E34: Wuhan lab leak theory, India's "traceability" law, Coinbase's fact check, Big Tech's Hollywood takeover

Jason’s prediction had two key parts: (1) that, within roughly 90 days of late May 2021, the U.S. government would already know that COVID-19 came from a Wuhan lab, and (2) that after the 90‑day review it would then begin concrete policies to become more economically independent from China, triggered by that knowledge.

1. U.S. government “already knows” it was a lab leak
   • President Biden ordered a 90‑day intelligence review of COVID‑19’s origins on May 26, 2021; the report was delivered August 24 and a declassified summary released August 27, 2021. The assessment explicitly stated that the U.S. Intelligence Community was divided and reached no definitive conclusion on whether the virus came from natural spillover or a lab incident. Four agencies plus the National Intelligence Council, all with low confidence, leaned toward a natural zoonotic origin; one agency (the FBI) assessed with moderate confidence that it was most likely a lab‑related incident; others remained undecided. (en.wikipedia.org)
   • Even years later, U.S. agencies remain split. In 2023 the Department of Energy shifted to a low‑confidence assessment favoring a lab‑related incident, and in 2025 the CIA reportedly judged a lab origin “more likely” but still with low confidence—while other agencies still favor zoonosis. (en.wikipedia.org) There is still no government‑wide, high‑confidence finding that the virus leaked from a lab.
   • WHO‑convened studies and much of the scientific literature continue to view zoonotic spillover as the more likely pathway and describe a lab leak as “extremely unlikely” or at least unproven, underscoring that origins remain unresolved. (en.wikipedia.org)
   Given the official, declassified record and subsequent updates, the claim that the U.S. government already knew the lab‑leak theory was true within that 90‑day window is not supported; at most, one agency leaned that way with moderate confidence while others did not.

2. Post‑review economic “independence from China” starting then
   • Major U.S. efforts to reduce supply‑chain dependence on China were already underway before the May 31, 2021 episode and before the 90‑day origins review concluded. Biden signed Executive Order 14005 ("Buy American") on January 25, 2021 to strengthen domestic sourcing for federal procurement. (en.wikipedia.org)
   • On February 24, 2021, he signed Executive Order 14017 on America’s Supply Chains, launching a 100‑day review of critical supply chains (semiconductors, large‑capacity batteries, critical minerals, pharmaceuticals) aimed at strengthening U.S. manufacturing, diversifying sources, and reducing vulnerabilities from concentrated foreign supply—with China repeatedly cited as a dominant, risky supplier in follow‑on reports. (cisa.gov) Those reviews and recommended actions were announced June 8, 2021, still before the COVID‑origins intel report was completed.
   • In parallel, Congress was already moving on the U.S. Innovation and Competition Act (USICA)—later folded into the CHIPS and Science Act—which explicitly aimed to reshore semiconductor production, bolster supply‑chain resilience, and “counter China.” The Senate passed USICA on June 8, 2021, again predating the August 2021 origins assessment. The final CHIPS Act, signed in August 2022, is framed as strengthening U.S. supply chains and competing with China, but it was the continuation of this pre‑existing trajectory, not a new turn taken right after the 90‑day review. (en.wikipedia.org)
   • Later initiatives (e.g. Indo‑Pacific Economic Framework, export controls on advanced chips, further tariff measures) also stress supply‑chain security and competition with China, but they are not tied by the administration to any definitive finding that COVID‑19 leaked from a lab; they reflect a broader, ongoing strategic shift that began earlier.

Taken together, the best available public evidence shows:

• The 90‑day intelligence review did not conclude that COVID‑19 definitively came from a Wuhan lab; the U.S. government still does not “know” this with high confidence.
• Policy moves to reshore manufacturing and reduce dependence on China were already in motion before May–August 2021 and were not newly initiated after the 90‑day review on the basis of a confirmed lab‑leak finding.

Because both central components of Jason’s prediction (government already secretly knowing lab leak, and a subsequent, causally linked policy shift starting after the 90‑day review) are contradicted by the observable record, the prediction is best classified as wrong.

Evidence since 2021 supports parts of Chamath’s thesis (a durable shift toward resilience-focused, more regional supply chains with very large associated investment), but contradicts the stronger claim that the prior era of globalization and just‑in‑time supply chains is effectively “over.” Because the prediction bundles these together and uses vague notions like “trillions of dollars of opportunity,” it can’t be cleanly classified as simply right or wrong.

1. Globalization “as we know it” clearly has not ended.

• Global trade remains near historic highs: world trade as a share of GDP was about 63% in 2022 and remained above mid‑2000s levels in 2023–24, indicating that the world economy is still highly integrated rather than deglobalizing in a sharp way. (ycharts.com)
• WTO data show that after a brief contraction in 2023, merchandise trade volumes are forecast to grow again in 2024–25, not collapse. (wto.org)
• OECD work on global value chains and a 2025 OECD report on economic security both find no clear evidence of large‑scale reshoring or major fragmentation of international supply chains so far; instead they describe a plateau/“slowbalisation,” with continued deep cross‑border production linkages. (oecd.org)
• The ECB similarly concludes that there is no strong reshoring response inside Europe; firms are mainly diversifying suppliers and building inventories while trade in intermediate goods stays high. (ecb.europa.eu)
• An investment note from T. Rowe Price on “friendshoring” explicitly says that “reports of globalization’s death were exaggerated” and finds little evidence of an overall increase in reshoring of manufacturing back home, even as supply chains are reconfigured. (troweprice.com)

Taken together, mainstream data and analysis suggest that globalization has been reorganized and slowed, not ended. That undercuts the literal claim that “globalization as we know it is over” and that the old model has fully given way to a new one centered on onshoring.

2. But there is a sustained shift toward resilience, regionalization, and (selective) onshoring/friendshoring.

• A 2020 McKinsey survey found 93% of supply‑chain executives planned to increase resilience (redundant suppliers, nearshoring, regionalization, etc.) after COVID‑19; Capgemini research the same year found only 14% of firms expected a return to “business as usual,” with resilience a top priority. (mckinsey.com)
• NBER research by Alfaro & Chor on the “looming great reallocation” of global supply chains shows US sourcing has shifted away from China toward Vietnam and Mexico, with some stages of production moving back upstream (consistent with partial reshoring) rather than a simple reversion to pre‑2020 patterns. (nber.org)
• Policy has reinforced this shift: the CHIPS and Science Act, Inflation Reduction Act, and infrastructure legislation together catalyzed roughly $1 trillion in U.S. private investment by late 2024, with nearly $800 billion in announced manufacturing projects aimed at semiconductors, batteries/EVs, clean energy and related advanced manufacturing—explicitly framed around domestic capacity and resilient supply chains. (en.wikipedia.org)
• U.S. data show an extraordinary boom in manufacturing‑facility construction since 2021, led by chip fabs and green‑energy plants, which Treasury and Federal Reserve analyses link directly to these policies and the broader resilience/onshoring push. (home.treasury.gov)

So, the directional macro theme Chamath identified—greater emphasis on resiliency, regionalization, and strategic onshoring with very large associated capital spending—has clearly materialized. In that narrow sense, his “macro investing theme” exists in a meaningful way.

3. Evidence for outright reshoring and for quantifying “trillions of dollars of opportunity” is mixed and hard to pin down.

• Academic and policy work stresses that hard evidence of broad‑based reshoring is still limited; value chains are “sticky,” and firms often prefer nearshoring or friendshoring over full onshoring, especially outside a few strategic sectors. (paperity.org)
• Global trade and global value‑chain participation remain very high by historical standards, and bodies like the OECD and WTO emphasize diversification and regulatory changes rather than wholesale retreat from globalization. (oecd.org)
• On the other hand, when you aggregate:
  • the ~$1 trillion in U.S. private manufacturing/clean‑tech/semiconductor investment tied to resilience and domestic capacity;
  • similar—but harder‑to‑quantify—friendshoring and industrial‑policy waves in Europe and Asia;
  • plus the market capitalization and revenue growth of firms building chips, grids, automation, logistics, and other resilience infrastructure, it is plausible that the “opportunity set” associated with this theme runs into the low trillions of dollars globally. But this is an inference, not something cleanly documented in official statistics.

Because the prediction combines:

• a strong, falsifiable claim (that the old era of globalization/JIT is effectively over) that current data contradict; and
• a broader thematic claim (that there will be a long‑lasting, trillion‑dollar shift toward resilience, onshoring and related businesses) that is directionally supported but hard to measure precisely,

the overall forecast defies a binary right/wrong judgment. The macro trend he pointed to is real and very large, but globalization has not ended, and the exact magnitude and attribution of “trillions” in opportunity cannot be rigorously confirmed.

Given this mixture of confirming and disconfirming evidence plus fuzzy definitions, the fairest rating is “ambiguous.”

Evidence since 2021 points in the opposite direction of Chamath’s prediction on all three key elements (costs, margins, and commoditization):

1. Production costs have not continued to fall; they’ve risen.

   • Ampere Analysis and others estimate that the six biggest global content companies (Disney, Comcast/NBCU, Google/YouTube, WBD, Netflix, Paramount) will spend a record $126B on content in 2024, up 9% year over year, with Netflix’s own content spend rising from about $10.9B in 2020 to $16B in 2024. (mediapost.com)
   • In the UK, overall spending on high‑end TV grew 11% in 2024, even though the number of shows dropped from 223 to 181, which industry analysts explicitly attribute to rising production and talent costs; local broadcasters say they’re being priced out of premium drama. (theguardian.com)
   • Lists of the most expensive TV series show that big‑budget streaming shows like The Lord of the Rings: The Rings of Power (≈$58M/episode, 2022), Citadel (~$50M/episode, 2023), Secret Invasion (~$35M/episode, 2023), and upcoming Stranger Things S5 ($50–60M/episode) represent new highs in per‑episode budgets, not declines. (en.wikipedia.org)
   • Trade and analyst commentary across TV, streaming and music repeatedly notes that content costs continue to rise, both for traditional TV operators and for streamers like Netflix. (advanced-television.com) The 2023 WGA strike and concurrent SAG‑AFTRA deal further locked in higher minimums and residuals for streaming, structurally increasing labor costs. (en.wikipedia.org)

2. Margins have not structurally fallen because content got cheap; leading streamers’ margins have improved despite higher costs.

   • Netflix’s operating margin rose from about 19% in 2020 to 26.7% in 2024, with management targeting ~29% in 2025, even as analysts emphasize that its content costs are still rising and may reach $20–21B annually. (ainvest.com)
   • Disney’s direct‑to‑consumer streaming segment (Disney+, Hulu, ESPN+) swung from a $2.6B loss in fiscal 2023 to a $134M full‑year profit in 2024, and then to over $1.3B in streaming profit by fiscal 2025, helped by repeated price hikes and ad‑tier monetization, despite higher programming and production costs. (thewrap.com)
   • Other major platforms (Warner Bros. Discovery’s Max, Paramount+, etc.) also moved from heavy losses toward positive streaming EBITDA by 2024–2025; sector scorecards explicitly report mid‑teens or better EBITDA margins for Netflix and double‑digit margins for Disney’s entertainment streaming bundle. (thedailymesh.com) This is inconsistent with a world where falling content costs drive permanently thin or deteriorating margins.
   • It’s true that streaming margins are structurally lower than the legacy cable bundle’s peak ~30%+ network margins, but the actual trend since 2021 has been margin expansion via price increases and ad tiers while content remains expensive, not margin compression because content became cheap.

3. Professional content has not become an economic commodity that “almost doesn’t matter.”

   • The same Ampere data show the largest media groups driving content spending to record highs, with original content alone accounting for over $56B of their investment since 2022. This ongoing arms race in premium series and films reflects that distinctive content still materially differentiates platforms. (mediapost.com)
   • Analysts emphasize that “critical franchises and tentpole series remain crucial for differentiation” in streaming, directly contradicting the notion that individual shows are economically interchangeable. (stockscounting.com) Massive per‑episode budgets for flagship series (Rings of Power, House of the Dragon, Stranger Things S4–5) further demonstrate that platforms treat these titles as high‑stakes, non‑commodity assets. (en.wikipedia.org)
   • Viewing data from Nielsen/Gracenote show a power‑law outcome: about 6% of streaming titles account for roughly 75% of U.S. viewing minutes, with the rest described as “digital landfill” – indicating that most content underperforms, but a small set of hits captures the vast majority of value, which is the opposite of uniform commoditization. (linkedin.com)

Overall, from 2021 to late 2025 the empirical record shows rising production and talent costs, record‑high aggregate content spend, improving margins at leading streamers driven by pricing and ads, and continued economic importance of standout IP and franchises. That is materially at odds with Chamath’s claim that content costs would keep falling, margins would structurally deteriorate as a result, and most professional content would become an economic commodity that “almost doesn’t matter.”

The prediction covers roughly May 2021–May 2031, so as of November 2025 the full 10‑year period has not elapsed.

Evidence so far (2021–2025) largely supports Friedberg’s description:

• The number of major subscription streamers has remained high (Netflix, Amazon Prime Video, Disney+, Hulu, Max, Paramount+, Peacock, Apple TV+, etc.), and consumer surveys in 2024–2025 report that the average U.S. household pays for about four video streaming services and that 62% of subscribers believe there are too many options, reflecting fragmentation and the need to pick and choose across services. (fool.com)
• Articles on “subscription burnout” in 2025 attribute rising cancellations partly to content being split across many platforms, forcing people to juggle multiple paid or ad‑supported services. (techtimes.com)
• At the same time, the industry is starting to respond with bundles and consolidation: Disney has long sold a Disney+/Hulu/ESPN+ bundle and is planning to fold Hulu fully into Disney+ by 2026; Warner Bros. Discovery merged HBO Max and Discovery+ into Max; Paramount+ integrated Showtime; and telecoms like Verizon and T‑Mobile increasingly market multi‑service streaming bundles. (techradar.com)
• Analysts in 2025 are explicitly predicting that a “mega‑bundle” era could arrive in the next 2–3 years, which, if it materializes and becomes dominant before 2031, would undercut the idea of a full decade of fragmented, pick‑and‑choose competition. (businessinsider.com)

So far, the competitive, fragmented landscape Friedberg anticipated does describe the market reasonably well, but because his prediction is about conditions lasting for roughly ten years (through ~2031), it is too early to say definitively whether it will hold for the entire period. Hence the outcome must be marked as inconclusive (too early) rather than fully right or wrong.

Summary of what was predicted
Sacks argued that (a) once major Hollywood studios/content libraries are acquired by big‑tech platforms (Amazon, Apple, Google, etc.), they will effectively never change hands again (no more trading every ~10 years), and (b) big tech will ultimately "gobble up" most of the remaining major Hollywood content libraries and keep them.

What has actually happened (through Nov 30, 2025)

• Amazon completed its acquisition of MGM in March 2022 and continues to own it; there has been no sign of Amazon flipping or selling MGM’s library. (washingtonpost.com)
• Other major Hollywood studios and libraries (Disney, Comcast/NBCUniversal, Sony Pictures, Paramount Global, Lionsgate) have not been acquired by big‑tech platforms. Instead, the most significant post‑2021 reshuffling has been within legacy media and telecom:
  • AT&T spun off WarnerMedia and merged it with Discovery to form Warner Bros. Discovery in April 2022 – a media/telecom-to-media transaction, not a big‑tech takeover. (ir.wbd.com)
  • Warner Bros. Discovery later announced plans to split into two publicly traded entities, again a restructuring among traditional media assets, not a sale to big tech. (reuters.com)
• Apple, Google, and others have invested heavily in original production and streaming, but as of late 2025 they have not acquired most of the legacy studio libraries.

Why this is not yet clearly right or wrong

• The permanence part of the prediction (that once big tech owns a studio it will never trade again) cannot realistically be tested only a few years after Amazon–MGM closed; we have no counterexample yet, but also no proof this will hold over decades.
• The “gobble up the rest” part has not materialized so far: outside of MGM, big tech has not acquired the other major Hollywood libraries. However, Sacks did not give a specific deadline, framing it instead as a long‑run “end state.” The absence of these acquisitions by 2025 is evidence against the speed or inevitability of his scenario, but it does not decisively falsify a multi‑decade structural prediction.

Because (1) no major big‑tech divestitures have occurred to clearly disprove the “they’ll never be traded again” claim, and (2) the time horizon for “big tech will ultimately gobble up the rest” is open‑ended, the available evidence is insufficient to declare the prediction definitively right or wrong at this point. Hence the result is inconclusive (too early).

As of late 2025, Big Tech has not acquired or directly controlled the majority of the major Hollywood studios.

1. The current “Big Five” major studios are Walt Disney Studios, Universal, Warner Bros., Sony Pictures, and Paramount Skydance. They are owned respectively by The Walt Disney Company, Comcast (via NBCUniversal), Warner Bros. Discovery, Sony Group, and Paramount Skydance Corporation—not by Amazon, Apple, Alphabet/Google, Meta, Microsoft, or Netflix.(en.wikipedia.org) These five together control over 80% of the 2024 North American box office, while Amazon MGM Studios holds only about 3.4%.(en.wikipedia.org)

2. Since 2021, the only major studio acquisition by a large tech company has been Amazon’s purchase of MGM, completed in March 2022 and later folded into Amazon MGM Studios.(en.wikipedia.org) Industry references classify Amazon MGM as a mini‑major, not one of the Big Five majors, and its box‑office share remains far smaller than any of the Big Five.(en.wikipedia.org)

3. Paramount did merge with Skydance in 2025 to form Paramount Skydance, but Skydance itself is a media/production company backed by investors, not a consumer tech platform on the order of Amazon, Apple, or Google; the resulting parent is still a traditional media conglomerate rather than “Big Tech.”(en.wikipedia.org)

4. Where Big Tech has come to dominate is distribution and viewing time, not ownership of the legacy studios: by May 2025, streaming accounted for 44.8% of all U.S. TV usage, surpassing cable plus broadcast for the first time, with YouTube and Netflix leading overall TV streaming share and Amazon’s Prime Video also significant.(nielsen.com) But this is dominance of platforms, not of the underlying studio assets that produce most Hollywood films.

Because the Big Five studios remain primarily owned by traditional media conglomerates and only one mini‑major (MGM) has been acquired by a Big Tech company, Sacks’ normalized prediction that large technology companies would acquire and control the majority of major Hollywood studios and their content output has not materialized by November 30, 2025.

On the direction of the trend, Chamath’s prediction is supported by post‑2021 data.

• A Deloitte Digital Media Trends survey of U.S. Gen Z found that watching TV or movies at home ranks only fifth among their top entertainment activities, behind video games, music, browsing the internet, and social media; only about 10% named TV/movies as their favorite pastime, versus much higher shares in older cohorts. Commentators on the study explicitly warned this likely reflects a permanent shift rather than something that will converge to older generations’ habits as Gen Z ages.【1search0】

• A 2025 AP‑NORC poll shows most Americans now prefer to watch new releases via streaming at home rather than in theaters, and the North American box office, while modestly up year over year, remains more than 20% below pre‑pandemic levels, indicating that going out to tentpole movies has not regained its former centrality.【0news16】

• Surveys of younger audiences show their media attention is dominated by short‑form, social‑first content on smartphones. A Deloitte‑cited survey reported nearly half of Gen Z prefer social‑first content to traditional entertainment, and about 88% of 13–24‑year‑olds watch video weekly on their phones. A Gen Z media executive in that coverage flatly states that “social media is the new water cooler,” i.e., day‑after conversations now center on viral clips and creators rather than last weekend’s movie.【0search3】【0search0】

• Other polling finds large shares of Gen Z have never seen many historically canonical films, recognize fewer famous movie quotes, and are more likely to consume media via smartphones than to treat a TV or cinema screen as essential, underscoring the erosion of a shared movie canon across generations.【1news16】

• Industry research on viewing contexts notes there are now fewer tentpole, ‘water‑cooler’ TV/film moments overall, with much media consumption happening alone and fragmented across platforms, even as cinema still leads for the relatively rare co‑viewing occasions that remain.【0search1】

At the same time, there are notable counterexamples—Barbenheimer (Barbie and Oppenheimer’s 2023 double‑feature phenomenon) and youth‑driven events like A Minecraft Movie’s rowdy, cosplay‑filled screenings show that big films can still become cultural moments, including for Gen Z and Gen Alpha.【1search12】【1search14】【1news20】 But these are spikes against a backdrop where movies occupy a smaller share of young people’s attention and where shared conversation has migrated toward social platforms.

Overall, by late 2025 the evidence indicates that younger cohorts do still care about movies, but less and less in the traditional, culturally central, Monday‑morning‑water‑cooler sense compared with older generations, and that their primary cultural “center” has shifted to games and social media. That makes Chamath’s normalized prediction—long‑term decline in the cultural centrality of movies for younger generations—directionally right, even if his wording that his generation was the “last” to care is somewhat overstated.

By 2025 there is strong evidence that the “feature of the web” Chamath described—tools and business models that let creators own distribution and directly monetize their audience relationships—has in fact been built and adopted at scale.

• Direct‑to‑subscriber platforms. Substack provides publishing, payment, analytics, and design for subscription newsletters, podcasts, and video, letting creators send content directly via email rather than relying on social feeds. By November 2021 it already had over 500,000 paying subscribers, and by 2023–2025 it had grown to several million paid subscriptions and a $1.1B valuation, signaling a large, functioning ecosystem around direct audience monetization outside traditional platforms. (en.wikipedia.org) Notable independent media brands such as Bari Weiss’s The Free Press and Mehdi Hasan’s Zeteo were built around Substack-style direct subscriber relationships, then expanded into broader media companies—illustrating that creators can build durable businesses anchored in owned distribution. (en.wikipedia.org)

• Membership and fan‑funding at scale. Patreon enables creators to earn recurring revenue directly from fans via membership tiers. In 2024 alone, podcasters earned over $472M on Patreon across 6.7M+ paid memberships, with individual shows making tens or hundreds of thousands of dollars per month from listener support. (businessinsider.com) This is precisely the kind of direct monetization of the creator–audience relationship the prediction envisioned, and it is large enough to support many full‑time careers.

• Decentralized social graphs and portable audiences. Web3 projects such as Lens Protocol explicitly aim to let creators own their social graph and carry their audience between apps. Lens, launched by the Aave team on Polygon, is described as a smart‑contract based social graph “designed to empower creators to own the links between themselves and their community,” with user‑owned profiles/followers and modules that support direct monetization (e.g., paid “collects” of posts). (platoblockchain.net) This goes beyond just direct payments and directly tackles the dependency on centralized social platforms that Chamath was criticizing.

Collectively, these developments show that technology and business models now exist—and are in active, large‑scale use—that allow creators, influencers, and media personalities to (a) control their own distribution channels (email lists, membership platforms, decentralized social graphs) and (b) monetize those relationships directly via subscriptions, memberships, and on‑chain payments, rather than relying solely on centralized social media intermediaries. Since the prediction was that this would be figured out within the speakers’ lifetimes, and it has already been figured out and commercialized within just a few years, the prediction is best classified as right.

The prediction’s time horizon is explicitly “in a decade, 20‑somethings” from 2021—i.e., roughly 2031–2041. As of November 30, 2025, we’re only ~4.5 years in, so we’re far short of the lower bound of the 10–20 year window.

Substance-wise, the prediction is that a new generation will make it so that:

• Top creators (like Charli D’Amelio) can communicate with and monetize their audiences
• Without needing to go through intermediary platforms such as TikTok
• To the point where going through an intermediary “will not in the future make any sense” for such creators

Current reality (2025):

• Major influencers still rely heavily on intermediary social platforms (TikTok, Instagram, YouTube, Snapchat, Twitch) as their primary distribution and discovery channels; these platforms continue to mediate access to followers via algorithms, terms of service, and moderation/policy controls.
• There are growing direct‑to‑fan and multi‑platform monetization tools (email lists, Discord communities, Substack, Patreon, OnlyFans, Fanhouse, membership platforms, creator CRMs, and even emerging decentralized social protocols like Farcaster or Lens), but none has yet displaced the large social platforms as the default way top creators reach and grow their audiences.
• The structural power of large intermediaries remains intact: they still control feed ranking, recommendation, access, and often monetization rails. Big creators supplement with direct channels; they have not made intermediary‑controlled access “obsolete” yet.

Because:

1. The core structural shift Chamath describes (intermediary‑controlled access no longer making sense for top creators) has not clearly happened yet, and
2. We are still well before the earliest end of his 10–20 year window (2031),

there is not enough elapsed time to say whether the prediction ultimately proves correct or incorrect. It remains inconclusive (too early) rather than clearly right, wrong, or permanently ambiguous.

Evidence since 2021 shows that a large and growing number of platform‑native creators (YouTubers, TikTokers, Instagram influencers, etc.) now use big social platforms mainly as top‑of‑funnel and then move fans to independent, direct‑to‑fan monetization channels that they control more directly—very close to Chamath’s New‑York‑Times‑to‑Substack analogy.

Key points:

1. Mass adoption of independent, direct‑to‑fan monetization by social creators
   Patreon—originally built for creators whose audiences live on platforms like YouTube—has paid out more than $10 billion to creators and now supports over 25 million paid memberships, indicating that very large numbers of creators are monetizing fans off the big discovery platforms rather than relying solely on YouTube/TikTok ads or brand deals. (axios.com)
   A 2023–24 overview of the creator economy notes Patreon, Substack, OnlyFans and similar tools as the “go‑to” infrastructure for independent monetization, explicitly framing social platforms as discovery and these services as the business back end for creators. (andelek.com)

2. Creators deliberately funnel audiences off major platforms to owned channels
   Linktree’s creator report shows strong growth in outbound traffic from social profiles to Substack (+157% YoY) and Patreon (+33% YoY), which is exactly creators using their social‑media reach to move followers into direct subscription or community products they control. (orebic.plus)
   Business coverage of the creator economy repeatedly describes this shift as creators trying to “own their audience” instead of being dependent on algorithmic feeds, encouraging email lists, paid communities, and standalone membership sites. (joanwestenberg.medium.com)

3. Large, platform‑native creators building their own paid products, apps, and services
   Uscreen—a company whose core business is helping influencers launch their own subscription apps and sites—has helped creators earn over $600 million in subscription revenue and just raised a $150 million growth round, with its CEO noting that “more and more creators are looking to leverage and own their land, rather than rent that land via all those Big Tech companies.” This is direct evidence of sizable creators moving from YouTube/TikTok into owned distribution and monetization channels. (businessinsider.com)
   Creator‑owned streaming service Nebula, launched by and for YouTubers, has grown to about 680,000 subscribers and is described as the “largest creator‑owned streaming platform,” explicitly positioned as a place where established YouTube creators offer premium content outside YouTube’s ecosystem. (en.wikipedia.org)
   Membership platforms like Fanfix and OnlyFans report top TikTokers and other social‑native influencers earning six‑ and seven‑figure incomes by sending followers from Instagram/Snap/TikTok into subscription clubs hosted on those services—again, social for audience‑building, independent platforms for direct monetization. (en.wikipedia.org)

4. Scale and composition meet Chamath’s "significant number" bar
   By mid‑2020s, we see:

   • Billions of dollars flowing through Patreon, Substack, OnlyFans, Fanfix and similar platforms from fans to creators whose initial fame came on YouTube, TikTok, Instagram, Twitch, etc. (axios.com)
   • Dedicated tooling and funding (e.g., Uscreen’s $150M) specifically to help influencers build their own apps/sites for direct fan relationships, which wouldn’t exist at that scale if this were a marginal behavior. (businessinsider.com)

While most big creators still keep a presence on major platforms for reach, Chamath’s claim wasn’t that they would abandon those platforms—it was that they would build on them, then spin out into their own direct, Substack‑style channels for distribution and monetization. By 2025, that pattern is widespread and economically meaningful among large social‑native creators.

Given the data on payouts, subscriber counts, creator‑owned platforms, and the explicit strategic shift toward audience ownership, Chamath’s prediction is best judged as right.

Chamath frames this as something that will be figured out by crypto over the coming decades from 2021, not as a short‑term forecast. By November 2025 we are only about four years into that multi‑decade horizon, so it is too early to say definitively whether the prediction will ultimately prove right or wrong.

Empirically, there is clear progress in the direction he describes. Web3 social/creator platforms such as Lens Protocol and Farcaster give creators on‑chain identities, social graphs, and direct monetization via NFTs, pay‑to‑follow, and other smart‑contract mechanisms, partially shifting ownership and payments away from centralized intermediaries. (crowdfundinsider.com) The broader on‑chain creator economy (e.g., tokens and NFTs for creators, Zora/Base, etc.) is growing but remains small compared with the overall creator economy, which is still dominated by Web2 platforms like YouTube, TikTok, and Instagram and is projected to reach hundreds of billions of dollars mainly through traditional ad and sponsorship models rather than crypto rails. (tmcnet.com) At the same time, formal on‑chain reputation and credibility scoring systems are still active research areas and experimental deployments, not yet a mature, widely adopted standard for measuring and monetizing individual social value. (arxiv.org)

So as of late 2025, the direction of the prediction (crypto being used to help creators own distribution and link reputation to payments) is partially materializing, but the underlying problem is far from fully solved and the stated time window has not expired. Therefore the correct assessment today is that the prediction’s accuracy is still inconclusive.

Chamath’s claim was that, by mid‑2021, Big Tech was already in the “August” of its supremacy and that in the following years their dominance would start to wane under regulatory, competitive, and technological pressure. Looking at 2021–2025, the opposite has happened on most tangible metrics of power.

1. Market and profit dominance have increased, not waned.
The “Magnificent Seven” (Apple, Microsoft, Alphabet, Amazon, Meta, Nvidia, Tesla) expanded from about 20% of the S&P 500 a few years ago to roughly one‑third of the index by 2024, and about 37% by October 2025, an all‑time record in concentration. (gurufocus.com) They have driven a majority of S&P 500 gains in 2023–2024 and continue to produce outsized earnings growth relative to the rest of the market. (finance.yahoo.com) As of late 2025, several of these firms (Apple, Nvidia, and soon Alphabet) are in or near the $4T market‑cap club, and a small group of trillion‑dollar companies now accounts for about 41% of total S&P 500 value—much higher than a decade ago. (timesofindia.indiatimes.com) This is hard to reconcile with the idea that their supremacy has started to ebb.

2. In key technologies (cloud and AI), Big Tech is more central than ever.
In cloud infrastructure, AWS, Microsoft Azure, and Google Cloud control roughly 60–65% of global spending as of 2025, and that share has grown in the AI boom, with the “big three” capturing around two‑thirds of cloud spending. (indiekings.com) Hyperscalers—dominated by these same firms (plus Meta and Oracle)—account for more than 98% of AI infrastructure deployment, and are projected to spend well over $300–450 billion per year on data centers and AI hardware in the mid‑2020s, entrenching their control of compute and platforms. (gurufocus.com) Rather than opening space for smaller rivals, AI has largely reinforced Big Tech’s role as indispensable infrastructure providers.

3. Regulatory pressure has intensified, but has not yet materially reduced their power.
Regulators have clearly moved against Big Tech: the EU’s Digital Markets Act formally designated Alphabet, Amazon, Apple, Meta, Microsoft, and ByteDance as “gatekeepers” in 2023, imposing conduct rules on 20+ core platform services as of 2024. (digital-markets-act.ec.europa.eu) In the U.S., courts have found Google to hold illegal monopolies in both search and ad‑tech markets, the DOJ has filed a major antitrust suit against Apple, and the FTC has brought high‑profile cases against Amazon and Meta. (en.wikipedia.org) However, these actions have so far led mainly to behavioral remedies, ongoing appeals, and compliance obligations—not breakups or large structural loss of market share. The same firms still dominate search, mobile operating systems, social networking, app stores, digital advertising, and cloud.

4. Competition has emerged, but mostly around and through Big Tech rather than against it.
New or resurgent players like Nvidia, OpenAI, and various AI‑infrastructure challengers have grown rapidly, but Big Tech remains at the center of this ecosystem—as the primary cloud providers, chip customers, distribution channels, and investors. (investor.wedbush.com) The competitive landscape is more dynamic, yet it hasn’t translated into a broad-based erosion of the incumbents’ economic or platform power.

Overall, by late 2025 Big Tech faces more legal and political scrutiny, but its economic and technological dominance has, if anything, deepened. That makes the claim that we were already in the late‑summer phase of their supremacy in 2021—implying that their dominance would soon begin to ebb—look wrong based on the evidence to date.

Evaluation is against what actually happened with U.S. federal tax law in the 2021–2022 Biden period.

1. Capital gains tax rates
   Chamath predicted no movement on capital gains. Federal long‑term capital gains rates in 2021 and 2022 remained at 0%, 15%, and 20%, with only normal inflation adjustments to the income thresholds; Biden’s proposals to sharply raise the top rate on high earners never passed Congress. (seekingalpha.com)
   This sub‑prediction was essentially correct.

2. Corporate tax rate to 25% (not 28%)
   He predicted the regular federal corporate income tax rate would be increased from 21% to 25%, but not to 28%. In reality, the statutory federal corporate rate has stayed at 21% ever since the 2017 Tax Cuts and Jobs Act; neither the Build Back Better effort nor the Inflation Reduction Act changed that rate. (en.wikipedia.org)
   The Inflation Reduction Act did add a separate 15% corporate minimum tax on certain very large corporations, but that is layered on top of, and does not replace or raise, the 21% statutory rate. (en.wikipedia.org)
   This directly contradicts his prediction that the headline corporate rate would go to 25%.

3. “Really tighten the IP loophole” and make inversions/IP shifting to places like Ireland effectively impossible
   Biden and House Democrats proposed major international tax changes (e.g., higher and country‑by‑country GILTI, SHIELD, stronger anti–profit‑shifting rules) in early Build Back Better drafts, but those provisions were stripped out during negotiations. Multiple professional summaries of the final Inflation Reduction Act note that it omitted the international tax reforms needed to align U.S. rules with the 15% OECD global minimum tax and to more tightly police profit shifting. (withum.com)
   Separately, the United States never implemented the agreed OECD Pillar Two global minimum tax, and even later (in 2025) explicitly pulled back from that agreement, underscoring that no sweeping new regime was put in place to end use of low‑tax jurisdictions. (reuters.com)
   Existing post‑2017 rules (GILTI, BEAT, anti‑inversion provisions) already constrained some structures, but there was no 2021–2022 Biden legislation that made it “impossible” to move IP to low‑tax jurisdictions or to execute inversions in the strong sense Chamath described.

Overall judgment

• One narrow element (no federal capital gains hike in 2021–2022) was accurate.
• Two central elements—the predicted increase of the statutory corporate tax rate to 25% and the enactment of sweeping new IP/inversion‑blocking rules—did not occur.

Given that the main structural tax‑law changes he foresaw did not happen, the overall prediction is best classified as wrong.