The episode opens with back-to-back sponsor reads. Odoo pitches its all-in-one business management platform as an antidote to fragmented software stacks. Northwest Registered Agent offers a full business formation package with built-in privacy protections. BetterHelp closes with a notable data point from its 2026 State of Stigma report: 74% of Americans believe society still discourages asking for help — a figure that underscores the broader themes of societal pressure the show often engages with.

Alice Han opens the show solo — James is on holiday — and welcomes Ed Elson from ProfG Markets as her co-host for the week. The episode preview crystallizes three major themes that will drive the conversation: the financial decoupling between US and Chinese capital markets, whether China can produce a tech billionaire in Musk's league, and how governments on both sides are (or aren't) protecting workers from AI disruption. Ed Elson's pre-roll soundbite — 'China has demonstrated a willingness to forego money in the name of the common good... what I know about America is that the total opposite is true' — serves as a provocative thesis statement before the conversation has even formally begun.

Alice Han opens the main discussion with a quick Monday market check: Chinese indices finished strongly on the back of news that the US and Iran had reached a deal potentially reopening the Strait of Hormuz, with the CSI 300 up 2.4%, Shenzhen Composite up 3.4%, and the Shanghai Composite up 1.6%. She then sets the stage for the episode's two headline stories — SpaceX completing the largest IPO in history and making Elon Musk the world's first trillionaire, and the Pentagon's latest round of Chinese Military Companies designations covering Alibaba, BYD, and Baidu. Alice frames the week as historically significant, even slipping in a note that the Knicks won against the Spurs for the first time since the 1970s.

Ed Elson digs into the SpaceX IPO in detail, noting it originally targeted $75 billion before demand pushed the raise to $86 billion — making it the largest IPO in history by a wide margin. The $135 per share issuance price saw an initial 10–11% pop on the first day, below Ed's 25% prediction, but the stock subsequently climbed over 30% by the end of the first week. He calls this an early test of market appetite but warns that the next six months will be more telling, particularly as lockup periods expire and early investors may sell. More broadly, he flags that SpaceX is just the first in a wave of massive equity offerings: Google's $86 billion, Nvidia, Amazon, Meta, OpenAI, and Anthropic could collectively inject close to half a trillion dollars of new equity supply into markets — raising the question of whether there is enough investor capital to absorb it all without forcing liquidations of existing positions.

The discussion shifts to why Chinese investors are being excluded from the SpaceX IPO. Ed Elson's instinct is that this is a US-imposed restriction, but Alice Han flips the narrative: it's actually a two-sided story. Beijing's CSRC has cracked down on cross-border brokerages like Tiger Brokers and Futu Holdings, signaling that the Chinese government wants mainland capital allocated to domestic tech rather than inflating US valuations. Alice notes there is no definitive US legislative block on Chinese portfolio investment — CFIUS was weaponized for M&A, but not for stock market flows. High-net-worth individuals with offshore entities may still find loopholes. Meanwhile, China is building its own parallel IPO pipeline: Unitree (humanoid robotics), CXMT and YMTC (semiconductors), and an estimated 50 robotics companies preparing to list. The inescapable conclusion is that financial decoupling is accelerating from both directions simultaneously.

Ed Elson examines the Pentagon's Chinese Military Companies list with barely concealed skepticism. The formal consequence of being designated is minimal — primarily a prohibition on US government contracts and some reputational risk. Elson points out that if the logic is 'any company that could benefit China's military,' then virtually every Chinese company should be on the list, given the CCP's pervasive oversight. He questions whether this is substantive policy or performance art designed to signal hostility toward Beijing. Alice Han counters that the chilling effect is the real story: she describes Hollywood executives privately blown away by ByteDance's Seedance video AI platform, but whose legal teams refuse to allow any engagement for fear of political backlash. The shadow of the sanctions list, she argues, is vastly larger than the list itself — keeping corporate America spooked about China regardless of whether formal restrictions apply.

Ed Elson breaks down the Anthropic-White House conflict in detail. Anthropic released Fable-5, a powerful variant of its Mythos model. Amazon flagged alleged security concerns to the White House; Anthropic disputed them; the White House imposed export controls anyway, forcing Anthropic to pull the model entirely. Elson traces the root cause not to a specific rule violation but to Anthropic's prior declaration that it would set its own terms for Pentagon collaboration — a move that infuriated the administration. Pete Hegseth's Twitter response suggested the export control was not a surprise. The dynamic, Elson argues, mirrors the broader China question: the Trump White House divides the world into friends and enemies, and punishes enemies regardless of whether they've technically broken any rule. Alice Han adds that with David Sacks gone, Scott Bessent is now the primary policy voice on AI, and his financial and national security instincts are taking precedence over innovation concerns.

Alice Han takes a skeptical view of whether Washington's China technology restrictions are actually effective. She points to semiconductor export controls as a test case: China has been smuggling Rubin chips through third-party countries in the Middle East and Malaysia, and has rerouted trade through Southeast Asia and Latin America. The restrictions create friction but not a wall. On the talent side, the dilemma is more acute: approximately 38% of top AI talent in America are Chinese nationals. Tech companies privately admit they cannot build frontier models without this workforce. Any serious attempt to exclude Chinese nationals from sensitive AI projects — as Anthropic just announced for its Mythos and Fable-5 programs — would directly degrade US capabilities. Alice frames this as the administration needing to 'think long and hard about balancing those priorities,' since total decoupling on talent is effectively impossible.

Ed Elson leans into the nuclear-AI analogy that Dario Amodei has publicly championed, using the historical precedent of the nuclear arms race to argue that trying to prevent China from developing powerful AI is futile. The US built the bomb; the USSR and China figured it out anyway. Yet mutual assured destruction has kept nuclear weapons from being used for 80 years. Applied to AI: China will inevitably achieve frontier capabilities. The question is whether a similar deterrence dynamic emerges, and what US policy should look like in a world where China has AI as capable as or more capable than anything the US has built. Elson is struck by the absence of this question in Washington's current policy debate, which remains fixated entirely on prevention. Alice Han agrees, adding that this is ultimately a multipolar trap — each player's dominant strategy is to maximize their own AI capabilities, making mutual restraint structurally impossible.

The episode pauses for a sponsor break. Google Chrome promotes its Gemini AI integration for web browsing tasks. SoFi pitches its private student loan product covering up to 100% of school-certified costs with no origination fees. Indeed closes with a headline from its own data: sponsored job posts on Indeed are 95% more likely to result in a hire than non-sponsored listings, alongside a $75 credit offer for new advertisers.

Ed Elson uses his own calculations to frame the scale of Elon Musk's wealth in a way that most commentators miss. Musk's $1 trillion-plus is not just a large number — at 3.2% of US GDP, it exceeds Rockefeller's historically cited peak of 1.5% of GDP by a factor of two, arguably making Musk the wealthiest person in modern history on a relative basis. Ed then turns to China's equivalent: Zhang Yiming of ByteDance at a reported $93 billion is China's richest person — not even remotely in the same league. He poses the hypothetical of whether China could ever produce a trillionaire, noting the CCP has tolerated billionaires but might draw the line well before a trillion. The question is whether ideology or practical economics would win out in such a scenario.

The conversation zooms in on ByteDance — the world's most valuable private company — and the many reasons it's likely to stay that way. Alice Han argues ByteDance simply doesn't need to raise cash: their revenues slightly exceeded Meta's last year, and their CapEx plans of up to $70 billion in 2026 are far more manageable than Meta's spending. An IPO would invite more regulatory scrutiny from Beijing and complicate the already convoluted TikTok US situation, where Chinese data export rules create a legal minefield. Ed Elson introduces the 'China tax' concept — the persistent discount that the investment world applies to Chinese companies out of fear of arbitrary government intervention, the so-called risk of being 'jack-mauled.' At Meta's 10x sales multiple, ByteDance would be a $2 trillion company. Its $600 billion private valuation represents a $1.4 trillion gap explained almost entirely by political risk. Ed argues this discount leaves enormous value uncaptured and, if he were 'dictator of China for a day,' he'd loosen regulations on public markets.

The conversation turns to the CCP's paradoxical relationship with extreme wealth. Ed Elson wonders what Mao would think of a China that harbors billionaires worth tens of billions — Zhang Yiming at $93 billion, Ma Huateng at $63 billion, Colin Huang at $45 billion. Alice Han frames it as 'socialism with Chinese characteristics': the private sector employs 80–90% of the Chinese workforce and drives the innovation and productivity the state needs to achieve its geopolitical ambitions. The government has implicitly co-opted the tech billionaires as instruments of national strategy. But the tightrope is real: Jack Ma's brief disappearance following a regulatory speech shows exactly where the line is. Being useful is tolerated; being a political threat is not. Alice concludes that China's billionaires are 'tolerated with varying degrees of tolerance' — and a trillionaire would likely cross a line that even Xi Jinping's pragmatism can't ignore.

The second sponsor break covers three distinct products. Gusto pitches its all-in-one payroll and HR platform for small businesses, offering 3 months free on first payroll. Aven makes a striking pitch around the $1 trillion in US credit card debt carrying 23%+ interest rates, positioning its home equity Visa card as a dramatically cheaper alternative. Ferragamo closes the break with a Father's Day-themed ad for its fragrance and shoe collections.

Alice Han opens the labor market segment with a tale of two data sets. China's headline GDP growth looks healthy at 5% for Q1 2026. But underneath, the labor market is flashing warning signs. Youth unemployment for 16-to-24-year-olds remains elevated at 16.9%, while the slightly older 25-to-29 cohort has reached a record high of 7.7%. Perhaps most significant for the AI disruption conversation, nearly half of China's urban labor force has shifted into flexible gig employment — delivery drivers, ride-hailing, platform work — precisely the categories most vulnerable to automation through autonomous vehicles, drones, and humanoid robots. The macroeconomic story and the microeconomic reality are pulling in opposite directions, setting up the episode's central labor discussion.

Alice Han frames the central paradox of this section: China — a country known for authoritarian control — is proactively identifying and legislating against AI's harms to workers, while the US — a democracy — has essentially abandoned this role. China's government mouthpiece Workers' Daily recently ran a series identifying digital cloning, AI-cover firings, and algorithmic pay manipulation as specific threats requiring legal responses. Meanwhile, the Trump White House positioned any AI regulation as innovation-stifling — until it decided to punish Anthropic, at which point regulation was suddenly fine. Ed Elson names this the 'libertarian mind virus': an ideological inability to hold two truths simultaneously (free markets AND sensible guardrails). The result, he argues, is that Americans are more scared of AI than any other major population — with every reason to be, given 50,000 AI-attributed job cuts and no regulatory safety net in sight.

Alice Han goes deep on the specifics of China's official AI labor framework. Workers' Daily identified three categories of concern: companies requiring employees to train AI avatars of themselves (raising IP and data privacy issues), employers using 'AI transition' as cover for firings that Chinese courts have already ruled unlawful, and platform algorithms that obscure working hours and cut pay in ways workers can't document to defend their rights. The broader data is alarming: Chinese university researchers project AI could displace 278 million workers by 2049, one in three current employees. Job postings for college graduates already fell 22% in the first half of 2025. And protests are breaking out — notably in Wuhan, where drivers rallied against Baidu's robo-taxi expansion. Alice argues China is structurally more exposed to AI disruption than other economies precisely because of its gig economy dominance, and draws a parallel to the 1990s SOE restructuring when the government deployed iron rice bowl subsidies to cushion mass layoffs.

Ed Elson delivers the episode's sharpest analytical observation: proactive AI regulation isn't just good ethics, it may be good strategy. If workers trust the government to protect them, they'll embrace the technology. A KPMG and University of Melbourne survey quantifies the gap starkly: 90% of Chinese respondents are excited about AI versus less than 40% of Americans. Ed connects this to the data center opposition story — tens of billions blocked in 2025, 19 states considering restrictions — arguing that America's regulatory vacuum is generating a public backlash that is materially slowing infrastructure buildout. China's approach, by contrast, creates public buy-in. He acknowledges this doesn't require glorifying the CCP, but argues that the White House's libertarian reflex is creating an own-goal that could hand China a structural advantage in the AI buildout over the long term.

Alice Han challenges the common narrative that the US is comfortably ahead of China on AI infrastructure investment. US data center CapEx is projected at $750 billion in 2026 alone, while China's announced target is roughly half that by 2030 — numbers that seem to show a clear US lead. But Alice argues this comparison is deeply misleading because infrastructure and energy buildout costs per gigawatt are dramatically lower in China, meaning the actual capacity gap is far smaller than headline spending figures imply. Ed Elson extends the point to the model layer, noting that Deepseek's AI models are reportedly 96% cheaper than OpenAI equivalents — and that as US enterprises face spending caps on AI, cheaper Chinese alternatives become increasingly attractive. Alice delivers a striking anecdote: major US tech companies she encountered at a Hollywood conference are actively using the Chinese open-source model Qwen because it's cheaper, faster, and easier to fine-tune than OpenAI or Claude. The great AI decoupling, it turns out, is not happening at the enterprise level.

Alice Han closes out the substantive discussion, thanks Ed Elson for standing in for James, and reminds listeners that China Decode is a production of Prof G Media. She directs listeners to follow the show on their preferred podcast platform. The episode then runs out with two short outro sponsor reads: Stitch Fix pitches its personal styling service, and Mint Mobile — voiced by Ryan Reynolds — promotes its $15-a-month unlimited plan, before noting that making $15 bills to advertise the offer would, regrettably, be illegal.