AI Bubble Pops, Zuck Freezes Hiring, Newsom's 2028 Surge, Russia/Ukraine Endgame

As of November 30, 2025, the prediction covers a multi‑year horizon (“over the next several years,” normalized here as through at least 2028), so it is too early to determine whether the AI boom will persist without a sustained bust.

Evidence so far (late 2025) is broadly consistent with Sacks’s current view that we are in an AI investment boom/supercycle:

• Major analysts and banks still describe AI as driving a structural "supercycle" in markets and earnings, not a completed bubble and crash. (reuters.com)
• The Wikipedia “AI bubble” overview notes enormous and growing AI spending—US mega‑cap AI investment expected to exceed $1.1 trillion between 2026 and 2029 and total AI spending to surpass $1.6 trillion—alongside record valuations like Nvidia briefly reaching a $4 trillion market cap in 2025. This describes an ongoing boom, not a post‑bubble bust. (en.wikipedia.org)
• Forecasts from Dell’Oro and others project global data‑center capex—now heavily driven by AI accelerators—to grow at ~21% CAGR to around $1.2 trillion by 2029, with hyperscalers (Amazon, Microsoft, Google, Meta) responsible for roughly half of that spending. They explicitly expect strong AI‑related investment through the second half of the decade, even if there may be a short‑term slowdown around 2026. (datacenterdynamics.com)
• Recent reporting shows big tech continuing to raise AI and data‑center capex guidance for 2025–2026—individual firms planning tens to over a hundred billion dollars each, with aggregate AI infrastructure spending well into the hundreds of billions and characterized as an ongoing “AI capex frenzy.” (investorplace.com)
• Coverage in outlets like the Washington Post frames AI data‑center and software spending as a key force “propping up” the U.S. economy in late 2025, with Google, Meta, and Microsoft all signaling further increases in AI‑related capex into 2026. (washingtonpost.com)

At the same time, many analysts and commentators argue that parts of the AI market are in a bubble and predict a correction or “burst” in late 2025 or 2026, though this remains forward‑looking opinion rather than an observed, sustained bust at this point. (forbes.com)

Since:

• The boom is clearly still ongoing in late 2025 (so the short‑term component of his view hasn’t been falsified), but
• The core of the prediction is explicitly multi‑year (through at least 2028), and future investment cycles, valuations, and commercial outcomes between 2026–2028 are unknown,

there is not yet enough elapsed time to judge whether AI will "broadly remain in a boom phase rather than entering a sustained bust" over the full period Sacks described. Therefore, the correct classification today is inconclusive (too early).

Sacks’ claim is about the “next several years” of AI development following a sustained supercycle of investment and progress rather than a short hype spike followed by collapse. As of November 30, 2025, only a bit over three months have passed since the August 22, 2025 episode—far short of “several years,” so the prediction’s full time horizon hasn’t elapsed.

Available data so far show:

• AI and especially generative AI funding hit record highs in 2024 and continued rising into 2025, with generative AI companies raising about $56B in 2024 and even larger totals projected for 2025. (techcrunch.com)
• In early and mid‑2025, AI captured a majority of global VC funding (over 50–60% of all venture capital), with large late‑stage “mega‑rounds” and massive capex by big tech firms (Google, Microsoft, Meta, Amazon) on AI infrastructure. (faf.ae)
• Reports throughout 2025 describe an ongoing AI investment boom or “gold rush,” alongside warnings of possible bubble dynamics and future corrections, but no clear evidence of a broad collapse by late 2025. (fortune.com)

These facts are consistent with the early phase of a supercycle rather than a popped bubble, but they do not settle the multi‑year claim Sacks made. A sharp correction or collapse could still occur after November 2025. Since the prediction specifically concerns the trajectory over the next several years, and that period has not yet played out, it is too early to say whether he was right or wrong.

Therefore the correct classification as of November 30, 2025 is: inconclusive (too early to tell).

The prediction is about a negative event over a multi‑year window: that “over the next few years (through at least 2028)” we won’t see large AI models enter a rapid, recursively self‑improving loop that yields a single runaway superintelligent AGI far ahead of all others, and that instead progress will remain incremental and competitive across multiple providers.

As of November 30, 2025:

• There is no credible public evidence of any AI system undergoing hard‑takeoff‑style recursive self‑improvement to become a runaway superintelligent AGI. Major labs (OpenAI, Anthropic, Google DeepMind, Meta, xAI, etc.) continue to report improvements via larger models, better training data, new architectures, and tools, but not via autonomous self‑modification that rapidly explodes capabilities beyond all others.
• The AI landscape remains competitive and multi‑polar: several labs provide cutting‑edge frontier models (e.g., GPT‑4‑class and successors, Claude‑class models, Gemini‑class models, etc.), and none is acknowledged as a qualitatively “runaway superintelligence” far beyond the rest. This matches the spirit of the prediction so far.

However, the prediction explicitly extends through at least 2028, and we are only at 2025‑11‑30. It could still be falsified by events in 2026–2028. Because the claim is “this won’t happen in the coming years,” you can’t call it definitively right before the end of the specified window.

So:

• So far (2025), observable reality is consistent with Sacks’s prediction.
• But it’s too early to say whether it holds for the entire period through 2028.

Therefore the correct classification today is "inconclusive" (too early to fully judge), even though current evidence aligns with the prediction.

The prediction’s timeframe is “within roughly the next few years (by around 2029).” As of November 30, 2025, we are only about three months past the podcast release (August 22, 2025), so it is too early to judge whether this 2029‑ish claim will ultimately be correct.

What we can say so far:

• There is a visible industry trend toward more modular and specialized approaches, e.g.:
  • Mixture‑of‑experts (MoE) architectures and router‑based systems in both research and production settings.
  • Specialized / small language models (often called SLMs or task‑specific models) such as Microsoft’s Phi series and many on‑device models targeting phones and edge devices, marketed as lower‑cost and more efficient per token than frontier LLMs.
• There is also strong economic and regulatory pressure to reduce energy use and cost per token in large‑scale AI deployments (e.g., data‑center power constraints, GPU scarcity, and cloud cost concerns discussed widely in AI industry reporting in 2024–2025).

However, the prediction contains three separate long‑range components, each aimed at around 2029:

1. A significant transition of AI toward architectures built from networks of smaller, specialized models.
2. Achieving roughly 10x–100x reductions in energy and dollar cost per token relative to 2025 levels.
3. A resulting substantial increase in total token production versus 2025.

Current public information (as of late 2025) is not sufficient to:

• Confirm that SLM‑network architectures will dominate over large monolithic frontier models by ~2029.
• Verify an achieved 10x–100x aggregate reduction in cost/energy per token versus 2025 baselines, at production scale across the industry.
• Measure industry‑wide token production in 2029 relative to 2025.

Because the target year (≈2029) has not yet arrived, and the necessary market‑wide metrics are not yet observable, the prediction cannot be definitively evaluated.

Therefore the correct status as of November 30, 2025 is: inconclusive (too early to tell).

As of November 30, 2025, only a few years have elapsed since the start of the current generative‑AI surge (late 2022–2023) and only a few months since Chamath’s August 2025 statement. His claim is explicitly about a “multi‑decade innovation cycle” (20+ years beyond 2025), so by definition it cannot yet be confirmed or definitively falsified.

Available evidence does show that the AI wave is still very active rather than clearly having peaked and ended:

• Major AI companies remain among the world’s most valuable firms; for example, Alphabet’s market cap is nearing $4 trillion in late 2025, driven largely by AI products and infrastructure, following Nvidia’s earlier move into the $4T tier, both underpinned by heavy AI investment and demand. (reuters.com)
• Reports on an “AI bubble” note corrections and fears (e.g., Nvidia’s sharp but temporary stock drop in early 2025 and widespread debate about ROI), yet they also highlight continued or even accelerating AI spending: global AI investment is projected in the trillions of dollars over the next several years, treating AI as a foundational technology rather than a short‑lived fad. (en.wikipedia.org)
• Large international AI summits and conferences (e.g., the 2025 AI Action Summit in Paris and the 2025 World Artificial Intelligence Conference in Shanghai) show sustained, system‑level commitment by governments and industry to long‑term AI development. (en.wikipedia.org)

However, these facts only show that the cycle is still underway in 2025; they do not prove it will persist for “multiple decades.” Likewise, while there is concern about bubbles and overvaluation, no decisive collapse has occurred that would prove the cycle will not last that long. Because the horizon of the prediction (20+ years beyond 2025) extends far into the future, there is simply not enough elapsed time to judge its ultimate accuracy.

Given that the key claim is about long‑term duration, and neither confirmation nor clear falsification is yet possible, the only defensible assessment as of 2025 is inconclusive (too early to tell).

The prediction is explicitly about conditions through at least the late 2020s (i.e., out to around 2028–2029). As of today, November 30, 2025, fewer than three years of that period have elapsed, so we cannot yet evaluate whether AI startup funding, corporate capex, and strategic M&A will remain elevated throughout the stated horizon.

We can already say that:

• AI investment and capex remained very high in 2024–2025 (e.g., continued multi‑billion‑dollar data center and GPU spend by hyperscalers, strong AI funding rounds, and active AI-related M&A), which is consistent with the near‑term part of the claim.
• However, the core of the prediction is that this elevated activity will not collapse like a burst bubble through the late 2020s, which cannot be confirmed or falsified until that time has actually passed.

Because the main time window of the forecast is still in the future, the only defensible evaluation today is that it is too early to tell whether the prediction ultimately proves correct.

The prediction gives OpenAI roughly four years from August 2025—i.e., until around August 2029—to reach ~2 billion daily active users, about one‑tenth of Facebook/Meta’s then‑current revenue, and a valuation on the order of $1.5 trillion. As of the current date (November 30, 2025), that deadline is still almost four years away, so it is too early to know if this target will be met.

Available data only show partial progress so far: estimates and CEO statements put ChatGPT at around 800 million weekly active users in 2025, with daily actives on the order of 120–180 million—well below 2 billion DAU.【1search0】【1search2】【1search3】 OpenAI is valued in the low hundreds of billions as of 2025 (roughly $300–500 billion, depending on the private-markets quote), not yet near $1.5 trillion.【1search0】【1search2】【1search1】 Internal forecasts cited in the press project $125 billion in annual revenue by 2029 and billions of users by 2030, but those are plans rather than outcomes.【1search5】【1search7】 For comparison, Meta’s 2024 revenue is about $164.5 billion, implying that one‑tenth of Facebook/Meta revenue is on the order of $16–19 billion per year.【2search3】

Because the forecast horizon (through ~August 2029) has not yet passed, we cannot definitively say whether Chamath’s prediction is right or wrong at this time; it remains unresolved.

The prediction concerns the late 2020s and explicitly ties its timing to the 2028 U.S. election cycle. As of November 30, 2025, that election cycle has not yet begun in earnest, and there is still substantial time for party dynamics and leading figures to change.

While Alexandria Ocasio-Cortez and other democratic-socialist–aligned figures are nationally prominent and influential within the Democratic Party, it is too early to determine whether a clearly socialist-aligned politician (such as AOC) will, by the 2028 cycle, become the major national standard-bearer or leading candidate of the party. The key condition of the prediction is specifically about their status by 2028, not just rising prominence before then.

Because the forecast’s target period has not yet arrived, the prediction cannot be judged as right or wrong at this time. It is therefore inconclusive (too early).

As of November 30, 2025, the 2026 U.S. midterm elections have not yet taken place (they are scheduled for November 2026). Because the full general‑election campaign, media narrative, and post‑election analyses of voter priorities are still in the future, we cannot yet know whether economic issues will in fact dominate over social or foreign‑policy issues. There is therefore not enough evidence at this time to judge whether this prediction is right or wrong.

As of November 30, 2025, neither the 2028 Democratic presidential primaries nor the 2028 general election have taken place, so we cannot test the core of Chamath’s prediction (whether a self‑identified socialist could or does win the Democratic primary, and whether such a candidate then loses the general under the Electoral College).

Key points:

• The 2028 U.S. presidential election is scheduled for November 7, 2028, and the primary calendar and candidate field for that cycle are not yet fully determined or completed as of late 2025. The Democratic nominee has not been selected, and the general election result is unknown.
• There have been no completed 2028 presidential contests that would empirically confirm or falsify the claim that “a socialist can absolutely win” the Democratic primary but “cannot win” the general election under current Electoral College conditions.
• While there have been ongoing debates about reforming or abolishing the Electoral College, no constitutional amendment or nationwide compact has taken effect to replace it as of 2025; the Electoral College system remains in place for upcoming presidential elections, including 2028. This preserves the structural premise of the prediction, but doesn’t yet show whether a socialist nominee would in fact lose in the 2028 general election.

Because the necessary events (the 2028 primaries and general election) have not occurred, the prediction cannot yet be judged as right or wrong. Therefore the correct status is "inconclusive" (too early).

The prediction explicitly concerns what will happen “over the coming years” to self‑identified socialist or highly left‑wing movements in the U.S. The podcast was released on 22 August 2025, and today is 30 November 2025, only a little over three months later. Multi‑year political momentum and support trends (especially for ideological movements) are typically judged over at least one or more election cycles (e.g., 2026 midterms, 2028 presidential), which have not yet occurred. Even if current polls or news showed short‑term increases or declines, that would not be enough to determine whether these movements will continue gaining momentum versus plateauing or declining over the coming years. Therefore, there is not yet enough elapsed time or data to label the prediction as right or wrong.

As of November 30, 2025, the 2028 Democratic presidential primary field is not set and no one has formally declared a 2028 presidential candidacy among the three governors named.

Current public positions:

• Gavin Newsom (CA) – Has not announced a run, but has openly said he will give a 2028 presidential bid “serious thought” after the 2026 midterms and acknowledges he’d be “lying” if he claimed he wasn’t considering it. He is actively building national profile and infrastructure that could support a run, but remains undeclared. (thewrap.com)
• Gretchen Whitmer (MI) – Has not announced a campaign but explicitly said she “can’t rule anything out” regarding a 2028 Democratic presidential bid, keeping her options open while focusing on her governorship and the 2026 midterms. (news.bloomberglaw.com)
• Wes Moore (MD) – Has not announced a run and, in fact, has publicly ruled out running for president in 2028, stating on NBC’s “Meet the Press” that he is “not running for president” and intends to serve a full second term as governor if reelected. (aol.com)

The prediction claims that all three will ultimately run in the 2028 Democratic primary. Because the primary is still in the future and filing/ballot deadlines have not passed, any of these governors could still change course, including Moore despite his present denial. Political figures do sometimes reverse earlier statements about not running.

Therefore, it is too early to say definitively whether all three will or will not run. The prediction is trending against realization (given Moore’s current stance), but it has not yet been conclusively proven right or wrong, so the appropriate status is inconclusive (too early).

The normalized prediction claims that from 2025 onward, Germany and the broader EU will permanently cease purchasing Russian oil.

• Germany: Evidence indicates Germany halted Russian crude oil imports at the start of 2023 and has not resumed them. Statista data show Germany has imported “almost no” Russian crude oil in the last two years up to December 2024, and German officials have stated that the country no longer relies on direct Russian energy imports.(statista.com) As of late 2025 there are no reports of Germany restarting Russian oil purchases, so Germany alone is consistent with the prediction so far.

• European Union as a whole:

  • The EU sharply reduced but did not end Russian oil imports. A European Parliament briefing (Nov 20, 2025) notes that Russian crude oil fell from 27% of EU crude imports in 2021 to 3% in 2024, and it outlines a roadmap to end Russian oil imports only by the end of 2027, implying ongoing imports into 2025.(europarl.europa.eu)
  • Statista data show the EU still imported about 710,000 metric tons of oil and petroleum products from Russia in August 2024.(statista.com)
  • Shipping analysis for April 2025 finds that the EU still received about 6% of Russia’s seaborne crude exports, with Novorossiysk a major port for shipments to the EU, confirming continued purchases in 2025.(bssc.pl)
  • As late as November 28, 2025, AP reports that Hungary, an EU member state, “continued reliance on Russian oil and gas” and had secured exemptions allowing it to maintain energy imports from Russia.(apnews.com)

Because at least one EU member (and thus the EU as a bloc) is still importing Russian oil after the prediction date, the claim that the EU would, from 2025 onward, permanently stop buying Russian oil is already false. Even though we cannot yet evaluate “never” in an absolute sense, the condition that no EU oil purchases from Russia occur from 2025 onward is clearly not met.

Since the normalized prediction is about Germany and the broader EU, and the EU part has been falsified, the overall prediction is wrong.

As of 30 November 2025, Russia’s full-scale war against Ukraine is still ongoing, with major missile and drone attacks continuing and no announced, formal Russia–Ukraine peace treaty or ceasefire agreement in place, despite multiple U.S.-brokered negotiation attempts. (reuters.com) Jason’s forecast was about the probability (roughly 50%) that Trump would secure such a deal at some point during his current 2025–2029 term; since that term is far from over and the prediction was explicitly probabilistic rather than a categorical claim, it cannot yet be evaluated as clearly right or wrong.