Big Fed rate cuts, AI killing call centers, $50B govt boondoggle, VC's rough years, Trump/Kamala

Chamath’s prediction in the Sept 20, 2024 episode was that once all All‑In Summit clips were released and “baked for a couple of months,” their combined views on YouTube and X would be “around 50 million.” This is explicitly framed in the transcript as a forward‑looking estimate based on roughly 20M early views for about half the clips. (podscripts.co)

What we can verify:

1. YouTube channel totals. The main All‑In Podcast YouTube channel has about 148.6M lifetime views as of Nov 23, 2025; SocialBlade showed roughly 120M views around mid‑2025. (speakrj.com) That 120M figure has to cover:

   • All regular podcast episodes (hundreds of videos since 2020), plus
   • Summit recordings from 2022, 2023, 2024, and early 2025. Given this, it’s unlikely (though not impossible) that 2024 summit clips alone contributed ~50M views on YouTube by early 2025, but the aggregate numbers don’t let us isolate the summit share.

2. Per‑video YouTube views for key 2024 summit talks. Third‑party sites that scrape YouTube give sample view counts such as:

   • Elon Musk | All‑In Summit 2024: ~0.9–1.5M views depending on snapshot. (glasp.co)
   • John Mearsheimer & Jeffrey Sachs | All‑In Summit 2024: roughly 1–2M views across different snapshots. (ytscribe.com)
   • Peter Thiel | All‑In Summit 2024: around 0.7–0.8M views. (metapodcast.net) Even assuming similar or somewhat lower numbers for other big‑name sessions (Sergey Brin, Travis Kalanick, Marc Benioff, JD Vance, Megyn Kelly, etc.), a reasonable estimate for YouTube views across all 2024 summit long‑form videos lands in the single‑digit to low‑teens millions, not obviously near 50M.

3. Historical benchmark from 2023 summit. A LinkedIn recap of the 2023 All‑In Summit notes that the top two panels (Elon Musk; Mearsheimer/Sachs) were at ~955k and ~918k YouTube views respectively, and that the hosts hoped future summit recordings could reach about 50M views on YouTube and X combined. (linkedin.com) This shows the 50M figure was aspirational and that, even after a year, the prior summit’s biggest videos were around the 1M mark each—again consistent with a total far below 50M purely on YouTube.

4. X (Twitter) views are not auditable. Chamath’s 50M target explicitly included X, but:

   • X does not provide a public, summable API endpoint for total video/impression views per campaign.
   • We can see that various summit clips (for example, those shared by the @theallinpod account and amplifying accounts like @tsarnick) were widely circulated, but their individual view/impression counts are not systematically archived by reliable third‑party analytics in a way we can sum across all clips. (gaganghotra.com)
   • No credible article or later podcast segment appears to report a consolidated, verifiable figure like “the All‑In Summit 2024 clips did 50M+ views across YouTube and X.”

5. Timing ambiguity. The prediction was “around 50 million … when all the clips are released and you let it bake for a couple of months.” (podscripts.co) We do not have:

   • a clear public date when all summit clips finished rolling out, nor
   • any third‑party time‑series data tying combined YouTube+X views to that 2–3 month post‑release window.

Given all this:

• We can say with some confidence that YouTube alone appears to fall well short of 50M views for the 2024 summit content in the relevant timeframe.
• However, the prediction was explicitly about YouTube plus X combined, and we lack trustworthy, aggregate X view data to either confirm or rule out the total.
• No later authoritative source (press, analytics firm, or the hosts themselves) publishes a clear combined number that would settle it either way.

Because the necessary cross‑platform view data is not publicly reconstructable—despite enough time having passed—the truth of Chamath’s “~50M combined views” prediction cannot be determined from available evidence, so the correct grading is ambiguous, not confidently “right” or “wrong.”

As of November 30, 2025, the end of 2026 has not yet occurred, so we cannot determine whether the federal funds rate will be in the 2–3% range by that time.

What we can say now:

• The Federal Reserve began cutting rates in 2024–2025 after its earlier hiking cycle, but the current federal funds rate target is still well above 2–3% (exact level depends on the latest FOMC decision, which as of late 2025 remains in the mid‑single digits range, not near 2–3%).
• Because the prediction is specifically about the level of rates by end of 2026, and that date is in the future, there is no way yet to classify the prediction as right or wrong.

Therefore the correct classification at this time is "inconclusive" (too early).

BEA data and subsequent analysis show that U.S. GDP revisions in the quarters after September 2024 were mostly upward or negligible, not “a couple of significant downward revisions” that broadly weakened the growth story.

Key points:

1. Q3 2024: Real GDP was estimated at 2.8% in both the advance and second estimates, then revised up to 3.1% in the third estimate. (bea.gov)

2. Q4 2024: Real GDP came in at 2.3% in the advance and second estimates and was revised up to 2.4% in the third estimate. (bea.gov)

3. Annual revisions for recent years: The 2024 annual update (released Sept. 26, 2024) revised 2019–2023 annual real GDP growth mostly up, not down. The 2025 annual update (released Sept. 25, 2025) explicitly notes that the percent change in real GDP for 2022–2024 was unrevised, meaning no downward reassessment of overall recent growth. (apps.bea.gov)

4. Q1 2025 (the one clear downward revision): Real GDP for Q1 2025 was initially reported at –0.3% (advance), then –0.2% (second estimate), and later revised down to –0.5% in the third estimate; journalists described this as the economy contracting more than previously thought. (bea.gov) The 2025 annual update then nudged this slightly further down to –0.6%. (bea.gov) This is one quarter with a materially more negative revision.

5. Q2 2025: Real GDP was revised up repeatedly, from 3.0% (advance) to 3.3% (second) and then to 3.8% (third/annual-update release). (bea.gov)

6. Net effect on the perceived economy: The 2025 annual-update article emphasizes that average real GDP growth from 2019–2024 is essentially unchanged from prior estimates, and that annual growth rates for 2022–2024 are either unchanged or previously revised up, not down. (apps.bea.gov) The one notable negative surprise is Q1 2025, not “a couple” of significant downward GDP revisions across multiple quarters.

Because only one quarter (Q1 2025) experienced a clearly significant downward GDP revision while surrounding quarters saw upward or flat revisions, Chamath’s prediction of multiple significant downward GDP revisions over the next few quarters—and a generally weaker picture emerging from revisions—did not materialize. Hence the prediction is best classified as wrong.

As of November 30, 2025, we are only ~14 months into a 2–3 year prediction window that runs roughly from September 2026 to September 2027 for the full outcome. It is therefore too early to decisively judge whether “massive disruption” has occurred or will occur in that time frame.

What we see so far (supports possible future disruption):

• Multiple surveys indicate that a majority of contact centers already use some form of AI (chatbots, assistants, analytics), with estimates around 50–65% adoption, and large shares planning further investment. (emerge.haus)
• Industry reports project rapid growth in call-center AI markets through the late 2020s, with strong economic incentives for automation and workforce reduction. (businessresearchinsights.com)
• Some forward‑looking analyses specifically forecast that by 2027 leading call centers could have 50–60% smaller human workforces due to AI, implying substantial disruption if these projections materialize. (emerge.haus)

What we don’t yet see (why it can’t be called “right” today):

• Current credible estimates still show most customer interactions handled by humans, with AI automating a minority of volume. Gartner, for example, projects only about 10% of agent interactions fully automated by 2026, which is significant but not obviously a “massive” displacement of the industry as a whole. (sci-tech-today.com)
• A 2025 Cavell study projects that global contact-center agent roles will increase from 15.3M (2025) to 16.8M (2029), concluding that AI will slow agent growth rather than cause net job collapse in the near term. (cxtoday.com)
• A Goldman Sachs–linked survey finds that, as of late 2025, only about 10% of firms have actually cut jobs due to AI, and many executives still consider AI too early for broad deployment, even though they expect more impact over the next few years—especially in customer support. (businessinsider.com)

Why the verdict is “inconclusive (too early)”

• Sacks’s prediction is about what will happen within 2–3 years, not what must be true just one year in. For it to be clearly wrong, we would need to reach at least the 3‑year mark (September 2027) without seeing industry‑wide, large‑scale displacement or transformation.
• For it to be clearly right already, we would need strong evidence that by late 2025 a “massive disruption” had already occurred—e.g., large, measurable drops in global call‑center employment or a clearly dominant share of interactions being AI‑handled. Current data instead shows rapid adoption and meaningful efficiency gains, but still incremental displacement and even net growth in agent headcount.

Given the remaining time in the forecast window and the mixed but rapidly evolving data, the only defensible status as of November 30, 2025 is “inconclusive (too early)”, not definitively right or wrong.

The prediction’s explicit timeframe is within roughly 2–3 years of September 2024, i.e. by about September 2026–September 2027. As of today (30 November 2025), that window has not yet elapsed, so it is too early to say definitively whether it will prove right or wrong.

Current evidence shows strong adoption of AI tools in contact centers, but not the outcome described (“all the level one customer support is going to get replaced by AI”):

• A 2025 Calabrio “State of the Contact Center” report finds 98% of contact centers are using some form of AI, mostly chatbots/voicebots and analytics, but describes AI as an add‑on to human agents and focuses on challenges balancing automation with human empathy, not wholesale replacement of front‑line staff. (businesswire.com)
• The Emerge Haus AI Atlas notes that as of 2024 fewer than 1% of organizations are piloting fully autonomous AI agents that handle live customer calls, with broader Level‑3 (autonomous agent) adoption projected only to start inflecting around 2025–2027. Even their aggressive forecast for 2027 is that perhaps ~50% of major call centers will use AI agents for some call types, implying substantial human involvement remains. (emerge.haus)
• A KPMG-based analysis of large U.S. firms reports that only about 11% have deployed AI agents at scale in customer service, and emphasizes that “human call agents remain the backbone” of support operations, with the vast majority of interactions still handled by people. (dinocajic.com)
• Broader enterprise surveys find that agentic AI adoption significantly lags the hype: Gartner-style projections put major AI resolution rates (e.g., ~80% of customer service issues handled by AI) closer to 2029, and note that most organizations still insist on human oversight. (itpro.com)
• A Goldman Sachs banker survey indicates only about 10% of firms have cut jobs due to AI so far, though many expect customer-support roles to be among the most affected in coming years—evidence of early impact but not mass replacement yet. (businessinsider.com)

So as of late 2025, level‑one call center work has not been “essentially all” replaced by AI; humans still handle a large share of first‑line interactions. However, because the prediction allows until roughly 2026–2027, there is still time for further adoption and job displacement. Given that the deadline has not passed and the trajectory could still change, the status of the prediction is best classified as **“inconclusive (too early)” rather than clearly right or wrong.

Available evidence by roughly one year after the podcast (through late 2025) shows that AI customer service software did not become a fully commoditized, low‑value startup category, even though basic AI features became common.

1. Major AI customer‑support startups are still capturing large, differentiated value. Sierra, an AI customer support startup, reported its ARR growing from about $20M to $100M in a year and was valued around $10B, after raising more than $600M. This is specifically an AI customer‑support company, not a foundation‑model provider, and it is expanding headcount and office footprint rather than being squeezed out by commodity competition. (theverge.com)
2. Incumbent application‑layer vendors doubled down on AI and maintained strong valuations. Intercom, whose core product is an AI‑driven customer service and messaging platform, rolled out Fin and then Fin 2, and was in 2025 discussions with investors about multibillion‑dollar valuations alongside other AI customer‑support vendors like Kore.ai. These companies are positioned as differentiated platforms built on top of foundation models, not as commodity utilities displaced by them. (en.wikipedia.org)
3. Market research describes a fast‑growing, multi‑vendor software market, not a commodity layer owned by a few model providers. Industry reports put the AI‑for‑customer‑service market at roughly $12–13B in 2024, projecting high‑20s CAGRs and tens of billions in value by 2030, with long lists of distinct software vendors (Zendesk, ServiceNow, Freshworks, Ada, Aisera, Intercom, LivePerson, etc.) participating in the space. This is consistent with a competitive applications market riding on foundation models, not one that has been “run over” and reduced to undifferentiated infrastructure. (globenewswire.com)
4. Commentary from practitioners and vendors frames AI features as ‘table stakes’ but stresses differentiation in implementation, trust, and integration. Multiple 2024–2025 pieces from customer‑service‑AI providers emphasize that simply having a chatbot is now “table stakes,” yet argue that real differentiation comes from agentic behavior, omnichannel workflows, governance, and domain‑specific tuning—i.e., application‑layer work on top of commoditizing models still matters and is where value is captured. (forbes.com)
5. Gartner/Reuters analysis points to a crowded and hyped vendor landscape, not pure model‑layer domination. Gartner estimated thousands of self‑described “agentic AI”/support vendors, with only ~130 considered technically legitimate, and predicted that many projects will be scrapped. That suggests oversupply and shake‑out risk—but also that a subset of vendors have meaningful, defensible technology and business value, rather than all value collapsing into generic foundation‑model APIs. (reuters.com)

Taken together, the data show a crowded, partially commoditizing feature space (everyone can plug into foundation models) but continuing strong value capture by differentiated AI customer‑service platforms and startups well past September 2025. That contradicts Chamath’s strong claim that you “cannot touch customer service” because it would be fully commoditized and “run over” by foundation models within about a year.

Evidence from OpenAI’s own materials shows that this prediction basically came true on both timing and substance.

• Preview then production: OpenAI announced OpenAI o1‑preview (with o1‑mini) on September 12, 2024 as an early preview reasoning model in ChatGPT and the API. (openai.com) Later, OpenAI documentation for ChatGPT states that “ChatGPT Enterprise and Edu customers will have access to the o1 model on December 5”, referring to the non‑preview o1 model. (help.openai.com) That’s about 2½ months after the September 20, 2024 podcast—within a “couple of months” window.

• Production build: In the product blog “OpenAI o1 and new tools for developers,” OpenAI describes o1 (not o1‑preview) as “the successor to OpenAI o1‑preview” and explicitly calls it “production‑ready”, listing key production features like function calling, Structured Outputs, developer messages, vision, and a reasoning_effort control. (openai.com) The same article notes that the snapshot o1‑2024‑12‑17 is the version being shipped in the API, clearly marking it as the production release of the o1 line. (openai.com) The API pricing page separately lists o1 as a regular, billable model, distinct from preview models, further confirming production status. (platform.openai.com)

• “Spectacular” / materially more capable: The o1‑2024‑12‑17 snapshot sets state‑of‑the‑art results on several benchmarks and is significantly stronger than o1‑preview—for example, AIME 2024 accuracy jumps from 42.0 to 79.2, MATH from 85.5 to 96.4, and SWE‑bench Verified from 41.3 to 48.9—while also using ~60% fewer reasoning tokens and adding major capabilities (vision, function calling, structured outputs). (openai.com) These are large, tangible gains over the preview model, consistent with a “spectacular” or materially more capable upgrade.

Given that: (1) a non‑preview, production‑ready o1 shipped around mid‑December 2024, roughly “in the next couple of months” after the podcast, and (2) it is clearly a significantly more capable successor to o1‑preview, Chamath’s prediction is best classified as right.

As of November 30, 2025, there is not enough realized data on post‑2020 venture capital vintages to say whether average VC fund returns will ultimately be 50–100% lower than historical norms.

Key reasons:

1. VC fund outcomes are inherently long‑dated. Typical venture funds are structured for ~10–12 years, and many now take 15–20 years to fully return capital; most value realization happens in years 7–10+ and often later. (commonfund.org) That means funds raised around 2020–2022 are still in their early or mid investment period in 2025.

2. J‑curve and lack of distributions show we are still early. Analyses of VC J‑curves show that the negative or flat part usually lasts several years, and Carta’s 2024 data notes that more than 60% of 2019‑vintage VC funds had not distributed any capital back to LPs even after five years. (carta.com) If 2019 vintages are still largely unrealized, 2020–2022 bubble‑era vintages are even further from their true, realized performance.

3. Interim performance is mixed and not predictive of final returns. Cambridge Associates’ 2023–2024 benchmarks show that recent VC vintages (2014–2022) have had weak to slightly positive annual returns, with many pre‑2020 vintages negative in 2023 and some post‑2020 vintages (like 2022) positive in 2023–2024. (cambridgeassociates.com) These are mark‑to‑model/mark‑to‑market snapshots, not final cash‑on‑cash outcomes, and industry analysts explicitly warn that early TVPI/IRR after a few years is not a reliable indicator of eventual fund multiples.

4. Current underperformance vs public markets doesn’t fix the long‑run multiple. Recent commentary notes that VC returns over the last year have been around -1%, versus roughly +20% for public equities, and distributions to LPs in 2023–2024 are far below peak years, contributing to weak fundraising and LP fatigue. (wsj.com) This indicates pressure on the asset class and raises the odds of lower eventual returns—but it still doesn’t tell us whether final average vintage‑year multiples will end up half of, equal to, or only modestly below historical norms.

Because:

• the relevant 2020–2022 "bubble" vintages are only ~3–5 years into what is likely a 12–20 year realization arc, and
• core outcome metrics like DPI (cash returned) and final net multiples are largely unknown for those vintages,

we cannot yet quantitatively determine whether average VC fund returns will be 50–100% lower than historical averages. There are signs consistent with stress and potentially lower future returns, but the prediction is about ultimate average realized performance, and that won’t be observable until well after 2025. Therefore the status of this prediction is inconclusive (too early to tell).

The prediction is framed over “the next several years,” so by November 30, 2025 only ~14 months have elapsed—too little to know whether VC fundraising will remain structurally lower than 2020–2021 for the entire horizon.

Data so far do line up with key parts of Chamath’s view: the number of active U.S. VC firms has already fallen by more than 25% from 2021 to 2024, indicating a substantial washout of managers as capital concentrates in a smaller set of large platforms. (ft.com) Fundraising volumes are also far below the 2021 peak: global VC fundraising in 2024 was about $160.6B versus $403.6B in 2021 (down ~60%), U.S. VC funds raised about $76.8B across 538 funds in 2024, and H1 2025 raised only $26.6B across 238 funds—on pace for the lowest fundraising year in a decade. (sourcery.vc) Capital is increasingly concentrated: in 2024, just 30 U.S. firms captured 75% of commitments and nine firms took roughly half, while by Q3 2025 the top 10 funds accounted for 42.9% of all capital raised and the count of emerging-manager funds hit its lowest level since 2015. (linkedin.com) All of this supports the “many VCs washed out, structurally lower fundraising so far” part of the thesis.

Where the evidence is less clear is the causal link that lower fundraising reflects startups needing “a lot less capital.” Median early‑stage round sizes have continued to rise, hitting record levels in 2024–2025, and analysts note that even with AI-driven efficiency, early‑stage check sizes are growing faster than inflation, not shrinking; instead, capital is being rationed into fewer, larger winners rather than broadly reduced because companies inherently need less money. (axios.com) At the same time, some AI and infrastructure startups remain extraordinarily capital‑intensive, with multi‑billion‑dollar rounds. (reuters.com)

Because (1) the time window he specified (“several years”) has not yet played out and (2) the mechanism he posits about companies fundamentally needing less capital is only partially supported and partly contradicted by current data, it is too early to say the full multi‑part prediction is definitively right or wrong; it is best classified as inconclusive so far, though early fundraising and firm‑consolidation trends are directionally consistent with his view.

As of 30 Nov 2025, there has not yet been a clear, widely adopted new mechanism that fundamentally “reinvents” or substantially restructures the U.S. IPO process, but the prediction explicitly refers to changes happening over the “coming years,” which extends beyond the ~14 months since the podcast.

Key points:

• The main ways U.S. companies access public-market capital remain traditional underwritten IPOs, direct listings (with or without a capital raise), SPACs/de‑SPACs, and reverse mergers—all of which existed before September 2024.

  • Nasdaq’s Direct Listing with Capital Raise (DLCR), which allows companies to both list and raise primary capital in the opening auction, was approved by the SEC in December 2022, well before the podcast, and is still framed as an incremental alternative rather than a replacement for the IPO process. (nasdaq.com)
  • Legal and practitioner guides in 2024–25 still describe the landscape as a menu of IPO, SPAC, direct listing, and reverse-merger options—"evolving" but not fundamentally re‑architected. (finsyn.com)

• Empirically, the backlog / bottleneck problem that Chamath is reacting to has not been resolved:

  • Reports in 2024–25 note that the number of U.S. public companies remains far below late‑1990s levels, even after a rebound in IPO volume, and many large firms (e.g., major tech names) continue to stay private longer. (barrons.com)
  • Private equity is still sitting on a very large overhang of unsold assets (roughly $1 trillion as of mid‑2025), with exits via IPOs and M&A constrained—evidence that no new, scalable exit channel has yet unlocked the backlog. (reuters.com)

• Regulatory signals point toward potential future changes but not a completed “reinvention” yet:

  • In 2025, SEC leadership has emphasized making public listings easier and is exploring rules for blockchain-based issuance and trading of securities, but these are early-stage regulatory directions rather than an already-implemented structural overhaul of the IPO mechanism. (barrons.com)

Given this, we can say:

• The conditions Chamath describes (too many private companies, constrained public-market access) still largely exist.
• The existing IPO process has not yet been clearly “reinvented” or supplemented by a genuinely new, widely used mechanism beyond the pre‑2024 toolkit (IPO, SPAC, direct listing, reverse merger, etc.).
• However, because his timeline is “in the coming years” and regulators/markets are visibly working on incremental reforms and potential new structures, it is too early to declare his prediction definitively right or wrong.

Therefore, the fairest evaluation as of Nov 2025 is “inconclusive (too early)” rather than “right” or “wrong.”

By November 30, 2025 it is too early to know how 2021–2022 venture vintages will ultimately perform.

Key points:

• 2021–2022 vintage funds are only ~3–4 years into their lives. Industry data shows that most VC funds don’t start returning meaningful cash (DPI) until years 5–7, and often don’t reach 1.0× DPI (return of paid‑in capital) until around year 8–10 or later.(phoenixstrategy.group) Final outcomes for these vintages won’t be clear until the early–mid 2030s.
• Carta’s recent benchmarks confirm that 2021–2022 vintages are lagging earlier vintages in DPI: only a small minority of 2021 funds had begun returning any capital after three years, and 2022 funds are also behind historical norms on both deployment and DPI.(carta.com) However, this mainly shows slower exits and illiquidity, not terminal multiples.
• Analyses of vintage‑year performance show the 2021 vintage currently has weaker IRR than surrounding years, with median IRR around zero or negative, reflecting peak 2021 valuations and the subsequent correction.(techcrunch.com) But IRR and TVPI for such young funds are mostly unrealized marks and can change substantially if a few large winners exit later.
• The prediction is about end‑of‑life performance and LP sentiment ("you could return capital, you're going to look like a hero") for the 2021–2022 deployment vintages. There is not yet systematic evidence that, on average, these funds will end up only around 1.0× DPI, nor that LPs broadly view a simple return of capital on these vintages as an unusually good outcome. Available data only shows underperformance and delayed liquidity so far, not final multiples.

Because (1) typical realization timelines run 8–10+ years, and (2) current metrics for 2021–2022 vintages are still largely unrealized and could move significantly, Chamath’s claim about their ultimate performance and how LPs will feel about a mere return of capital cannot yet be judged as right or wrong.

The macro condition Sacks described has largely occurred, but whether the consequence (“a new, sustained golden era…not a bubble”) has come true is not objectively resolvable yet.

1. Macro preconditions were basically met
• After the initial 50 bp cut on September 18, 2024, the Fed cut another 25 bp in November and 25 bp in December, taking the funds rate from 5.25–5.50% before September down to 4.25–4.50% by year‑end—i.e., roughly another 50 bp of cuts in 2024, and about 100 bp in total. (forbes.com)
• Additional cuts in September and October 2025 lowered the target range further to 3.75–4.00%, meaning rates moved down substantially from their peak while remaining far from zero. (forbes.com)
• U.S. inflation fell sharply from 2022–23 peaks; by 2024 headline CPI was about 2.9% and core about 3.2%, with the Fed itself characterizing inflation as moving sustainably toward its 2% goal and citing “cooling” inflation when cutting in late 2024. (jpmorgan.com)
On the narrow macro terms Sacks specified (more ~50 bp of cuts in 2024, inflation contained, rates down but not to zero), the condition was broadly satisfied.

2. Evidence for an AI‑driven tech/VC boom
• The Nasdaq Composite gained roughly 30% in 2024 and the Nasdaq‑100 about 26%, with further strong gains into 2025 and fresh all‑time highs in October 2025, driven heavily by AI‑centric names like Nvidia. (investor.wedbush.com)
• U.S. VC investment rebounded in 2024 to about $209B (up nearly 30% YoY), with AI startups capturing a record ~46% of that capital. (reuters.com)
• By 2025, AI accounts for an estimated 58% of U.S. VC dollars and roughly one‑third of deals; many venture‑backed tech companies are again growing revenue, and the tech IPO window is reopening. (prnewswire.com)
This is strong evidence of a powerful AI‑driven upswing in opportunity and performance across large parts of the tech and venture ecosystem.

3. But the “golden era, not a bubble” characterization is disputed
• The venture environment is highly uneven. The number of active U.S. VC firms fell more than 25% from 2021 to 2024 as capital concentrated in a handful of mega‑funds, and multiple reports highlight continued fundraising weakness and firm closures. (ft.com)
• Asia’s startup funding fell to a 10‑year low in 2024, with AI investment there notably lagging the global AI boom—so the “ecosystem” is far from universally flourishing. (news.crunchbase.com)
• At the same time, prominent voices warn explicitly of an AI bubble: Sam Altman has called the current AI market a speculative bubble; the Nasdaq‑100 has suffered one of its worst quarters in years amid “AI bubble fears”; and fiscal authorities like the UK’s OBR have modeled serious downside risks from a potential AI‑stock crash. (theverge.com)
• Other analysts argue valuations are “stretched but not yet at levels consistent with historical bubbles,” with the Nasdaq‑100 at ~28× forward earnings versus a 10‑year average of ~23×—high, but not definitively bubble territory. (nasdaq.com)
So there is genuine disagreement among informed observers about whether today’s AI‑driven tech market represents a healthy rerating or a classic bubble.

4. Why the prediction is ambiguous rather than clearly right or wrong
• Sacks’s claim bundles several qualitative and long‑horizon ideas: a “new, sustained golden era” comparable to the early‑2000s boom, and “not…a bubble‑level of excess.” These terms are inherently subjective and would normally require many years of data to judge. By November 2025 we have only about one year of post‑prediction history.
• The observable facts cut both ways: AI has clearly powered a major boom in tech markets and VC dealflow, but fundraising stress, regional weakness, and persistent “AI bubble” warnings complicate any claim that the era is both broadly golden and clearly non‑bubble.

Because the macro precondition largely occurred, but the downstream claim about a “sustained golden era” without bubble‑level excess depends on contested, long‑term, and partly subjective judgments, the outcome cannot be assessed as definitively right or wrong at this point. Hence the result here is ambiguous.

Donald Trump did in fact win the November 5, 2024 U.S. presidential election, defeating Kamala Harris with 312 electoral votes to 226 and winning the popular vote roughly 49.8% to 48.3%. (en.wikipedia.org) This means Jason’s stated view that there was a “pretty good” / better‑than‑even chance Trump would win ended up aligned with the actual outcome.

On the mechanism, post‑election analyses show that pre‑election polling once again slightly underestimated Trump’s support by about 2–3 percentage points nationally and in key swing states, generally biasing toward the Democrat. (theguardian.com) Explanations emphasize things like nonresponse bias and turnout modeling problems—Trump‑leaning voters being less likely to answer or complete polls, or breaking late toward Trump—rather than clear evidence of large numbers of voters actively lying to pollsters. (pewresearch.org) That pattern is at least directionally consistent with Jason’s idea that some Trump supporters were not fully visible in polling or social circles, though experts still debate how big a classic “shy Trump” (social‑desirability) effect really is.

Because the core, checkable part of the prediction—Trump having a strong chance and ultimately winning in November 2024—came true, and the claimed existence of a non‑open Trump‑supporting bloc is broadly consistent with the observed, modest undercount of Trump in polls (even if the exact cause is disputed), the fairest overall judgment is that this prediction was right.