Summary

TikTok has become an unprecedented legal-operational hybrid: banned by PAFACA yet usable for Americans who already installed it. Oracle and Akamai accepted President Trump’s promise of non-enforcement and restored service, while Apple and Google kept ByteDance apps out of their stores because statutory exposure is $5,000 per user across roughly 170 million TikTok users. Without store access, TikTok cannot distribute routine feature, bug, or security updates: “It is Schrödinger’s app.”
TikTok is now a bargaining chip in a wider US-China negotiation over tariffs, ownership, and executive power. Trump delayed enforcement for 75 days and floated giving the United States half of an asset supposedly worth “like, a trillion dollars” with a permit; meanwhile, China signaled for the first time that it might accept a divestiture as Trump linked TikTok’s fate to threatened tariffs. Kevin expects another extension rather than a completed deal; Casey predicts a deal, potentially enabled by China’s tariff calculations.
Stargate captures Silicon Valley’s emerging Trump playbook: give the president credit for private investment already underway, then seek a permissive policy environment. OpenAI, Oracle, and SoftBank presented a $100 billion AI-infrastructure venture—potentially reaching $500 billion—as something that “couldn’t have” happened without Trump, although construction had begun and no government funding was planned. The tangible benefit for AI companies was Trump’s repeal of Biden’s executive order, leaving what Kevin called essentially “no regulatory infrastructure” governing development.
TRUMP and MELANIA converted presidential attention into speculative assets with paper valuations above $10 billion and unusually direct family economics. Trump-related entities reportedly control about 80% of TRUMP’s supply and 35% of MELANIA’s; the former’s tokens unlock over roughly three years, limiting an immediate rug pull but not transaction-fee income. A Coinbase director estimated that the Trump family had already earned $58 million in fees by Saturday, while critics warned that anonymous purchases create a “new and worrying avenue of political corruption.”
The meme-coin boom is exposing the gap between crypto’s promised financial utility and its realized attention economy. Trump-supporting investor Nick Carter called the launches “preposterous” and said their creators were “plumbing new depths of idiocy,” yet Coinbase and Kraken rushed to list the coins and monetize trading. Kevin still predicts that a newly issued meme coin will briefly reach a $100 billion market cap during 2025 before crashing.
Pump.fun has collapsed the cost of manufacturing speculative markets around viral attention. The roughly year-old Solana platform turns coin creation into a few clicks and invites users to declare, “I’m ready to pump”; promotion then runs through TikTok, X, and Discord until momentum attracts traders trying to exit before the crash. The result is a “crypto carnival” where Butthole Coin reached a $40 million market cap and a Shoggoth coin inspired by Kevin’s reporting reached $31 million.
Chris Hayes argues that attention has become a profoundly valuable, monetizable commodity connecting platforms, politics, crypto, and AI. Meme coins are, in his phrase, “the purest monetization of attention,” while haptic alerts and algorithmic feeds compel users in ways that no longer feel volitional. AI could become a useful filter against spam and information overload, but Casey countered that TikTok is already an AI “meme servant,” and the next generation may create, distribute, and screen increasingly irresistible media.

Deep dive

1. TikTok became “Schrödinger’s app”

Casey’s shorthand captured the operating paradox: TikTok went dark for about 12 hours, returned for existing users, but remained legally banned and absent from US app stores. Its shutdown screen credited Trump in advance with promising “a solution to reinstate TikTok once he takes office.”
The outage swept beyond TikTok. ByteDance-linked Lemon8, CapCut, and Marvel Snap also disappeared; Casey compared the latter’s removal to “a Thanos snap,” particularly striking because few users realized its publisher was owned by ByteDance.
Oracle, TikTok’s US infrastructure provider, and Akamai, its content-delivery network, accepted Trump’s assurance that service providers would face no liability and flipped their systems back on. Apple and Google did not, because a future administration could still enforce the statute.
That split creates slow operational decay. Existing installations work, but ByteDance cannot ship the weekly—or sometimes every-few-days—updates that add features, repair bugs, and patch security vulnerabilities, potentially leaving American users frozen on aging software into April.

2. The app-store holdout prices the rule-of-law risk

PAFACA permits a 90-day presidential extension tied to a divestiture process, but Biden declined to grant one. Trump instead ordered a 75-day enforcement delay after the law had passed Congress, been signed by Biden, and survived Supreme Court review.
Apple and Google face a stated fine of $5,000 per user for facilitating the banned app. Against roughly 170 million US TikTok users, the theoretical exposure reaches hundreds of billions of dollars—“ruinous fines” that a verbal promise from one president cannot fully neutralize.
Casey’s concern extended beyond TikTok: the incoming president effectively said a duly enacted and upheld law “does not apply because I say so.” Kevin’s version of the institutional problem was simpler: if that works, “Schoolhouse Rock would have something to say about that.”
TikTok therefore tests more than content policy. Service providers are assigning different values to presidential assurances, statutory text, future enforcement, and political favor—and producing materially different access decisions from the same legal facts.

3. Washington disowned the ban just as it took effect

The Biden administration called TikTok’s shutdown threat “a stunt” and said it saw no reason for companies to act before Trump entered office. Casey found that “profoundly embarrassing”: Biden’s signature tech law declared the service illegal and threatened enormous fines, then his administration acted surprised when companies complied.
Republican resistance was also sparse. Senator Tom Cotton warned that anyone hosting, distributing, or facilitating “communist-controlled TikTok” could incur “hundreds of billions of dollars of ruinous liability,” but Casey stressed how few lawmakers repeated what had recently been their national-security case.
The hosts preserved the unresolved contradiction: Congress had portrayed TikTok as capable of spying on Americans, misusing data, spreading propaganda, and sowing dissent. Once the ban arrived, most politicians either opposed enforcement or went silent, raising Casey’s question: “Was this app really that big of a problem?”

4. TikTok moved onto the US-China tariff chessboard

Trump proposed that ByteDance sell through some unspecified joint venture and “give half to the United States of America,” claiming TikTok was worthless without a permit but worth “like, a trillion dollars” with one. He mentioned Elon Musk and Oracle founder Larry Ellison as possible buyers.
Casey’s honest non-answer: “What is Donald Trump talking about? I truly do not know. And I will not speculate.” A literal 50% Treasury stake would resemble partial nationalization, while government ownership could also give Trump leverage over moderation and promotion.
The decisive change may be China’s first signal in roughly five years of coverage that it could tolerate ByteDance divesting TikTok. Trump explicitly connected the app to threatened tariffs on Chinese goods, making ownership one component of a negotiation with potentially much larger economic consequences.
The hosts split on the 75-day outcome. Kevin expects no completed deal and predicts another extension; Casey expects a deal, with China potentially trading a TikTok sale for tariff terms that do less damage to its economy, possibly to someone it can influence such as Musk.

5. Stargate illustrates the new corporate courtship

Casey’s emerging playbook for tech under Trump: let him take credit for plans already in motion, praise his role, wait for the spotlight to move, and continue operating. TikTok’s shutdown message was one specimen; Stargate supplied the clearest one.
OpenAI, Oracle, and SoftBank announced a separate AI-infrastructure company committing $100 billion, potentially rising to $500 billion, for data centers serving OpenAI as the expected major customer. The project had already broken ground, and its announced financing was entirely private.
Sam Altman nevertheless told Trump, “We couldn’t have done this without you, Mr. President.” Casey called that false and warned that CEOs were helping construct a fiction in which every success belongs to Trump and every failure belongs to an enemy.
Kevin supplied the transactional logic: Trump repealed Biden’s AI executive order, which had imposed what Kevin described as light safety guardrails. With essentially no federal regulatory infrastructure remaining, unrestricted deployment may be enough to motivate the flattery.

6. Political flattery buys access but not durable alliances

Kevin expects “churning and thrashing” because Trump’s closest first-term relationships typically had a half-life of only a year or two. Being useful today does not guarantee that OpenAI, Oracle, or another favored company remains aligned with him tomorrow.
Casey’s longer-term objection was institutional rather than tactical: “lying creates bad karma.” Even if coordinated praise works while every major company participates, blatantly misstating government involvement may eventually trap executives inside commitments and narratives they helped create.
The TikTok and Stargate stories share the same mechanism: regulatory discretion has become economically valuable, and companies can improve their position by publicly transforming ordinary private decisions into presidential victories.

7. TRUMP and MELANIA monetized office before inauguration

Trump announced TRUMP through gettrumpmemes.com on Friday, several days before taking office; Melania Trump announced MELANIA on Sunday. Neither token conveyed any stated benefit beyond ownership of a purely speculative instrument.
Both surged into the upper ranks of cryptocurrencies, reaching paper valuations above $10 billion at points. Kevin stressed that this did not mean billions had been realized, though some traders posted that they had made millions during the volatility.
Available reporting indicated that Trump and affiliated entities controlled roughly 80% of TRUMP’s total supply, while Melania-related holders controlled 35% of MELANIA. Kevin kept the ownership hedge intact because meme-coin structures lack conventional disclosure requirements.
That concentration creates an anonymous channel for influence. Kevin’s concern was that an overseas actor seeking US business could buy TRUMP without leaving the conventional trace of a payment, allowing people to “cash in on the presidency” while enriching the family’s position.

8. Vesting limits the rug pull, but fees pay immediately

Kevin described the classic rug pull: reserve tokens, generate hype, then sell into demand and abandon the market as the price collapses. TRUMP’s insiders reportedly cannot unload their entire allocation immediately; tokens are distributed over about three years.
Market mechanics provide another restraint: selling the whole position would destroy the price before holders could realize its quoted value. Casey jokingly translated that nuance into “a safe and good investment”; Kevin immediately answered, “No, don’t do that.”
The more immediate revenue stream is transaction fees. Even without selling the family stake, repeated buying and selling can generate income; a Coinbase director estimated that the Trump family had made roughly $58 million from fees by Saturday.
Kevin’s standing 2025 prediction remains unmet but live: a newly released meme coin will briefly reach a $100 billion market capitalization before crashing. TRUMP approached the thesis without yet crossing his threshold.

9. Crypto’s serious-build narrative collided with its incentives

Industry criticism was unusually blunt. Trump-supporting crypto investor Nick Carter called the coins “preposterous” and their creators’ conduct “plumbing new depths of idiocy”; podcaster Scott Melker called them “a gratuitous cash grab.”
Casey’s pushback—worth keeping: if these coins are embarrassing cash grabs, which crypto assets are not? Kevin distinguished tokens with potential payment or smart-contract utility from meme coins whose only function is buying, selling, and hoping someone else arrives later.
During the Biden crackdown, crypto companies had argued, “Let us build. Let us cook,” so the industry could demonstrate serious, transformative financial applications. Instead, after the floodgates opened, the first act was a president and first lady launching speculative coins.
The contradiction became commercial when Coinbase and Kraken—companies associated with the serious-use case—rushed to list TRUMP and profit from trading. The episode’s conclusion was not that utility is impossible, but that platforms currently get paid faster by speculation.

10. Pump.fun industrialized the conversion of virality into markets

Pump.fun, about a year old, lets users launch Solana meme coins with a few clicks, eliminating much of the coding, forking, and smart-contract work previously required. Its entry button makes the intent explicit: “I’m ready to pump.”
Kevin called the platform a “LARP” and “crypto carnival” built around capturing attention, raising a token’s price, and exiting before collapse. Promotion runs through TikTok, X, and Discord until momentum gains enough gravity that traders pile in to time the unwind.
The game-like absurdity carries real valuations. Butthole Coin, marketed as “the foundation of flatulent finance,” reached a $40 million market cap; Shoggoth, inspired by Kevin’s 2023 AI-meme column, reached $31 million, prompting him to disclaim any involvement.
Casey saw “vast swaths of American life” becoming gambling products. Kevin sharpened that into the “financialization of news”: where viral figures once linked a Patreon, inauguration pastor Lorenzo Sewell immediately launched Lorenzo coin and said proceeds would support his church.

11. Attention capitalism works because demand predates the phone

In The Siren’s Call, Chris Hayes begins with compulsion: a phone buzzing on the table triggers the brain as if “there’s a predator rustling in the bushes.” People may enjoy paying attention to strange things, but today’s allocation no longer feels sufficiently volitional.
Responsibility starts with the platforms whose leaders lined up at the inauguration, though Hayes included Amazon’s product-selling attention business and crypto’s speculation in the same system. The meme coin, he said, is “the purest monetization of attention” because it exists almost entirely in purchasers’ minds.
The demand side is boredom: why do humans want their attention taken? Hayes cited Pascal’s conclusion that humanity’s troubles stem from an inability to sit alone, evidence that restless consciousness long predates TikTok, television, and radio.
Yet Hayes resisted calling boredom universal. Among Australia’s Warlpiri people, the concept reportedly lacks a native word and is described with the English import; in his framing, modernity imports both “a lexeme and an experience.”

12. The crisis is alienation, not merely shorter attention spans

Casey challenged the decline narrative with four-hour podcasts and two-and-a-half-hour YouTube videos. If people voluntarily consume long, niche media, claims that young people simply cannot focus fail to describe what audiences demonstrably do.
Hayes separated empirical questions—whether attention spans shortened or anxiety rose—from the philosophical question of a good life. Even if randomized trials found no measurable harm from playing video games 13 hours daily, he would still argue that the person was not living well.
Podcast architecture supports his case: RSS is an open protocol without a gatekeeper deciding what deserves attention, allowing unlikely four-hour formats to “sink or swim.” Kevin countered that listeners may give them only 10% background attention while driving, folding laundry, or washing dishes.
Casey added that niche podcasts succeed because they cover subjects no friend or relative wants to discuss. Hayes embraced that democratization but traced it to the pre-commercial internet, before platforms consolidated the same open-ended discovery into controlled attention environments.

13. The problem reaches elders, creators, and self-government

Kevin raised the uncomfortable critique that established media figures may resent audiences for looking elsewhere. Hayes conceded the point but cited MrBeast—a master of attention engineering who reportedly finds the system oppressive—as evidence that beneficiaries can also feel trapped.
Hayes rejected the reflex to blame children: “I actually think the kids are in better shape than the elders.” They are more native to screening information, while the deeper problem affects him, his parents, and everyone else.
His family example made the hypocrisy concrete. Seeing his child play a video game beside a picture-in-picture video felt wrong—until the child observed that Hayes watches television while using his phone “literally all the time.” His response: “Fair point, but not you.”
Hayes uses book reading as a personal benchmark. Writing The Siren’s Call “binding myself to the mast,” forcing him through demanding philosophy and rebuilding weakened muscles; his categorical claim was that collective self-government “hinges on our ability to do that.”

14. AI could filter the attention market or perfect it

Kevin offered two futures: AI generates hyper-personalized media and emotionally sticky companions, or it reads the flood on users’ behalf and returns only what matters. Hayes wanted “the five dankest memes” from an AI servant but admitted “zero trust” in his own forecast.
His closest analogy was the late-1990s tech boom: genuine innovation mixed with “ludicrous froth” and products attempted before the technology was ready. Hayes said he was not using ChatGPT and felt mixed about engaging with a new thing, while Kevin and Casey said they use it “every day for everything” and “constantly for everything.”
Kevin’s optimistic specimen was spam filtering, which reduced unwanted email without requiring user effort. He also knew someone who configured a ChatGPT task to deliver a morning list of everything Trump had done the previous day, ranked by importance.
Casey’s rebuttal was structural: TikTok already is an AI meme servant, trained on engagement to select what each person will enjoy. If that state of the art is already “spooky good,” AI may create the TikToks, distribute them, and then power a second AI that watches them for the user.

15. Media form changes when every moment has an exit

Casey’s struggle with Roberto Bolaño’s 900-page 2666—about 250 pages in—felt partly like a damaged attention span and partly like cultural evolution. Books occupy less of his friends’ and family’s conversation than when he studied English, reducing the social reinforcement for finishing them.
Hayes argued that the attentional environment changes the work itself. Films from the 1970s and 1980s could move slowly because viewers had paid, entered a theater, and had nowhere else to go; current creators write, edit, and speak faster because audiences always have alternatives.
Casey preserved a useful final objection: contemporary media is not uniformly accelerated. “Watch the fourth episode of any Netflix show. They didn’t need to make it”—a reminder that abundant supply can produce both frantic pacing and conspicuous bloat.

How the New Administration Will Impact Crypto, AI & Tech Globally w/ Ben Horowitz & Salim Ismail Carlos Delatorre, CRO @Harness: Why Every Sales Rep Should Do Pipeline Generation | E1251