interest costs will tend to slow growth and investment over decades, but they do not claim an observable, realized collapse in innovation in the first couple of years.(crfb.org)

Given the multi‑year horizon implied by the quote and the inherently slow, noisy nature of innovation data, it is premature in 2025 to say whether U.S. long‑run innovation has actually been reduced in the way he predicted, even though the fiscal and interest‑rate backdrop is moving in the direction he described.

Bottom line: The fiscal and rate environment so far is broadly consistent with his mechanism (large deficits, lower inflation and Fed cuts, yet still‑elevated long‑term yields and cost of capital). But the core outcome—depressed long‑term innovation and investment “over the coming years”—cannot yet be clearly observed or falsified. Therefore the prediction is inconclusive at this point in time.

Public data for U.S. Customs and Border Protection (CBP) show that fiscal year (FY) 2023 did in fact set a new all‑time record for migrant encounters at the southern (U.S.–Mexico) border, higher than FY 2022.

• Multiple summaries of CBP data report 2,378,944 encounters at the Southwest border in FY 2022, which at the time was the highest number ever recorded there. (abc15.com)
• The same data series shows 2,475,669 encounters at the Southwest border in FY 2023, a new record and about a 4% increase over 2022. (newsweek.com) A House Homeland Security Committee factsheet and op‑ed using CBP numbers likewise note that FY 2023 Southwest border encounters were “more than 2.4 million,” exceeding the 2,378,944 recorded in FY 2022. (homeland.house.gov)
• The Migration Policy Institute’s 2024 statistical overview, based on CBP data, states that of more than 3.2 million total encounters at all U.S. borders in FY 2023, nearly 2.5 million were at the U.S.–Mexico border, calling this “the highest number of border encounters on record,” and explicitly noting it surpassed the nearly 2.4 million in FY 2022. (migrationpolicy.org)
• Some news outlets had earlier summarized FY 2022 as “about 2.7 million migrant encounters along the southern border,” matching the podcast’s rough “~2.7 million” reference, but those appear to be using a broader or differently labeled tally; in any case, later FY 2023 data still show higher totals than FY 2022 on those broader national measures as well. (goodmorningamerica.com)

Because FY 2023 southern‑border encounters clearly exceeded FY 2022’s totals under both the standard Southwest‑border series and broader nationwide-enforcement tallies, the core claim that FY 2023 would set a new all‑time annual record for migrant encounters at the U.S. southern border is supported by the data. The prediction is therefore right.

Joe Biden won the 2024 U.S. presidential election and was reelected president, defeating Donald Trump. Multiple major outlets report that Biden secured both the popular vote and the necessary Electoral College majority, confirming his victory.

Because the prediction was that the border wall issue "could cost him the election" (i.e., lead to him losing in 2024), and Biden in fact did not lose the 2024 election, the prediction is wrong.

(If one interpreted the statement more loosely as merely "this will be a major liability" without the explicit electoral-loss claim, that might be debated—but the quoted line is specifically about costing him the election, which did not happen.)

As of November 30, 2025, only about 2 years have passed since the 2023 prediction. The claim was that widespread, fully deployed autonomous robotaxi/self‑driving services are likely about 10 years away, i.e., around 2033. We are still well before that target date, so it is too early to determine whether the prediction will ultimately be correct or not. Therefore, the outcome is inconclusive (too early to judge).

As of November 30, 2025, the prediction is partly borne out so far, but the core "10+ years" timeline for the U.S. Northeast cannot yet be fully evaluated.

1. "It’s now" in Phoenix / California / Austin

• Waymo has operated a fully driverless, paid robotaxi service in metro Phoenix since 2020, and by 2023 had expanded it into what was then the largest contiguous paid robotaxi area in the world, including multiple suburbs and 24/7 airport service at Phoenix Sky Harbor. (waymo.com)
• By 2024–25, Waymo had opened citywide or large-area commercial robotaxi services in San Francisco, Los Angeles, and Austin, with anyone able to download the app and hail fully autonomous rides in those sunny, largely grid-based metros. (cnbc.com)
• Industry overviews in 2025 note that driverless robotaxis are a regular reality in cities like Phoenix, Austin, Los Angeles, and San Francisco, confirming that such deployments are indeed “viable now” in these climates and urban layouts. (axios.com)
  → This strongly supports Jason’s claim that high‑reliability robotaxi deployments are viable in dry, relatively simpler cities in the early–mid 2020s.

2. "10+ years out" for Boston / U.S. Northeast

• In Boston and the broader Northeast, there is still no fully driverless commercial robotaxi service. Waymo’s presence in Boston in 2025 is limited to short-term manual testing with human drivers, explicitly not operating in autonomous mode and not open to the public, and the city states this “does not represent the start of permanent operations.” (cambridgema.gov)
• Waymo’s own site lists Boston, New York, Minneapolis, etc. under “Driving experience in” / testing, while its commercial “Serving riders in” list includes only Phoenix, San Francisco Bay Area, Los Angeles, Austin, and Atlanta. (waymo.com)
• Boston officials and local reporting highlight steep political and technical resistance, citing non‑grid, narrow streets and harsh winter conditions, and note that Waymo has not yet validated fully driverless operation in snow and standing snow. (boston.com)
  → So far, reality is consistent with Jason’s view that the Northeast will be “last” and faces extra difficulty, but we are still years away from 2033, so we can’t yet confirm or falsify the specific “10+ years” horizon.

Conclusion

• The near‑term part of the prediction (robotaxis are already viable in Phoenix / parts of California / Austin) has clearly been vindicated by events since 2020.
• The long‑term part (high‑reliability deployments in Boston‑like, snowy, non‑grid cities are more than 10 years away from 2023) cannot be decisively judged before ~2033; we only know that such deployments have not happened yet and that substantial obstacles remain.

Because the prediction, as phrased, includes that long no‑earlier‑than‑2033 claim, its overall truth value is still too early to call, hence the rating "inconclusive" even though current evidence leans in the direction Jason described.

Evidence shows that Waymo did in fact launch a commercial, fully autonomous robotaxi service in Los Angeles within a relatively short time after October 2023.

Key timeline:

• In March 2024, the Los Angeles Times reported that Waymo One would begin offering paid, driverless taxi service in Los Angeles, following CPUC approval, initially with a small fleet and a waitlist of riders. (latimes.com)
• Waymo’s own blog later stated that it had begun commercial operations in LA earlier in 2024 and that riders had already taken hundreds of thousands of paid trips there. (waymo.com)
• On November 12, 2024, Waymo announced that Waymo One in Los Angeles was now open to all riders (no waitlist), offering fully autonomous rides 24/7 across roughly 80 square miles from Santa Monica through Hollywood and USC. (forbes.com)
• Waymo’s LA service page, updated in 2025, confirms that it is operating a fully autonomous, publicly available ride-hailing service in Los Angeles, alongside Phoenix and San Francisco. (waymo.com)

Given that the podcast prediction on 7 October 2023 was that Waymo would launch in LA ‘soon’ after already running 24/7 in San Francisco and Phoenix, and full commercial robotaxi operations began in LA within about 5–13 months (well within the implied ‘next few years’), this prediction clearly came true.

The prediction’s time window is 4–5 years from 2023, i.e., roughly 2027–2028. As of today (November 30, 2025), that deadline has not yet arrived, so we cannot definitively judge it.

Based on current public information, Tesla’s Full Self-Driving (FSD) is still officially classified as an SAE Level 2 driver-assistance system, meaning the human driver must supervise at all times and remain responsible for the driving task; Tesla itself warns that FSD does not make the car autonomous, and U.S. regulators also treat it as Level 2, not Level 4/5. However, whether Tesla will reach a credible Level 4/5 system by 2027–2028 is inherently about the future and cannot yet be confirmed or falsified.

Because the prediction’s target date is still in the future, the correct status is inconclusive (too early to tell).

Friedberg’s prediction is that the Las Vegas Sphere concept will lead to dozens of similar immersive venues worldwide within roughly a decade after 2023.

As of November 30, 2025:

• There is one operational Sphere in Las Vegas, opened in September 2023.(en.wikipedia.org)
• A second full‑scale Sphere in Abu Dhabi has been formally announced via a partnership between Sphere Entertainment and Abu Dhabi’s Department of Culture and Tourism, with Abu Dhabi confirmed as the second location.(investor.sphereentertainmentco.com)
• Earlier plans for a major London Sphere were withdrawn in January 2024 after planning permission issues.(en.wikipedia.org)
• Abu Dhabi’s deal gives it exclusive rights to develop additional Spheres across the Middle East and North Africa, but these are only options, not yet specific, funded projects.(thenationalnews.com)
• Sphere Entertainment is exploring smaller, cheaper “mini‑Spheres” and broader immersive deployments, but as of now no concrete sites or build schedules have been committed publicly.(theverge.com)
• Other immersive dome venues (e.g., Cosm in California and Texas) indicate a broader move toward Sphere‑like experiences, but they do not yet amount to “dozens” of venues clearly derived from the Las Vegas Sphere model.(apnews.com)

Only about two years have elapsed since the prediction, and his normalized horizon is roughly ten years after 2023 (through ~2033). With at least one confirmed additional Sphere and contractual frameworks for more, expansion is underway but still in an early phase and far short of “dozens” so far. However, there is no decisive evidence yet that such growth cannot occur over the remaining ~8 years.

Because the prediction’s timeline has not yet expired and current evidence neither guarantees nor rules out reaching “dozens” of Sphere‑like venues by around 2033, the outcome is too early to call.

As of November 30, 2025, there isn’t enough evidence yet to say whether this decade‑scale prediction is right or wrong.

What has happened so far

• The original Las Vegas Sphere opened in 2023 at a cost of about $2.3 billion, consistent with Friedberg’s reference point for the first unit’s capex. (theverge.com)
• Sphere Entertainment’s CEO James Dolan has told analysts the company is designing smaller “mini‑Spheres” with ~5,000 seats that would be “much cheaper” and faster to build than the Las Vegas original, directly matching the idea of scaled‑down, lower‑capex versions. However, no specific locations, budgets, or build timelines have been announced yet. (theverge.com)
• The only confirmed additional full‑scale Sphere so far is a planned venue in Abu Dhabi, structured as a licensed franchise. Abu Dhabi’s tourism authority will fund construction; the venue is expected to be similar in scale to Las Vegas, and there are ambitions for further Spheres in the Middle East and North Africa, but no other concrete sites have broken ground. (reviewjournal.com)
• Plans for a second large Sphere in London were formally withdrawn after political and planning opposition, removing one major prospective international location. (euronews.com)
• Other companies (not Sphere Entertainment) are indeed building smaller immersive dome venues—for example, Cosm’s 65,000–70,000 sq ft domed venues in Los Angeles (open) and Dallas/Atlanta (in progress)—which are clearly lower‑capex than a $2.3B mega‑arena, but these are still a handful of sites, not a global IMAX‑style network. (prnewswire.com)
• Analysts remain skeptical about Sphere’s profitability and scalability, noting high losses and slow, costly expansion, which suggests uncertainty about whether a large global rollout will ultimately happen. (barrons.com)

Why the prediction is still undecided

• Friedberg’s claim included two main components: (1) smaller, much cheaper Sphere‑like venues (on the order of a few hundred million dollars) would emerge, and (2) these would be rolled out widely worldwide over “the coming decade or so,” analogous to IMAX proliferation.
• By late 2025, the direction of travel is partially consistent with the first part (design work on mini‑Spheres and independent smaller domes like Cosm), but the scale and density of global deployment are nowhere near an IMAX‑like footprint yet—nor would we reasonably expect them to be only ~2 years into a 10‑year horizon.
• There is also no clear falsification: Sphere Entertainment is still pursuing a global network (starting with Abu Dhabi and MENA rights), and other immersive dome operators are expanding, so a broad rollout could still materialize later in the 2020s or early 2030s.

Given the explicitly long time frame (“over the coming decade or so”), and the fact that both success (substantial rollout) and failure (stalled expansion) remain plausible, the correct assessment as of 2025 is that the prediction’s ultimate accuracy is inconclusive rather than clearly right or wrong.