E86: Macro outlook: jobs, housing, inflation + Dutch farmers protests & EU climate missteps

Evidence through late 2025 shows that San Francisco’s SoMa/downtown core has not been successfully revitalized as a dense office/commuter district, and city leadership has largely shifted away from that goal.

Key points:

• Office vacancy in San Francisco’s downtown reached record levels around 36% by late 2023 and was still roughly one‑third vacant in 2025, far above pre‑pandemic levels and among the worst recoveries in North America. (hotair.com)
• As of Q2 2025, the downtown core’s total office vacancy rate was about 35.1%, with only a very slight quarter‑to‑quarter improvement—hardly consistent with a “successfully revitalized” dense office district. (en.wikipedia.org)
• Mayor London Breed herself acknowledged in 2023 that “San Francisco’s Downtown as we know it is not coming back,” and her roadmap focuses on tax breaks, rezoning, arts/entertainment, and attracting new industries rather than restoring the old commuter‑office model. (sfstandard.com)
• In Central SoMa—explicitly planned pre‑COVID for roughly 8 million square feet of new office space—a 2025 city study notes that the shift to remote work has stalled office projects, leaving about one‑third of existing office space vacant and concluding that new office development is “unlikely to be profitable.” The city is now proposing to waive commercial requirements to make more housing feasible, effectively retreating from the original dense‑office vision. (sfchronicle.com)
• City fiscal projections expect downtown office property values to remain far below pre‑pandemic levels and not fully recover until after 2034, reinforcing that a quick or straightforward office‑district revitalization has failed. (sfchronicle.com)

There are modest green shoots—AI firms leasing space, programs like Vacant to Vibrant, and some retail revival—but these amount to partial, mixed‑use adaptation rather than a restoration of SoMa/downtown as a thriving, dense office/commuter hub powered by the mayor’s 2022‑era plans. On the prediction as normalized (“it will not be successfully revitalized as that kind of district in the foreseeable future; current plans will fail”), the available evidence to date aligns with Jason’s view.

Multiple brokerage reports for Q4 2022 show San Francisco’s office market was far from 40% vacant or 30 million square feet empty by year‑end 2022:

• Transwestern reported the market closed 2022 at a record-high 23.7% vacancy, with 8.6 million sq. ft. of sublet space equal to 9.5% of total office stock. That implies total stock of about 90.5 million sq. ft. (8.6 ÷ 0.095). Applying the 23.7% vacancy rate to that stock yields roughly 21.5 million sq. ft. vacant, well below 30 million.

  Available sublet space totaled 8.6 million square feet, representing 9.5% of total office stock… vacancy of 23.7% in Q4 2022. (transwestern.com)

• CBRE’s official San Francisco Office Figures Q4 2022 put overall vacancy even higher, at 27.6%. Using the same ~90.5 million sq. ft. stock, that implies roughly 25 million sq. ft. vacant (0.276 × 90.5M), still materially below 30 million and well under 40% of inventory. (cbre.com)

• Cresa’s Q4 2022 San Francisco Office Occupier’s Guide likewise reports a 23.6% total vacancy rate and 31.9% total availability at year‑end 2022, consistent with vacancy in the mid‑20s percent, not 40%. (cresa.com)

Subsequent reporting shows San Francisco didn’t hit roughly 30 million sq. ft. of empty offices until later in 2023, when CBRE data cited by the San Francisco Chronicle and other outlets showed about 33–34% vacancy and nearly 30 million sq. ft. vacant—still short of 40% and occurring about a year after the prediction’s deadline. (sfchronicle.com)

So by December 31, 2022, the best available data indicate:

• Total office stock ≈ 90–91 million sq. ft., not 75 million.
• Vacancy ≈ 23–28%, not 40%.
• Vacant space ≈ 21–25 million sq. ft., not 30 million.

Because both the magnitude of vacant space and the vacancy percentage were substantially below what was predicted, and the total inventory figure was understated, the prediction is incorrect for its stated timeframe.

OPEC+ decided on June 2, 2022 to accelerate its planned output hikes, agreeing to add 648,000 barrels per day in both July and August 2022 instead of smaller monthly increases; this decision explicitly covered the August period Chamath was talking about and was reaffirmed at the June 30 meeting.(thenationalnews.com) As part of these higher group targets, Saudi Arabia’s official OPEC production quota for August 2022 was set at 11 million barrels per day.(statista.com)

According to OPEC’s own Monthly Oil Market Report (September 2022), Saudi Arabia’s crude oil production in August 2022 averaged 10,904 thousand barrels per day (10.904 mb/d), up from 10.744 mb/d in July and 10.559 mb/d in June.(scribd.com) An independent analysis using IEA/EIA data likewise cites Saudi output at about 10.9 mb/d in August 2022, contrasting it with a much lower level in mid‑2023.(tradingview.com) Broader 2022 statistics from the U.S. EIA and other analyses show Saudi crude production averaging about 10.4–10.6 mb/d for the year, consistent with Saudi output being near 10 mb/d before these mid‑2022 hikes and close to 11 mb/d afterward.(eia.gov)

Chamath’s claim was that by August OPEC+ would implement a capacity (i.e., production/quota) increase that would take Saudi Arabia from roughly 10 mb/d to about 11 mb/d. In practice, OPEC+ did approve the accelerated increases, Saudi’s official August quota was indeed 11 mb/d, and actual production that month was effectively 11 mb/d when rounded (10.9 mb/d). Given normal rounding and the informal context of the statement, the prediction matches what happened in both policy (quota) and realized output terms, so it is best judged as right.

This was an explicitly conditional forecast: extreme price spikes ($180–$380 Brent) were said to occur if Russia cut oil exports by 3–5 million barrels per day (mb/d) from early‑2022 levels and other producers did not quickly offset the loss.

What actually happened:

1. Russian exports never fell by 3–5 mb/d.

   • The IEA’s September 2022 Oil Market Report estimates Russian total oil exports at 7.6 mb/d, only about 0.39 mb/d below pre‑war levels, not 3–5 mb/d.(iea.org)
   • The December 2022 IEA report shows Russian exports at 8.1 mb/d, the highest since April 2022, with production just 0.2 mb/d below pre‑invasion levels.(iea.org)
   • Other analyses similarly find Russian crude exports remained around or above pre‑war volumes after initial disruptions, as flows were re‑routed to buyers such as China and India.(aa.com.tr)
     In other words, the sharp 3–5 mb/d export collapse assumed in the scenario never occurred.

2. Other producers did add supply.
   IEA reports from mid‑2022 onward note rising global supply from OPEC+ (especially Saudi Arabia and the UAE) and non‑OPEC producers like the US and Canada, which helped keep the market "comfortably supplied" rather than in the extreme undersupply assumed in the scenario.(iea.org)

3. Oil prices never reached $180, let alone $380.
   Brent crude peaked around $133–139/bbl in early March 2022 during the initial shock of Russia’s invasion, then fell back.(en.wikipedia.org)
   Annual data for 2022–2024 show Brent averages near $100 in 2022 and around $80–83 in 2023–2024, far below $180.(app.macrotrends.net)
   But since the prerequisite 3–5 mb/d export cut never occurred, prices never had a chance to test the scenario.

Because the key condition (a 3–5 mb/d Russian export cut without rapid offset elsewhere) did not happen, the forecast about what prices would be under that condition cannot be empirically judged as right or wrong. It remains a counterfactual scenario analysis rather than a testable prediction about the world that actually unfolded.

Evidence from official Saudi and Aramco statements shows that Saudi Arabia’s maximum sustainable oil production capacity was already around 12 million barrels per day (bpd) well before 2024, and the main capacity-expansion projects started years earlier, not in 2024.

Key facts:

1. 12 mbpd capacity existed before 2022

   • A March 11, 2020 Aramco press release says the Ministry of Energy directed Aramco to increase its maximum sustainable capacity (MSC) from 12 million bpd to 13 million bpd, which necessarily means 12 mbpd was already in place by 2020. (aramco.com)
   • In 2021 and May 2022, multiple reports state that Saudi Arabia’s current production capacity was about 12 million bpd, and that the plan was to boost this to more than 13 million bpd by 2026–2027. (thenationalnews.com)
     These sources contradict the premise that Saudi still had to build capacity to get from roughly 11 to 12 mbpd starting in 2024.

2. Expansion work to go beyond 12 mbpd began before 2024

   • Aramco outlined that the extra ~1 mbpd to reach 13 mbpd would come from major field projects (Berri, Marjan, Zuluf, plus others), describing incremental additions out to 2027—plans already active and being executed by 2022. (arabnews.jp)
     This again contradicts the claim that “they can only start the work in 2024.”

3. The 2024–2027 build-out Chamath described never materialized

   • In January 2024, Saudi Arabia ordered Aramco to halt the plan to raise capacity to 13 mbpd and to maintain MSC at 12 mbpd, effectively cancelling the further expansion. (al-monitor.com)
   • By October 2025, Aramco’s CEO was still describing the maximum sustainable capacity as 12 mbpd, which Aramco can sustain for a year, confirming no new step-up above 12 had been completed. (reuters.com)

Putting this together:

• The prediction specified that work to increase capacity from ~11 to 12 mbpd would only start in 2024 and finish in 2027.
• In reality, Saudi capacity was already ~12 mbpd by 2020, and the subsequent expansion projects toward 13 mbpd had started before 2024 and were then frozen in early 2024 rather than following the 2024–2027 build schedule Chamath described.

Because both the starting date (“can only start the work in 2024”) and the completion timing (“they’ll be done in 2027”) are inconsistent with what actually happened—and even the premise (that they still needed to build to 12 mbpd) is false—the prediction is wrong.

Evidence since mid‑2022 shows a clear global trend toward wider use of cap‑and‑trade and carbon taxes to price negative externalities in production and industry, matching Friedberg’s directional forecast.

• The World Bank’s State and Trends of Carbon Pricing reports show steady expansion of carbon pricing instruments: 73 instruments covering ~23% of global GHG emissions in 2023, 75 instruments covering ~24% in 2024 with US$104B in annual revenue, and 80 instruments covering ~28% of global emissions in economies representing about two‑thirds of global GDP by 2024–2025.(worldbank.org) These are overwhelmingly applied to power and industrial production (steel, cement, chemicals, etc.), aligning with his claim about production and industry.

• The EU has significantly strengthened and broadened its Emissions Trading System (ETS): reforms under the Fit for 55 package extend ETS coverage and tighten caps, add a new ETS II for buildings, road transport and additional fuel‑using sectors, and phase in carbon pricing for maritime transport.(consilium.europa.eu) The EU is also introducing the Carbon Border Adjustment Mechanism (CBAM), which effectively applies a carbon price on high‑emission imports such as cement, steel, electricity and fertilisers, directly internalising production externalities and raising the cost of high‑emission methods relative to cleaner technologies.(pwc.nl) Other jurisdictions are launching or expanding ETSs (e.g., China’s planned expansion of its ETS to steel, cement and aluminium, Vietnam’s new ETS for steel, cement and power, and Türkiye’s forthcoming national ETS), further supporting the global rollout he anticipated.(reuters.com) Revenues from these systems are increasingly earmarked for green investment, reinforcing the economic incentive to adopt low‑emission alternatives.(worldbank.org)

• In agriculture, progress is slower and politically contentious but moving in the predicted direction of using explicit pricing: Denmark has agreed the world’s first direct carbon tax on livestock emissions, to start in 2030 and rise by 2035, explicitly intended to cut agricultural GHGs and fund a green transition in the sector.(reuters.com) New Zealand, after initially preparing to price agricultural emissions via its ETS, removed agriculture from the scheme in 2024 but still plans a separate farm‑level pricing system by 2030, indicating not a rejection of the idea but a change in design and timing.(environment.govt.nz) Meanwhile, carbon pricing on fertiliser and other inputs via CBAM will effectively raise the cost of more carbon‑intensive agricultural inputs within the EU.(agriland.ie)

Taken together, these developments show that since the prediction in July 2022, governments have indeed been increasingly rolling out cap‑and‑trade and tax‑based systems to price negative externalities across production and industry, with early but real movement into agriculture. The mechanisms function as he described: they make high‑emission options more expensive and use revenues and complementary policies to support lower‑emission technologies. Although the long‑term end state is not yet reached and politics remain contentious, the trajectory so far closely matches the forecast, so the prediction is best classified as right.

Jason explicitly predicted on the podcast that “Bezos is gonna run for president in 2024,” framing it as the reason Bezos retired and was giving money away. (podscripts.co)

However, the official records and comprehensive overviews of the 2024 U.S. presidential election show no presidential campaign or formal candidacy by Jeff Bezos in that cycle. The main article on the 2024 United States presidential election lists the party nominees and major candidates; Bezos does not appear among Republican, Democratic, or independent/third‑party contenders. (en.wikipedia.org)

Biographical and news coverage of Bezos during the 2024 cycle describe him as Amazon founder, Blue Origin owner, and Washington Post owner, including detailed discussion of his controversial decision to stop the Post from endorsing a presidential candidate in 2024—but make no mention of any personal presidential run or FEC filing. (en.wikipedia.org) In subsequent years he is reported instead as launching and leading the AI startup Project Prometheus, again with no indication he ever ran for office. (en.wikipedia.org)

Since the 2024 election has concluded and there is no record of Bezos declaring or conducting a presidential campaign, the prediction that he would run for president in 2024 is wrong.

Public records and comprehensive summaries of the 2024 U.S. presidential election list the declared candidates from the major parties (e.g., Donald Trump, Kamala Harris, Joe Biden) and from third parties/independents; Jeff Bezos does not appear among any filed or nominated presidential candidates for 2024. (en.wikipedia.org)

Biographical coverage of Bezos through and after the 2024 cycle likewise describes him as Amazon founder, Blue Origin owner, and Washington Post owner, including his controversial decision regarding the Post’s 2024 endorsement and his later interactions with President‑elect Trump, but never as a 2024 presidential candidate or FEC‑registered contender. (en.wikipedia.org)

Given the election is completed and there is no evidence that Bezos ever formally declared a presidential candidacy or ran in any party primary or as an independent, Friedberg’s prediction that Jeff Bezos would not run for president in the 2024 cycle is right.

The prediction stated that the 2024 U.S. presidential general election would be contested between Ron DeSantis and Jeff Bezos as the nominees of the two major parties.

In reality, the 2024 general election was between Republican nominee Donald Trump and Democratic nominee Kamala Harris. Trump, with running mate JD Vance, was the Republican candidate, and Harris, with running mate Tim Walz, was the Democratic candidate. (en.wikipedia.org)

Ron DeSantis ran in the Republican primaries but dropped out after the Iowa caucuses and endorsed Trump; he did not become the Republican nominee. (en.wikipedia.org) Jeff Bezos did not run in any major party primary and was not a presidential nominee of either major party. No reputable election or campaign record lists him as a 2024 presidential candidate or nominee. (en.wikipedia.org)

Because neither DeSantis nor Bezos secured a major-party nomination and the general election was instead Trump vs. Harris, the podcast prediction is wrong.

The prediction was that the All-In podcast would continue releasing episodes at least through episode 100 without ending or permanently disbanding before that point.

Evidence:

• Episode index summaries for All-In with Chamath, Jason, Sacks & Friedberg show a continuous run from E86 on July 8, 2022 (“Macro outlook: jobs, housing, inflation + Dutch farmers protests & EU climate missteps”) through E87, E88, … up to E100 on October 14, 2022 (“Reflecting on the first 100 shows, fan questions, nuclear threat, markets, Amazon & more”), with no indication of a cancellation or disbanding in between. (allinpodcastsummary.com)
• A video/summary page for E100 confirms that this milestone episode was indeed produced and released by the All-In Podcast on October 14, 2022. (glasp.co)
• As of 2025, the podcast is still active: a directory page lists many episodes well past E140 and into mid‑2025, and a reference entry notes that the show has 240+ episodes, implying continued operation long after episode 100. (pods.ee)

Since the show clearly released episode 100 and remained intact at least through that point (and indeed far beyond it), Jason’s prediction that they would “make it to 100” without ending or permanently disbanding was correct.