E117: Did Stripe miss its window? Plus: VC market update, AI comes for SaaS, Trump's savvy move

As of November 30, 2025, it’s too early to definitively judge a prediction about how the 2023–2025 venture capital vintages will perform over their full life. VC funds typically take 8–12+ years to realize and distribute returns, and industry benchmarks like Cambridge Associates’ US Venture Capital Index still treat 2014–2022 as the main “meaningfully sized” vintages as of 2024; 2023+ funds are only beginning their investment and markup phase, not their exit phase. (cambridgeassociates.com)

Partial evaluation of the interest-rate premise is possible. The Fed did raise the target range to 5.25–5.50% by July 2023, and the effective federal funds rate stayed around 5.3% through late 2023 and the first half of 2024. (global-rates.com) However, the Fed began cutting rates in late 2024, with the target range down to 4.25–4.50% by December 2024 and then to 4.0% and 3.75–4.0% after further cuts in September and October 2025. (theglobalstatistics.com) So rates did not stay around 5.5% for many years; they were at that level for roughly 12–15 months before a clear easing cycle.

For the VC performance outcome, only early, noisy indicators exist. Carta’s Q2 2025 data show 2023-vintage funds are about 58% deployed, with significant dry powder remaining, and that the very best 2023 funds (90th percentile) have strong IRRs (30%+), largely due to AI-driven markups—but median performance for recent vintages is weak, and power-law outcomes are concentrated in a minority of funds. (carta.com) AngelList/LP commentary based on large datasets of 2017–2024 funds reports median IRR for the 2023 vintage around –1% and median TVPI for 2021–2023 vintages hovering just under 1.0x, with DPI essentially zero—consistent with very early-stage, largely unrealized performance rather than clear outperformance. (nehaldesai.com) Broader benchmarks show the US VC index delivered –3.4% in 2023 and only a modest rebound in 2024, with private-market funds (including VC) underperforming US public equities over multiple horizons. (cambridgeassociates.com)

In summary:

• The mechanical rationale (rates at ~5.5% "for the foreseeable future") has only partially held: rates peaked there but have already been cut materially.
• The core claim—that 2023–2025 VC vintages will, in aggregate, be unusually strong “power-law years” versus surrounding vintages—concerns outcomes that will not be observable until well into the 2030s. Current data show mixed, highly preliminary results, not definitive outperformance.

Because the main outcome variable (full-cycle vintage-year returns for 2023–2025 funds) cannot yet be known, the prediction must be rated "inconclusive" rather than right or wrong at this time.

It is too early to evaluate this prediction.

Friedberg is talking about the current and upcoming early‑stage venture cohorts in a high‑rate environment, i.e., funds investing roughly in the early‑ to mid‑2020s. Standard VC and early‑stage fund lives are ~10–12 years, and performance (net IRR and net multiple) is typically only meaningfully known well after year 7–10, once exits and write‑downs have largely played out. Public benchmarks (e.g., Cambridge Associates, PitchBook, Preqin) only have early, highly volatile marks for 2020–2024 vintages, and no one can yet say what their final net annual returns will be, nor the final distribution of winners vs. losers across funds.

Because:

• 2020–2025 vintages are only 0–5 years old as of late 2025, and
• VC performance for a vintage can change dramatically between year 3 and year 10,

we cannot yet know whether the industry‑wide average will settle near 6% real annually or whether ~80% of funds will end up with negative real returns and only a small minority will be big winners. Therefore the prediction’s correctness is inconclusive (too early to tell).

As of November 30, 2025, the prediction cannot yet be evaluated. The funds that raised the roughly $1T of venture capital during the 2020–2022 boom are only about 3–5 years into their lives, and venture funds typically take at least 7–10 years—and increasingly 15–20 years—to fully return capital to LPs, with DPI usually remaining well below 1.0x in the first 5–7 years. (phoenixstrategy.group) Public data on recent vintages shows that 2021–2022 VC funds have very low realized DPI so far (often around 0.1–0.3x or even zero), and many 2021 funds have not yet made any distributions at all, underscoring that most value is still unrealized. (equitybee.com) While we do know that (a) U.S. VC assets under management surpassed roughly $1.1T by 2022 and global fundraising/dry powder reached record levels—consistent with Chamath’s “about $1T raised” framing—and (b) exits and valuations have fallen sharply since the 2021 peak (global VC exit value dropped from about $1.4T in 2021 to roughly $309B in 2022), it is impossible at this stage to say whether the final DPI on this capital will converge to ~1.6x paid-in (≈$1.6T in distributions) or end up materially higher or lower. (nvca.org) No comprehensive study as of 2025 projects ultimate distributions for the 2020–2022 VC fundraising cohort in the precise way Chamath describes, and the realization horizon he himself assumes (“over many cycles” / “out of the trillion dollars we’ve raised”) is on the order of 10+ years, which has not yet elapsed. Therefore, the quantitative prediction about ending up with only ~$1.6T in distributions and ~$3.9T in lost paper value remains untestable today.

Available evidence by November 30, 2025 shows strong movement in the direction of Sacks’s prediction, but it does not clearly confirm that “a large share” of customer-support inquiries overall have already “effectively disappeared” due to in‑product AI assistants.

What has clearly happened (supports the direction of the prediction):

• Major SaaS and productivity vendors (e.g., Microsoft 365 Copilot, Google Workspace Duet/Gemini, Salesforce Einstein Copilot, HubSpot AI, Intercom Fin, Zendesk AI agents) have embedded AI assistants directly into their products to answer help questions, draft responses, and resolve common issues in‑product rather than routing to humans or separate ticket portals. These deployments are specifically marketed as reducing escalations to human agents and deflecting tickets.
• Case studies and vendor reports (from companies like Intercom, Zendesk, Ada, Forethought, and others) describe substantial ticket deflection (often 20–40%+ of incoming requests for certain workflows or problem types) when AI chatbots/assistants are deployed inside the product experience, reducing contact volume that reaches human support.
• Many startups now build products with AI help directly in the UI (tooltips, chat sidebars, command palettes) so users get real‑time, contextual assistance rather than filing tickets at all. This matches the mechanism Sacks described: “assistant gets built into the tool directly, so you never get to the point” of contacting support.

Why the overall truth value is ambiguous rather than clearly right or wrong:

• There is no broad, independent, industry‑wide data yet showing that across software products as a whole a “large share” of support inquiries have already disappeared. The best evidence is fragmented vendor case studies and marketing claims, which are not a reliable aggregate measure.
• Impact is highly uneven by sector and company size. Large tech/SaaS firms and AI‑forward startups have implemented substantial in‑product AI support; many traditional enterprises, regulated industries, and smaller vendors still rely heavily on human agents and conventional ticket systems.
• Overall customer support employment and spending have not visibly collapsed. While some organizations report efficiency gains and slower headcount growth in support, there isn’t strong macro‑level evidence that human‑handled support volume has already dropped by the kind of “large share” implied by the prediction.
• The prediction is somewhat qualitative (“a lot… just go away,” “large share”), so judging whether the threshold has been met requires a quantitative benchmark that the public evidence does not supply.

Conclusion

By late 2025, product‑embedded AI assistants are clearly handling many more customer questions in‑product and reducing the need for traditional support interactions in a growing portion of the software industry. However, because the scale of this effect across the whole market cannot be reliably measured and varies widely by company and sector, we cannot definitively say that a large share of all such inquiries has already ‘effectively disappeared.’ The claim is therefore best scored as ambiguous rather than clearly right or wrong based on currently available evidence.

Chamath claimed that breast‑cancer surgeries have about a 30% error rate (typically interpreted as positive margins/re‑operations after breast‑conserving surgery) and that AI‑driven, laser‑guided robotic surgery that characterizes tumors with 100% accuracy could drive that error rate to essentially zero.(podscripts.co) The 30% figure is broadly consistent with historical data: population studies report re‑operation rates around 20–30% after breast‑conserving surgery and global estimates put positive margins in breast cancer at roughly 20–70%.(insightplus.mja.com.au) As of 2024–2025, however, typical re‑excision rates in large real‑world cohorts are still on the order of 10–20% (for example, ~17–18% in US Medicare data, 15–16% in national registries, and ~12–13% in some optimized single‑center series), meaning a substantial error rate persists.(pubmed.ncbi.nlm.nih.gov) New adjunct technologies—including devices like MarginProbe and confocal scanners for intraoperative margin assessment—can cut re‑operation rates from ~25–30% down to about 10–11%, but they still do not achieve near‑zero error in broader practice.(academic.oup.com) AI and robotics are emerging: deep‑learning systems for intraoperative margin evaluation and early multicenter series of robot‑assisted breast‑conserving surgery show promise (with very low re‑operation in small, selected cohorts), but these are pilot‑stage, not widely deployed, and they do not yet demonstrate system‑wide elimination of surgical error.(arxiv.org) Because Chamath framed this as something that "can and should" go to zero without a concrete time horizon, and current evidence shows progress but not anywhere close to universally near‑zero error by 2025, the prediction has neither come true nor been clearly falsified yet; it remains too early to judge definitively, so the result is best classified as inconclusive.

As of late 2025, the prediction is only partially playing out and it’s too early to declare it clearly right or wrong.

Pricing: Direct‑to‑consumer full‑body MRI providers like Prenuvo and Ezra still commonly charge well above $1,000 per full‑body scan (e.g., Prenuvo around $2,499+; Ezra $1,350–$2,695 in many offerings). (bodyspec.com) However, in May 2025 Function Health’s acquisition of Ezra introduced a full‑body MRI scan marketed at $499, explicitly framed as a big drop from prior ~$1,500 pricing. (prnewswire.com) So the price floor Jason imagined ($500–$1,000) is just beginning to appear, but is not yet the norm across the market.

Adoption: Usage remains niche. Prenuvo reports roughly 150,000 scans total since 2018, which is tiny relative to the general population, and coverage is largely limited to affluent or highly health‑conscious consumers. (axios.com) Major medical organizations and mainstream health outlets still do not recommend full‑body MRI as a routine screening tool for asymptomatic people, citing high cost, false positives, and lack of proven outcome benefit. (verywellhealth.com) That is far from “everybody’s doing it.”

AI‑driven early detection: This part of the vision is materializing. Companies such as Ezra use FDA‑cleared AI (e.g., Ezra Flash and related tools) to speed scans, clean images, and assist in analysis and reporting, explicitly positioning the service as AI‑enabled early cancer/condition detection. (ezra.com) Function Health + Ezra also frame their $499 scan as combining large‑scale lab and imaging data to enable predictive, preventive care. (prnewswire.com)

Because the price trend and AI usage are moving in the predicted direction, but widespread general‑population adoption plainly has not happened and no clear time horizon was given, the fairest assessment as of 2025‑11‑30 is that the prediction’s ultimate outcome is still inconclusive (too early to tell).

The prediction is about the outcome of Gonzalez v. Google LLC at the U.S. Supreme Court and its implications for Section 230 protections for recommendation algorithms.

What actually happened:

• The Supreme Court issued its decisions in Gonzalez v. Google LLC and the related case Twitter, Inc. v. Taamneh on May 18, 2023.
• In Twitter v. Taamneh, the Court unanimously held that the plaintiffs had not adequately stated a claim for aiding and abetting under the Anti-Terrorism Act, resolving the liability questions in favor of the tech platforms.
• On the same day, the Court issued a short, unsigned opinion in Gonzalez v. Google that vacated and remanded the case to the Ninth Circuit in light of the Taamneh ruling, rather than narrowing Section 230. The Court explicitly declined to reach the plaintiffs’ broad theory that recommendation algorithms should fall outside Section 230, and it did nothing to restrict Section 230’s application.
• Commentators and legal analysts widely interpreted this pair of decisions as a substantial win for Google and other platforms, because the Court neither imposed liability in Taamneh nor cut back Section 230 in Gonzalez. The practical effect was to preserve the broad immunity status quo, including for algorithmic recommendations, pending any future, more targeted case.

These outcomes are documented in the Supreme Court’s opinions and summarized in legal analyses, all of which note that the Court avoided narrowing Section 230 and effectively left existing protections in place while ruling for the platforms on the underlying liability theories.

Match to the prediction:

• The prediction: that in Gonzales v. Google the Supreme Court would "rule in favor of Google" and that Big Tech would "win this one," with broad Section 230 protections preserved for recommendation algorithms.
• The reality: The Court did not adopt the plaintiffs’ theory, did not narrow Section 230, and issued decisions in Taamneh and Gonzalez that are broadly regarded as wins for Google and similar platforms, maintaining the de facto protection of recommendation algorithms under Section 230.

While the Court’s procedural posture (vacate and remand rather than a merits ruling squarely expanding 230) is more technical than the podcast’s simplified framing, in practical and legal-effect terms the prediction was directionally accurate: Google and Big Tech prevailed, and broad Section 230 immunity—including for recommendation algorithms—remains intact.

Therefore, the prediction is best classified as right.

As of 30 November 2025, the Russia‑Ukraine war is still ongoing, and its ultimate outcome for the United States—in terms of total cost, duration, clarity of objectives, and long‑term strategic consequences—remains uncertain.

Key points:

1. War not concluded
   The Russia‑Ukraine conflict, begun with Russia’s full‑scale invasion on 24 February 2022, has not reached a stable political settlement or clear end state. Front lines have fluctuated, periodic offensives and counteroffensives continue, and there is no agreed peace treaty or armistice that would let analysts definitively tally costs or strategic outcomes relative to the 1991 Gulf War.

2. Costs and duration are still moving targets
   – U.S. military and financial assistance to Ukraine has surpassed many tens of billions of dollars and continues to be debated and extended in Congress. Future aid levels, reconstruction commitments, and indirect economic costs (e.g., energy, supply chains, deterrence posture in Europe) could still rise or fall substantially.
   – The duration dimension is inherently open‑ended: while the 1991 Gulf War had a clear, short combat phase (January–February 1991) and a well‑defined coalition objective (expel Iraqi forces from Kuwait), the Russia‑Ukraine war does not yet have either a clear terminal date or agreed political outcome.

3. Strategic consequences are still unfolding
   Medium‑ and long‑term effects for the U.S.—on NATO cohesion and expansion, Russia’s military and economic strength, China’s calculus on Taiwan, global defense spending patterns, and the future of the rules‑based order—are still in flux. Some effects (e.g., an enlarged and more unified NATO) can be read as strategically positive for Washington; others (e.g., sustained higher defense and energy costs, risk of escalation, strain on U.S. munitions stockpiles) could prove negative. Which side ultimately dominates the ledger is not yet knowable.

4. The prediction is about the ultimate comparison
   Sacks’ claim was that this war “is going to turn out much worse than the Iraq war [Gulf War] did in 1991” for the U.S. overall. Because the Russia‑Ukraine war’s end state, total cost, and full strategic impact are not yet determined, we cannot conclusively say today that this has turned out worse—or that it definitively won’t.

Given these uncertainties and the ongoing nature of the conflict, the fairest assessment is that the prediction’s truth value cannot yet be determined, so the outcome is inconclusive (too early) rather than clearly right, wrong, or permanently ambiguous.

Evidence since early 2022 supports Sacks’s core prediction that sanctions would inflict a modest macro‑level GDP hit rather than a “major collapse,” and that continued trade with non‑Western countries would cushion the blow.

1. Observed GDP hit was low single‑digit, not a collapse

   • Russia’s GDP fell by about 2.1% in 2022, far less than the double‑digit collapse many early forecasts predicted.(en.wikipedia.org)
   • By 2023 the economy rebounded with ~3.6–4.1% real growth, with continued positive growth reported into 2024, again inconsistent with the notion of a post‑sanctions collapse.(ru.wikipedia.org)
     These figures line up with Sacks’s “3–4% hit” characterization and clearly do not describe a major economic collapse.

2. Sanctions underperformed early expectations and were cushioned by non‑Western trade

   • Analyses have repeatedly noted that sanctions “might not be having the impact the West had hoped”, with the IMF and others highlighting Russia’s continued output and growth despite restrictions.(the-independent.com)
   • Reporting on the war’s economic impact stresses Russia’s surprising resilience, with growth resuming in 2023–24, heavily supported by redirected energy exports and trade with countries that did not join Western sanctions (e.g., China and India).(theguardian.com)
   • The IMF and other observers explicitly describe Russian trade being redirected from sanctioning to non‑sanctioning countries, matching Sacks’s reasoning that “enough other countries [are] willing to do business with them.”(the-independent.com)

3. Sanctions are still harmful, but that does not contradict the prediction
   Sanctions have clearly hurt Russia’s long‑term prospects—higher inflation and interest rates, fiscal strain, technology import restrictions, and growing dependence on war spending and a narrow set of trading partners.(en.wikipedia.org) But the prediction we are evaluating is narrower: whether sanctions would cause a major collapse versus a modest, low‑single‑digit GDP hit with continued trade via non‑Western partners. On those specific points, post‑2022 data and mainstream analyses align with Sacks’s view.

Given that Russia saw only a small initial GDP contraction, returned to growth, and avoided a macroeconomic collapse largely by re‑routing trade to non‑sanctioning countries, the prediction is best classified as right.

By November 5, 2024, enough time had passed to evaluate this prediction, and subsequent events contradict it on multiple key points.

1. No U.S. recession tied to the Ukraine war before the 2024 election.
   The National Bureau of Economic Research’s business‑cycle chronology shows the last U.S. recession ended in April 2020, with the most recent peak in February 2020 and no later recessions dated as of 2024–2025. (nber.org) GDP growth remained positive in 2023 and 2024 (e.g., U.S. real GDP grew about 2.9% in 2023 and 2.8% in 2024, with Q4 2024 annualized growth at 2.3%), indicating a slowdown but not a significant recession. (barrons.com) While some European economies, such as Germany’s, experienced Ukraine‑war‑related recessions due to energy shocks, this did not translate into a U.S. recession. (en.wikipedia.org) So the antecedent of the conditional (“a significant U.S. recession attributable to the Ukraine war before 2024”) never occurred.

2. Trump nevertheless won the 2024 election.
   Donald Trump defeated Kamala Harris in the 2024 U.S. presidential election, winning 312 electoral votes to 226 and carrying all major swing states. (en.wikipedia.org) This directly contradicts the speaker’s assertion that Trump had essentially no other way to win besides the hypothesized recession‑based strategy.

3. Trump’s actual campaign message centered on immigration, crime, and “Bidenomics,” not primarily on blaming a war‑induced recession or the military‑industrial complex.
   Analyses of Trump’s 2024 campaign show he emphasized immigration, the southern border, crime, and inflation/Biden’s economic record as his dominant themes. (washingtonpost.com) Trump and his campaign communications repeatedly framed the election around a “broken border,” violent crime by undocumented immigrants, and the failures of “Bidenomics,” as illustrated by official campaign emails and rally transcripts. (donaldjtrump.com)

   Trump did criticize the cost of aid to Ukraine and claimed he could quickly end the war, but reporting on his rhetoric and his own post‑election interview with TIME depict this as part of a broader foreign‑policy posture, not as a central, economic blame‑the‑war campaign built around an explicit anti–military‑industrial‑complex frame. (politico.com)

4. How this maps to the normalized prediction.
   The normalized claim is: if the Ukraine war drives a major U.S. recession, Trump’s main viable path to victory is to blame that downturn on the war and the foreign‑policy establishment, running an anti‑military‑industrial‑complex campaign—and that he has no other realistic route to win. In reality: (a) the war did not cause a U.S. recession before the election; (b) Trump did win; and (c) he did so with a message centered on immigration, crime, and inflation, not the posited recession‑and‑Ukraine narrative. This shows the forward‑looking judgment that he had “no other way of winning” besides that scenario was falsified by events.

Because both the key condition (a war‑induced U.S. recession) failed to materialize and Trump won via a different message than the one described as his only viable path, the prediction is best classified as wrong.

As of 30 November 2025, Russia has not used nuclear weapons in the Ukraine war, nor has a scenario of Russia facing an imminent, clearly decisive defeat and then choosing between defeat and nuclear escalation actually occurred. The war remains ongoing, with front lines shifting and Russia maintaining control of significant occupied territory, so the specific test condition—“rather than accept clear defeat”—has not been met.

The prediction is about Russia’s willingness to escalate up to and including nuclear use and about its supposed enduring escalatory dominance over NATO. Those are strategic-intent claims, not directly observable facts, and so far they have neither been clearly validated (e.g., via nuclear use or overt nuclear brinkmanship at the moment of clear defeat) nor clearly falsified (e.g., Russia accepting an obvious, final military defeat without serious escalation). Therefore, given the state of the war and absence of nuclear use as of late 2025, the correctness of this prediction cannot yet be determined.

Evidence since the Feb 2023 episode shows China and Russia have indeed moved into an increasingly close, quasi‑allied alignment broadly opposed to U.S. interests, even if they stop short of a formal treaty alliance.

Key points:

• Deepening strategic partnership: Xi and Putin have repeatedly reaffirmed their pre‑2022 “no limits” partnership. In May 2024 Putin’s state visit to China was explicitly framed as a “new era” visit that underscored a deepening strategic partnership, coinciding with the 75th anniversary of diplomatic relations.(en.wikipedia.org) Analyses describe this as a comprehensive partnership and strategic cooperation “just short of a conventional alliance.”(isdp.eu)

• Bloc‑like opposition to U.S. leadership: Xi and Putin now routinely position their partnership as an alternative pole to the U.S.-led order. A 2025 summit in Moscow saw them condemn U.S. tariffs, sanctions, and what they called Washington’s “dual containment” of Russia and China, while presenting their relationship as a stabilizing counterweight.(washingtonpost.com) A U.S. congressional commission likewise warns that Xi is building an “alternative world order” centered on China and aligned with anti‑democratic states like Russia and North Korea.(axios.com) This is consistent with Sacks’s claim of a new axis‑like bloc opposed to U.S. interests.

• Ukraine war and U.S. policy as a binding force: Western (especially U.S.) support for Ukraine and sanctions on Russia have made China an economic and political lifeline for Moscow—via surging trade, energy purchases and financial links—while Beijing avoids condemning the invasion.(aljazeera.com) NATO’s 2024 summit statement explicitly accuses China of becoming a “decisive enabler” of Russia’s war through large‑scale support for its defense industrial base, referring directly to their “no limits” partnership.(foreignpolicy.com) That is precisely the dynamic Sacks described, where the Ukraine conflict and U.S. responses pull Beijing and Moscow closer together.

• Quasi‑allied, even if officially “not an alliance”: Beijing still publicly insists that China–Russia relations are not aimed at any third party and do not constitute a formal alliance. Yet Xi simultaneously affirms the long‑term, “no limits” nature of the relationship and emphasizes that the two are “good neighbours that cannot be moved apart,” while analysts note they are increasingly “challeng[ing] a US‑led global order” together.(asiaone.com) Think‑tank and economic analyses highlight real limits and asymmetries in the partnership, but still characterize it as a durable, strategically aligned front against Western pressure.(capitaleconomics.com)

Given this trajectory between early 2023 and late 2025, the core of Sacks’s prediction—that U.S. policy around Ukraine and concern over China’s support for Russia would push Beijing and Moscow into an increasingly close, quasi‑allied bloc opposed to U.S. interests, undermining the old Cold War aim of keeping them apart—matches observed reality. The partnership is not a formal alliance, but it functions in practice much like the “new axis bloc” he warned about, so the prediction is best judged right.

Jason argued that Quora (via Poe and its Q&A corpus) would become the number one AI application company going forward, overtaking incumbents like ChatGPT in practical consumer relevance, and that Reddit could likewise become a major AI answers platform if it launched a chatbot.

As of late 2025, market data shows the opposite:

• ChatGPT remains the dominant consumer AI app by a wide margin. A World Bank–based analysis of generative AI tools for March 2024 found ChatGPT had ~2.34 billion visits, accounting for 82.5% of web traffic among 40 tools; Quora’s Poe ranked only third at ~43.4 million visits—orders of magnitude smaller. (visualcapitalist.com)
• By 2025, multiple analytics sources estimate hundreds of millions of monthly active users for ChatGPT’s app alone (e.g., ~546M MAU in April 2025 and ~769M MAU by October 2025), with ChatGPT having about 5x more users than any other AI app. (backlinko.com) ChatGPT is also reported at roughly 700 million weekly users across platforms. (techradar.com)
• Rankings of the most popular AI apps worldwide in 2024–2025 consistently place ChatGPT at #1 in MAUs, downloads, and in‑app revenue, with Quora/Poe absent from the top-tier list of global leaders. (backlinko.com) There is no evidence that Quora has surpassed ChatGPT or become the leading AI application company.

On Reddit:

• Reddit did exactly what Jason envisioned and launched an AI chatbot, Reddit Answers, starting tests in the US in December 2024 and expanding to India and other regions in 2025. It surfaces conversational answers based on Reddit posts and comments. (investing.com)
• However, available market and traffic reports on top generative AI tools and apps do not list Reddit Answers among the major global AI chatbots; instead, Reddit’s primary AI-related role appears to be as a content/licensing provider to companies like OpenAI and Google, and as a plaintiff in scraping lawsuits against Anthropic and Perplexity. (siliconindia.com) This indicates it has not emerged as a dominant, consumer-facing AI answer platform on par with ChatGPT.

Given:

• ChatGPT’s entrenched and growing dominance in users, traffic, and revenue;
• Poe’s and Quora’s status as secondary players without any sign of surpassing ChatGPT; and
• Reddit’s chatbot remaining a feature mainly for navigating Reddit content rather than a top-tier global AI app,

Jason’s core prediction—that Quora would become the number one AI player in consumer AI apps, overtaking ChatGPT, and that Reddit would likewise become a leading AI answers platform once it launched a chatbot—has not materialized by late 2025.

Therefore, the prediction is wrong.