E99: Cheating scandals, Twitter updates, rapid AI advancements, Biden's pardon, Section 230 & more

By roughly two years after the October 2022 podcast (i.e., by October 2024), Hans Niemann had not become the world’s number one over‑the‑board (classical) chess player. On the September 2024 FIDE rating list he was ranked 16th in the world with a classical rating of 2733, while Magnus Carlsen remained world #1.(chess.com) As of the November 1, 2025 FIDE list, Carlsen is still #1 at 2839 and Niemann is ranked 20th with a rating of 2729; he has never held the #1 spot.(en.wikipedia.org)

However, the prediction was explicitly conditional: “if he’s not cheating, then … he should rapidly become the world’s number one.” Whether Niemann has ever cheated over‑the‑board in elite classical chess has not been definitively established by any authoritative body. The high‑profile Carlsen–Niemann dispute ended in an August 2023 confidential settlement between Niemann, Magnus Carlsen, and Chess.com that included Niemann’s reinstatement and explicitly involved no admission of wrongdoing by any party, with Carlsen stating there was no determinative evidence Niemann cheated in their 2022 game.(grokipedia.com) This resolution leaves the truth of the antecedent ("he’s not cheating") unresolved.

Because the forecast is framed as a conditional on an empirically unsettled premise, we cannot determine its truth value from observed outcomes alone. If one ignores the condition, the implied claim that Niemann would become world #1 within a couple of years is clearly false; but taking the stated conditional form seriously, the prediction is best classified as ambiguous rather than definitively right or wrong.

Elon Musk’s acquisition of Twitter closed on October 27, 2022, when he completed the $44 billion deal and took the company private. (techcrunch.com)

The original merger agreement from April 25, 2022 set the purchase price at $54.20 per share in cash, valuing Twitter at about $44 billion. The deal ultimately closed on these same financial terms, without a renegotiated lower price. (cnbc.com)

The podcast was released on October 7, 2022, and the closing on October 27, 2022 is 20 days later—well within a reasonable interpretation of “the next few weeks.” Since the acquisition did close within a few weeks of early October and on essentially the originally agreed terms, Chamath’s prediction was accurate.

As of late November 2025 (a bit over three years after October 2022), several of the cited “vertical” problems are indeed at or beyond human level (e.g., chess/Go via AlphaZero/AlphaGo and related systems; poker via Libratus/Pluribus, which achieved superhuman performance against top pros years earlier).(en.wikipedia.org) However, the prediction was that all major vertical AI problems, including self‑driving and GPT‑style language, would be effectively solved in short order.

Self‑driving is clearly not “effectively solved.” GM has shut down its Cruise robotaxi unit after years of large losses and a serious pedestrian-dragging incident, citing long timelines and high costs.(apnews.com) Waymo has had to issue large software recalls and remains under NHTSA investigation after multiple crashes with barriers and other low‑impact incidents, and its service is still limited to specific cities and conditions rather than general, everywhere, all‑weather autonomy.(nypost.com) Tesla’s Robotaxi launch in Austin requires safety monitors, has triggered an NHTSA investigation for erratic driving behavior, and its Full Self‑Driving remains a supervised driver‑assistance system facing ongoing regulatory and legal scrutiny, not a solved, unsupervised Level‑4/5 system.(en.wikipedia.org) Collectively, this shows that one of the flagship verticals Jason referred to—self‑driving—remains far from broadly, reliably solved.

Other verticals he implicitly lumps in, like image generation and GPT‑style language models, have made large strides but still show clear unsolved issues. 2025 overviews of image generation note the field as "mature" for many professional uses, yet different tools still trade off quality, control, speed, and prompt alignment rather than representing a single, fully‑solved capability.(superallen.org) Microsoft’s rollback of a DALL·E‑3 variant in Bing due to significant quality regressions underlines that even this narrow vertical continues to exhibit non‑trivial failure modes.(theverge.com) Large language models in 2024–2025 still hallucinate and misreason often enough that entire research threads and benchmarking efforts are devoted to quantifying and mitigating these errors, including in law, science, code, and embodied agents—evidence that the "GPT‑3‑type" language problem is not effectively solved either.(academic.oup.com)

Because at least one of the central example domains (self‑driving) remains very clearly unsolved in practice, and others (LLMs, image gen) still have widely acknowledged, fundamental shortcomings rather than being "done," the generalized claim that "we're going to solve every vertical AI problem in short order" has not come true within the few‑year window he implied. Therefore, the prediction is best classified as wrong.

The ISIS/YouTube cases pending as of October 2022 were Gonzalez v. Google LLC (about YouTube recommendations and ISIS) and the related Twitter, Inc. v. Taamneh Anti‑Terrorism Act case.

On May 18, 2023, the Supreme Court:

1. Twitter v. Taamneh (9–0 for Twitter and other platforms)
   The Court unanimously rejected the plaintiffs’ claims that Twitter, Google, and Facebook had "aided and abetted" ISIS under the Anti‑Terrorism Act and held that the allegations failed to state a claim. (supreme.justia.com) This was a win for the platforms, not a narrowing of their liability.

2. Gonzalez v. Google (no ruling cutting back Section 230)
   In Gonzalez, the Court explicitly declined to address whether Section 230 protects YouTube’s recommendation algorithms, vacated the Ninth Circuit’s judgment, and remanded for reconsideration in light of Taamneh—"punting" on Section 230 and leaving its broad protections in place for the time being. (cnbc.com)

Media and legal analyses at the time uniformly characterized these outcomes as avoiding limits on Section 230 and effectively preserving broad immunity for big tech, not as rulings "against" Section 230 protections. (cnbc.com)

Chamath’s prediction was that the Court would go 9–0 against maintaining broad Section 230 protections for big tech in this context (i.e., a unanimous ruling finding no Section 230 protections here). In reality, the Court (a) ruled unanimously for the platforms on non‑230 grounds in Taamneh and (b) declined to weaken or even opine on Section 230 in Gonzalez. The net effect was to preserve broad Section 230 protections, not to reject them.

Therefore, the prediction is wrong.

Available data through late 2025 shows TikTok has not surpassed YouTube in overall global size or total engagement, and in fact remains substantially smaller.

User base / reach

• DataReportal-based estimates compiled by Backlinko put YouTube at about 2.5–2.54 billion active users in 2024–2025, nearly half of all internet users. (backlinko.com)
• Multiple 2025 summaries (DataReportal / Statista–based) place TikTok’s monthly active users around 1.6 billion, i.e., well below YouTube’s user base. (dreamgrow.com)
• Data.ai–indexed comparisons in the Digital 2025 report show YouTube as the most used social platform globally, with an active-user index of 100 versus TikTok around 48–48.3—meaning TikTok’s active user base is less than half of YouTube’s. (scribd.com)
• A separate social-media statistics compilation, based on DataReportal, explicitly states that TikTok’s active users are about 44.7% of YouTube’s, with YouTube ad reach ~2.5B vs. TikTok ~1.6B. (electroiq.com)
• Reuters reporting in September 2025 describes TikTok as having “over 1 billion” monthly users globally (and 200M in Europe), again clearly below the 2.5B+ range consistently cited for YouTube. (reuters.com)

Total time / engagement at platform scale

• The Digital 2025 Global Overview notes that YouTube still accounts for the greatest overall share of time spent on social platforms, with total time on YouTube exceeding TikTok and Instagram combined, even though TikTok has very high per-user engagement. (scribd.com)
• MarketWatch’s 2025 profile of YouTube’s 20th anniversary similarly describes YouTube as a media powerhouse with over 2.7 billion monthly active users and more than 1 billion hours watched per day on TVs alone, underscoring its continued dominance in total viewing time and revenue (about $36.1B in ad revenue in 2024). (marketwatch.com)

Was TikTok already “almost as big” in 2022?

• An August 2023 research paper by Touchstorm / VideoAmigo, doing an apples-to-apples comparison of global audiences, found that in 2023 YouTube had 2.3× more global unique active users than TikTok and roughly 2× the mobile audience. (newswire.com) Given TikTok user counts and bans in India, that ratio implies TikTok in 2022 was not “almost” as big as YouTube either.

Where TikTok does lead

• App Annie / data.ai and subsequent coverage show TikTok surpassed YouTube in average watch time per user on Android phones in the US and UK by 2020–2021 and continues to have very high per-user daily minutes. (thenationalnews.com) That supports a narrower claim about per-user engagement in certain markets, but not about global overall scale.

Assessment vs. the normalized prediction Friedberg’s normalized prediction was that TikTok was already nearly as big as YouTube in October 2022 and would soon surpass YouTube in overall size/scale (usage or engagement). As of late 2025, across the main global metrics that reasonably capture “overall size/scale” (total active users, ad reach, and total time spent), YouTube remains clearly larger—often by a factor of ~1.5–2× or more. TikTok has grown and leads on some per-user engagement measures, but it has not overtaken YouTube in overall global scale.

Therefore, the prediction is wrong.