Alibaba's Joe Tsai on US-China Rivalry, AI Future, Owning the Nets/Liberty, Caitlin Clark's Impact
Summary
- Joe Tsai rejects the idea that AI will have a permanent winner. Model leadership changes “every week,” model development may lack winner-take-all network effects, and the meaningful contest is adoption: a survey he cited showed Chinese-firm AI usage rising from roughly 8% last year to nearly 50%, alongside open-source models and smaller 1.7B-, 4B-, and 8B-parameter options.
- Alibaba is already translating AI into lower labor intensity and higher revenue. Tsai said the company can hire fewer people, has announced no AI-driven layoffs, and may have AI writing about 30% of its code; embedding it in e-commerce, maps, and food delivery is also improving consumer experience and expanding the user base.
- China’s government is going “all in” on AI despite acute employment and wealth pressures. Its AI Plus policy targets 90% penetration of AI agents, devices, and similar technologies by 2030, while roughly 10 million annual college graduates face 18% unemployment among 16-to-24-year-olds and average home prices have fallen about 30% during the property slump.
- Tsai rejects treating China as an existential enemy. He described China as focused on economic development and citizen well-being and rhetorically asked how many years—30 or 40—it had gone without starting a war, while acknowledging friction over perceived U.S. containment. A host argued that power is zero-sum and America must win AI and chips; Tsai countered that AI is a long marathon, not winner-take-all, and “almost like air.” Another host said medicine and biology leave room for cooperation.
- Caitlin Clark had an extraordinary economic impact on the WNBA. Tsai said the metrics he cited—including viewership, ticket sales, and sponsorship—rose “almost 4x,” while resisting attempts to frame her rivalry with Angel Reese primarily around race and emphasizing the broader influx of college talent.
- Tsai’s operating answer to complexity is radical focus. After returning as Alibaba chairman, he reframed six businesses as two—e-commerce and cloud computing, with an AI element—while describing the Brooklyn Nets candidly as being in “a rebuilding year” after using five draft picks and holding one 2026 pick they hope will be good.
Deep dive
1. Caitlin Clark changed the WNBA’s economic baseline
Tsai’s number was unequivocal: after Clark entered the league, the metrics he cited—including viewership, ticket sales, sponsorship, and other measures—rose “almost 4x.” He called her economic impact “extraordinary” and “undeniable.”
On whether the league had done enough to protect Clark, Tsai focused instead on her ability to withstand its physicality. He noted that she is a smaller point guard whose durability had been questioned, but said she had proven able to handle the league. Rivalries are useful, he argued, but attempts to cast Clark versus Angel Reese primarily as racial conflict obscure the skill entering the league.
His supporting example was Sonia Citron, a rookie-of-the-year potential candidate whom the host had not heard of before she entered the league. Tsai thought she had just broken the league’s record for highest three-point shooting percentage.
2. Basketball’s rulebook is product management
When the host contrasted a tougher WNBA with NBA flopping, Tsai pushed back on comparing the leagues: both are highly physical and athletic when viewed from the floor.
His sharper formulation was institutional: the NBA competition committee “should be called the product committee,” because rules, officiating, transition take-fouls, and even the three-point line directly determine what fans see.
A host challenged whether Boston taking 43 threes constitutes a good product. Tsai’s answer: “There are 18 different ways you can throw up 43 threes in a game.”
3. Alibaba moved from free markets through regulation to focus
Tsai joined Alibaba because Jack Ma could “instill faith in people.” In Ma’s apartment, 12 to 15 very young people—roughly recent college graduates—listened to a teacher about a decade older paint a compelling vision. Tsai linked Ma’s leadership to the teacher’s ability to communicate, identify talent, and accept that students may become more successful than he is.
Alibaba’s first 15 of 26 years were largely free-market, organic growth, followed by extreme e-commerce competition as internet platforms sought to monetize traffic. ByteDance, though not conventionally viewed as an e-commerce company, became one of Alibaba’s fiercest competitors.
Regulation followed excessive competition and platforms’ monopolistic behavior. Tsai said privacy and antimonopoly rules were beneficial and that Alibaba now knows “what the red lines are,” making the operating environment more predictable.
Returning as chairman roughly 18 months earlier, he rejected describing Alibaba as six businesses: “We’re in two businesses, e-commerce and cloud computing,” with an AI element. He said that focus was crucial to getting teams to execute.
4. AI leadership will rotate; diffusion is the contest
A host’s realist case was explicit: economics need not be zero-sum, but power is, and America should remain strongest by winning in AI and chips. Tsai acknowledged that a patriotic American wants America to win, but rejected the finish-line metaphor: “There’s no such thing as winning the race.”
More broadly, Tsai said he did not agree that China is an existential threat. He characterized China as focused on its own development and citizens’ well-being, rhetorically asked how many years—30 or 40—it had gone without starting a war, and acknowledged Chinese concern that the U.S. is trying to contain its rise.
Model rankings change weekly, and Tsai is unconvinced that models possess marketplace-style network effects. His definition of winning is “who can adopt it faster,” not who produces the strongest model. He argued that resources should go toward adoption and diffusion, not only model development.
While individual hyperscalers invest about $80 billion annually, China is emphasizing open source and smaller models—1.7B, 4B, and 8B parameters—that can run on phones and laptops. A survey Tsai cited showed Chinese-firm AI use rising from 8% to nearly 50%.
5. AI is lifting Alibaba while China absorbs job anxiety
Alibaba can avoid some hiring, though Tsai stressed it has announced no AI-driven layoffs. His rough estimate was that AI writes “maybe 30%” of company code; consumer-facing deployment in commerce, maps, and food delivery is producing revenue uplift through better experiences and user growth.
The host raised driver displacement and cited protests in Wuhan, but Tsai said China’s government remains “all in.” He said its recently launched AI Plus policy calls for 90% penetration of AI agents, devices, and similar technologies in society by 2030.
Tsai said there is little general discussion of AI replacing jobs, but acknowledged significant job anxiety: China graduates about 10 million college students annually, and unemployment among 16-to-24-year-olds, including graduates, is 18%. He also cited a property slump lasting four or five years, with average home prices down about 30% and the resulting negative wealth effect still lingering.
On AGI, Tsai said government officials do not discuss it much because they believe they could control it better than U.S. society, though fears linger. Dario Amodei’s 5-to-10-year forecast led Tsai to guess “probably 20 years”; he based that on “general” intelligence needing to apply principles to scenarios it has never seen before.