China Biz Talk - Baiguan Radio

www.baiguan.news
  • 5.0 (1)
  • Investing

Actionable insights for successful China-related business and investment, with a human touch www.baiguan.news

  1. 3d ago

    Field Notes: AI, Robotics, Commercial Space and the Case for Niches - Baiguan Radio #45

    We are already gathering feedback for the 2027 Spring Baiguan China Tour. Your preferences will help shape our program. So make sure to fill in our survey here. It won’t take more than 2 minutes of your time. Episode in brief * Recapping Baiguan Autumn Tour: Robert recaps the recent Baiguan Tours to Beijing and Shenzhen to visit AI and robotics companies. * Beijing is a deeper tech base than its reputation. Robert says talent (Tsinghua, Peking, the AI research institutes), proximity to policymakers and regulators, and the legacy of Baidu, Meituan and ByteDance explain why Moonshot AI and Z.ai were founded there. * Robotics looks like a bubble, but a useful one. General-purpose humanoids disappoint in open settings. Narrow, commercial use cases already work, from a robot-run pharmacy to cleaning and dish-carrying robots. * Durable value sits in unglamorous components. Dexterous hands and gears are concentrated, hard to copy and priced on cost leadership, not hype. * Commercial space is gated by approvals, not technology. Chinese rocket firms say slow government approval limits launch cadence, and officials tolerate failure less readily than a Musk-led company like SpaceX can. * China is not uniformly “full speed ahead.” Robotaxis are held to pilot zones for employment reasons even though the technology is mature. * The strongest takeaway is niche strategy. Robert argues that as AI models produce generalized, mid-distribution answers, value shifts to peripheral niches, which Smallrig illustrates in hardware and a founder’s lunch talk illustrated for AI. Why the tours went to Beijing Robert has just finished the first two tours of the autumn series: a week in Beijing with the independent media company Caixin, covering AI, robotics and commercial space, followed by a shorter Shenzhen trip on robotics and advanced manufacturing. Participants were foreign investors and corporate executives from the US, UK, France and Southeast Asia, with some overlap between the two tours. Olivia notes that most people would expect an AI and robotics itinerary to center on Hangzhou or Shenzhen. Wu says participants raised the same question. His answer is the talent pool: Beijing has Tsinghua and Peking, where Hangzhou has Zhejiang University, and he says the founding teams of Moonshot (Kimi) and Z.ai trained in Beijing’s AI research institutes. He adds that demand for robotics is heavily policy-driven, so proximity to government helps, and that commercial space is highly regulated, so companies stay near regulators. The internet era points the same way, since Baidu, Meituan and ByteDance were all headquartered there. AI: three companies, three cultures Moonshot AI (Kimi) Wu’s account is that Moonshot differs from DeepSeek, whose focus he describes as “high intelligence with lowest cost.” Moonshot, in his reading, is aiming at the top of the intelligence range, building models with trillions of weights and emphasizing science and medical research. It is also an open-weight company that has worked closely with international cloud service providers. Wu describes a licensing arrangement in which heavy commercial users above a threshold share revenue with Moonshot. Because the models are hosted on the provider’s own infrastructure, he says, there is no concern about a Chinese company holding the data, and customers keep more control than they would with a closed-source vendor such as Anthropic. He judges Moonshot the best of the Chinese AI firms at working with international stakeholders. The culture was young: staff mostly under 30, meeting rooms named after rock bands, and the group met in the Radiohead room because the founder loves rock music. The schedule shifted at short notice because the model companies were busy, which put Moonshot on day one. Z.ai and a pivoted model company Z.ai, maker of the GLM models, felt different: a full showroom and a practiced reception style for delegations, including local governments, which Wu notes matches the state-sector clients in Z.ai’s own base. A third model company, which Wu chose not to name, stopped training its own models about two years ago, saying it lacked the money, and became an AI application and forward-deployed engineering company. Participants were skeptical of its product that gives CEOs visibility into everything in the company, because, as Wu puts it, some executives do not want to see everything and then carry the liability. Wu still counts the visit as useful for showing where different firms sit on the value chain and why they make different choices under different constraints. “Different companies make different decisions given the different constraints.” Robert Wu Robotics: a bubble with real commercial ground What worked in the real world Wu says the highlights were not the dancing robots. The first was an underground pharmacy run by robots, serving a steady stream of Meituan and Alibaba delivery riders. He stresses it was a live operation, not a showroom. The case for automation is a pharmacy that should ideally run around the clock, where a human night shift is costly and arguably unnecessary. Where it fell short At a robot-staffed convenience store, a company employee asked the robot for a yellow juice, and the robot replied that it had nothing like that despite orange juice and other yellow bottles on the shelf behind him. Wu says that was a letdown for participants. His reading is that robots cannot yet operate freely across scenarios, but within a controlled setting, such as picking medicines or moving boxes out of a disorganized pile, some general capability is clearly developed. Revenue without subsidy Wu cites a large service-robot maker (Pudu Robotics) with revenue he puts above RMB 1 billion. Its core products are cleaning robots and cat-shaped dish-carrying robots popular in Japan, with humanoids a secondary line. Wu says its revenue is commercial, not driven by subsidies. The company survived an early existential crisis during Covid supply chain disruption by finding a use case in cleaning robots and by focusing on higher-margin international markets from the start. Wu believes it would survive even if the robotics bubble bursts. “We shouldn’t overestimate the intelligence of robots at this stage, but also we shouldn’t underestimate what the robots are already able to achieve from a commercial perspective.” Robert Wu Pockets of durable value Humanoid making is fragmented in Wu’s account, because components are available and almost anyone can assemble a robot. Two parts of the chain are different: * Dexterous hands. The top maker claims 70 to 80 percent of the market. Wu says the hand is the most complex and highest-value part of the body, and that no supply chain for its joints and gears existed, so the maker built from scratch in Beijing and location did not matter. * Gears. Wu visited a listed company and described gears as “the chips for the industrial world.” In his telling, it has become the least-cost producer in a product boring enough that newcomers are not drawn in, which creates a durable moat. * Data collection. Origin Flow, founded about a year ago and already through four funding rounds, makes a wristband that reads hand signals. It can serve as a computer input device for disabled users or as a data collection tool for training robot hands. It has launched a first kit and plans to extend to full-body data collection. “All throughout the robotics value chain, you find these pockets of value that are not easily involuted away.” Robert Wu Implication for investors. Wu’s consensus view is that Chinese robotics is entering, or is already in, bubble territory, which he previously called “necessary” and says this trip reinforced. His practical filter favors firms with clear unit economics, concentrated market position or cost leadership, over humanoid demonstrations. Commercial space: ambition capped by approvals The planned company visits did not receive clearance, given the sensitivity of the sector. Instead, local government hosted a sem… This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.baiguan.news/subscribe

    Field Notes: AI, Robotics, Commercial Space and the Case for Niches - Baiguan Radio #45

  2. Oct 20

    Golden Week Spending, China’s Housing Affordability Push, EU Trade, Manus After Meta, and the AI Access Fight - Baiguan Radio #44

    Episode in brief * Golden Week was healthier than the headlines. Total trips and spending both rose. Only spending per trip fell, by about 2%. Robert reads that as money spreading to smaller towns and to services, not as weak demand. * Beijing is working on the monthly payment, not the price. A central-government mortgage interest subsidy for some first-time buyers is, in Robert’s view, a more important housing measure than the market has noticed. He cites Baiguan data showing a pickup in secondary-market transactions, but new-home sales are not benefiting. * The EU-China car arrangement is, in Robert’s reading, more theater than substance. The “cut by half” is measured against an EU projection, so actual exports can still grow, and both sides can claim a win. * The Manus outcome supports a less alarmist reading of Beijing. Robert argues the unwinding of the Meta deal was a reasonable use of existing review powers, and the company’s first independent funding round shows it is still allowed to grow. * Consumer AI in China is held back by platform access, not model quality. More than 150 apps have reportedly blocked a ByteDance AI assistant from operating them, which fits Robert’s view that the large platforms are fighting over who controls the interface. Golden Week: Consumers Pay for Experience, and Spending Spreads Out Outbound travel is back, and shopping is no longer the reason Olivia opened with an observation from Kenya. At a national park that takes roughly a five-hour drive from Nairobi on rough roads, she found most safari vehicles full of young Chinese tourists, apparently travelling independently rather than in tour groups. Robert said his parents saw the same in Norway and Iceland and described a return to the post-Covid pattern of Chinese tourists “taking over the world.” He added a base-effect point: with China’s population so large, even 10% to 20% growth in outbound travel means millions more people. Both hosts tied this to a shift from shopping trips to experience trips. Robert pointed to a big Louis Vuitton venue in Shanghai that he doubts has an equivalent in Paris, and argued that China already offers plenty of shopping, so it is no longer a reason to fly. The same logic applies at home. He says domestic travellers increasingly choose smaller, cheaper, less crowded and less famous places over landmarks such as the Oriental Pearl Tower or the Forbidden City, which he describes as a past era of travelling mainly to bring home a photo. How negative are the spending numbers? Olivia cited figures from Baiguan’s Golden Week recap and noted that many headlines she saw about the holiday were negative because spending per trip fell. She added three offsetting factors: a longer holiday, more outbound trips, and more travel to lower-tier cities where money goes further. Robert said he was surprised anyone read the data as negative. Total trips and total spending rose, and the one weak reading, spending per trip, is consistent with more people travelling to cheaper places. He sees the dispersal of tourist money across the country as a positive that lays foundations for future growth there. Services growth, and why film is the wrong proxy Olivia pointed to the services categories as the bright spot and linked them to the government’s stated aim of expanding services consumption. Robert agreed and addressed the usual counterexample, cinema. He argues box-office weakness reflects a global decline in cinema-going that China shares, so it is a poor proxy for Chinese service demand. He also makes a distributional argument: services are delivered by people, and per-capita income in services tends to be higher than in goods industries, so service growth helps spread wealth and supports further consumption. Participatory film sets as an early signal The hosts discussed a format highlighted in Baiguan’s recap: at China’s largest film and television shooting base, in Zhejiang, tourists can pay several hundred RMB to play the lead in a short drama. Robert described a friend of a friend, previously in real estate, who is trying to convert a defunct commercial property into live movie sets with supporting actors for around 100 RMB per experience. Implication. Spending per trip is a weak standalone indicator. The episode points to trip volume, geographic spread and services categories such as sports and culture as the better gauges of where consumption is heading. Robert’s caution about the film-set business model is a reminder that the format is unproven commercially. Read more on Golden Week consumption here: Housing: Beijing Targets the Monthly Payment The measures Olivia framed housing as the main drag on consumer confidence and cited S&P Global Ratings, reported by CNBC, saying China’s property market is nearing a bottom based on comparisons with other countries. She described two recent measures. In August, Beijing tightened rules on selling unfinished homes. More recently, it announced that it will pay part of the mortgage interest for some first-time buyers of homes under 1.5 million RMB (about 220,000 USD) and under 120 square meters (about 1,300 square feet). Robert’s argument: affordability, not price Robert says this is one of the more impactful of a long string of incremental housing policies and that the market is not paying enough attention to it. The subsidy is only around 1% of a property’s value, which looks too small to matter. But he argues buyers think in monthly flows, comparing the mortgage payment with income, not in total price. It is also unusual for the central government to commit its own finances. His case rests on affordability measured in years of income. At the peak, he says, an average Shanghai apartment took almost 30 to 40 years of income, and 10 to 15 years in some smaller cities. Since then prices have fallen 30% to 50% or more in some places, interest rates have dropped, and the subsidy lowers the effective rate further, to around 2% for some buyers. Total personal income has risen over the same period, although people are saving more than they consume. He says some estimates put the improvement in affordability at as much as three times, for example in Shanghai. “You are paying less on a monthly basis for something that has less value with a higher income, which means the affordability of housing has gone up.” Robert Wu Robert reads current policy as a deliberate effort to keep pushing that affordability measure lower. His forecast is that the cycle ends when buying becomes obvious to the average buyer, for example when a mortgage takes about a third of income. What the early data show, and where the policy does not reach Citing Baiguan’s own data, Robert says secondary-market transactions in major cities such as Shanghai have grown about 20 times year on year since the subsidy was announced. Prices have not risen, and buyers are taking longer to negotiate and view more properties, but volumes are up. New-home sales are not benefiting because the policy targets smaller, cheaper homes, which are mostly in the secondary market, while new-build supply skews upscale. Why Beijing, and why now Robert argues that housing was historically a local-government responsibility, including construction, permits and market health, and that people have long urged Beijing to take more of it on. He sees early signs that it now is, which in his view means it can do more. On timing, he says conditions keep worsening and that people compare the present with 2024, before the September policy pivot. He thinks the country is not back there yet, but that policy is being applied pre-emptively. If the measure does not work, he expects the central government still has tools to push affordability further. This is a forecast, not an announced plan. He also tied the policy to broader consumption. Home purchases generate demand for renovation, furniture and appliances, and sellers receive cash to spend or reinvest. First-time buyers typically buy to start families. Olivia noted Shanghai’… This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.baiguan.news/subscribe

    Golden Week Spending, China’s Housing Affordability Push, EU Trade, Manus After Meta, and the AI Access Fight - Baiguan Radio #44

  3. Sep 28

    David Fishman on Power, Politics, and a Mega-Dam: Inside China's Energy Sector - Baiguan Radio #43

    About the Guest David Fishman is a principal consultant in the China practice of The Lantau Group, an energy economics and policy advisory that works with industrial power buyers and investors across solar, wind, nuclear and coal assets. He writes Crossing the River by Feeling the Stones, a Substack on China’s emerging and lower-tier cities. Episode in Brief * China’s “clean energy leadership” claim rests on a different definition than the West’s. Beijing frames leadership as enabling other countries’ decarbonization through mass technology exports and production volume, not necessarily as having phased out coal at home — coal alone still supplies roughly half of China’s 2025 generation mix. * China’s renewable-build targets are negotiated between industry and policymakers rather than imposed top-down, which Fishman says is why the country has repeatedly “underpromised and overdelivered” on clean-energy goals — and why the same pattern could hold on the 2030 peak-carbon target. * Energy is one of the sectors most directly run by the state, unlike China’s more bottom-up AI buildout. Local officials originate and pitch most projects; only genuinely national moves — cross-provincial power trading, a national data-compute strategy — originate in Beijing itself. * The Medog Hydropower Station, a $160–170 billion, 60-gigawatt project targeting 2033 completion, will generate roughly as much electricity annually as a mid-sized country consumes. * Western restrictions on Chinese clean-energy equipment (Chinese suppliers reportedly account for roughly 80% of EU inverter shipments) raise a tradeoff Fishman says hasn’t yet been convincingly argued: national security versus decarbonization speed. What “Clean Energy Leader” Actually Means Fishman opened by complicating the premise of the question itself: “leadership” needs to be defined, and China’s own version of it is not the one usually assumed in Western coverage. Beijing’s implicit standard is enabling other countries to decarbonize by supplying clean-energy technology at scale and low cost — solar panels, batteries, EVs — rather than a race to eliminate domestic emissions fastest. Both things can be true at once: heavy technology exports and a domestic grid still slightly over half thermal. China’s own 2025 generation mix backs up that tension. Fishman’s estimate: about 55% thermal (roughly 51% coal, ~3% gas), just shy of 36% combined wind, hydro and solar, another 2-3% nuclear, and the remaining ~5% split between biomass, wave and other small sources. On the pace of change, Fishman argued China’s official targets are structurally conservative because of how they are set: less a hard number imposed from above than a negotiation, with policymakers asking industry what it believes it can do, then testing whether it can do more. That process tends to produce targets industry is confident of hitting — and sometimes beating by a wide margin, since recent five-year build targets have in some cases undershot the country’s own trailing three-year average. Fishman’s read is that the country’s targets are calibrated to be beatable, which has been the recent pattern. “There’s been a lot of underpromising and overdelivering in the near recent past in the Chinese energy sector.” — David Fishman From Process Innovation to “Zero to One” Fishman also flagged a shift in the character of China’s clean-tech edge. For years the country’s advantage was “process innovation” — making an already-proven technology like solar cells incrementally cheaper, faster or more efficient, driven by intense domestic competition. He now sees genuine “zero to one” innovation emerging in areas that are not yet technologically proven anywhere, citing next-generation batteries and downstream hydrogen derivatives as examples. “Some zero to one innovation is happening in clean tech now in China, whereas China was previously known for one to one hundred, iterating an already existing concept.” — David Fishman Beijing Sets Direction, Local Governments Drive Execution Asked how national priorities actually turn into projects on the ground, Fishman drew a sharp line between “Beijing support” and provincial or municipal support. Most energy projects — a data center, a solar farm — originate locally: officials pitch the idea and seek approval upward. Only genuinely national-scale moves, like cross-provincial power trading (sending hydropower from Sichuan to eastern China) or a national data-compute strategy, originate in Beijing itself. “Local government officials in China are not principals. They are agents themselves. They are pursuing certain activities in response to their own incentives — they want career advancement, or they’re just trying to interpret national-level policy.” — David Fishman Fishman contrasted this with AI, where the buildout looks far less state-directed and more the product of disaggregated private activity responding to a policy-created ecosystem. Energy, the nuclear sector and grid development, by comparison, sit squarely inside the state’s direct sphere of influence. On policy risk specifically, Fishman was direct: change happens, but it is never a reversal — only a “course adjustment.” His example was the shift away from guaranteed feed-in tariffs toward market-priced electricity sales, a change individual power-plant operators dislike but that he says most economists view as the correct design move. Unlike the US, where recipients of a policy change can lobby directly and sometimes reroute it, in China that feedback runs through state research institutes and universities: affected companies can get details softened, but they cannot block a change they oppose. “It’s not a new direction. It’s never a reversal of previous policies. It’s a course adjustment.” — David Fishman Fishman’s framing: policy risk in China’s energy sector is directional and negotiable at the margins rather than existential, though a feed-in-tariff-style repricing can still hit the economics of an existing asset hard. Inside the Medog Hydropower Station The episode’s centerpiece was the Medog Hydropower Station, the subject of a recent Fishman article. The project features an unusually steep “hydraulic head”: the Yarlung River drops roughly two kilometers in elevation over a 200-kilometer horseshoe bend. Rather than following that bend, the plan is to drill a 40-kilometer tunnel straight through the mountain and route the river through it, capturing the elevation drop as power along the way. “This is not a low-hanging fruit project. This is the highest fruit on the tree type hydropower project.” — David Fishman Transmission is the open question. Fishman said only one route is publicly confirmed — a high-voltage direct-current (HVDC) line toward the Greater Bay Area in Guangdong, likely terminating in Huizhou, though he flagged that specific city as not fully certain. Sixty gigawatts will need several more lines; he estimated three to eight HVDC routes in total. On how that power gets absorbed once it lands, Fishman made a fungibility point worth remembering: grid operators elsewhere in the country cannot tell where incoming electricity originated, so new low-cost supply simply displaces whatever is the highest-cost or highest-emission generator in the local mix. Data Center Attitudes On data centers, Fishman said the topic barely registers with ordinary Chinese people, for a structural reason: residential electricity pricing in China is shielded from market movements, so there is no equivalent of the US fear that a local data center will push up household power bills. Chinese media coverage reinforces this by framing data centers as high-tech development news rather than a public-interest or community-impact story. Security Pushback Olivia raised US and EU restrictions on Chinese clean-energy equipment on cybersecurity and supply-chain grounds, citing a Wire China report that Chinese suppliers account for roughly 80% of inverter shipments into Europe. Fishman’s response set a specific test for whether that poli… This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.baiguan.news/subscribe

    David Fishman on Power, Politics, and a Mega-Dam: Inside China's Energy Sector - Baiguan Radio #43

  4. Sep 14

    Xi's Busy Travel Schedule, Long-Form AI Video, Dating Scam Bust by Anthropic, Moonshot IPO - Baiguan Radio #42

    Episode in Brief * Xi heads to Washington. Xi Jinping is set for his first US state visit since 2015, capping his busiest travel stretch since 2019. Expectations for the Trump-Xi meeting itself are low — stabilization, not a reset — though a Chinese business delegation is plausible and worth watching for who is (and isn’t) on it. * AI drama goes long-form, regulation follows. Mango TV has released what it calls China’s first full-length AI-generated drama, a 30-episode sequel made in five months instead of a year-plus. * A labor dispute goes global. Auto-lighting supplier Xingyu rescinded roles for about 100 new graduates; when a domestic complaint stalled, the graduates took their case straight to Xingyu’s European automaker clients, triggering investigations at Volkswagen and Mercedes-Benz. * Moonshot moves toward an IPO. Kimi-maker Moonshot is reportedly preparing a dual Shanghai/Hong Kong listing. Robert discusses how China’s frontier AI labs are unusually collaborative with each other because they see the competition as global, not domestic. * Anthropic disclosure meets Moonshot dispute. A new Anthropic threat report details a China-based dating-app scam network built on Claude, landing alongside an unresolved dispute over Moonshot allegedly rerouting Kimi queries to Claude. Xi’s State Visit to Washington Low Stakes, High Symbolism This has been Xi Jinping’s busiest travel stretch since 2019: stops in Kazakhstan and Egypt and at the BRICS summit in New Delhi, with a first state visit to Washington since 2015 still to come later this month. Robert Wu is skeptical that the Trump-Xi meeting itself will produce much: “the expectation is not very high.” He argues its main function is simply to keep principal-to-principal contact happening, not to reset the relationship. Unlike last year’s meeting in South Korea, which came amid an active trade dispute and drew intense speculation, this year’s meeting draws relatively little domestic Chinese attention — Wu says even policy circles are only lightly engaged, because China feels it has “found its own pace” in dealing with Washington. He does not rule out a narrow trade agreement or selective easing of controls, but cautions against reading Xi’s other stops as building toward the US visit; Chinese leaders typically bundle overseas trips together for logistical reasons rather than strategic sequencing, citing Xi’s 2013 stops in Mexico and Costa Rica ahead of a US visit as precedent. Will Xi bring business leaders? One open question is whether Xi will bring a delegation of Chinese executives, something he has mostly avoided since around 2020. Reuters and SCMP reporting suggests he might this time, and Wu thinks it plausible given that Trump’s own May visit to Beijing included Elon Musk, Tim Cook, and Jensen Huang. Nothing is confirmed. Wu specifically hopes to see China’s frontier AI founders included, though he calls this “very unlikely,” and flags that ByteDance founder Zhang Yiming was a conspicuous absence from the equivalent delegation during Trump’s Beijing visit despite ByteDance’s high stake in the relationship. Looming over the visit: Huawei’s criminal trial has just opened in New York, and Shenzhen hosts APEC in November — China’s largest hosting moment in over a decade. “I think right now it’s just good that the leaders of two countries get to meet every now and then in person and have this principal-to-principal communication. This really can help stabilize the relationship.” — Robert Wu China’s First Full-Length AI Drama Lands Appetite for more than micro Mango TV has released what it calls China’s first full-length AI-generated drama: a 30-episode sequel to Journey to the West, produced in five months versus the year-plus a conventional production would take. Mango’s stock rose in the days after release. Olivia frames this against the prior episode’s discussion of Niu Lai, an unpolished, human-made viral film that succeeded partly because audiences craved something authentic — raising the question of whether audiences now want the opposite. Robert doesn’t see AI and human content as being in conflict. He compares the current spread of AI-generated micro-dramas — now visible everywhere on China’s subways and in elevators — to the early mass-adoption phase of short video in China. He doesn’t think the government opposes the trend, reading it instead as an emerging industry Beijing is willing to let develop even as it disrupts incumbents: “many small-time actors and actresses lose a lot of their jobs” as some production sites lose work. His forecast is a split market: a large, cheap, always-available content tier that AI fully occupies, alongside formats built on the audience’s ability to physically meet or interact with real performers — concerts, tours, fan events — that AI cannot replace. He expects a middle category of AI-enhanced human performance to grow too, describing a friend’s account of a human-acted, choice-driven interactive drama filmed like a branching game, where the actors resisted being replaced by AI specifically because they wanted in-person contact with fans; he speculates AI could eventually make branching-script productions like this commercially viable by handling the otherwise unmanageable number of narrative paths. Olivia connects this to a physical-world parallel: a reconstructed palace complex she visited in Ordos, Inner Mongolia (formerly nicknamed a “ghost city”), built at large expense purely as a filming location — the kind of capital investment that could increasingly be replaced by virtual, AI-generated sets, with data centers taking the place of studio backlots. “There will be a mass market, really cheap, you-can-access-anytime type of content, and that will be completely occupied by AI. But then there’s also some work, some projects where humans are actively involved.” — Robert Wu A Labor Dispute at Xingyu Goes International Don’t test these recent grads Xingyu, one of China’s largest automotive lighting suppliers and a vendor to global automakers, hired roughly 400 new university graduates and then reportedly moved to terminate about 100 of them — offering the option to resign voluntarily or accept lower-paid assembly-line roles. A domestic complaint to China’s labor bureau reportedly went nowhere, so the affected graduates took their case directly to Xingyu’s European clients. Volkswagen Group China has opened a special investigation; Mercedes-Benz has confirmed it referred the matter to a specialist team. Xingyu has apologized, suspended its HR director, and offered a compensation package including a job-search subsidy and accommodation support. The company is already listed on the Shanghai Stock Exchange and is reportedly planning to add a Hong Kong listing. Robert calls the underlying dynamic simply sad: trust between Chinese employers and employees is low right now, and he says it runs in both directions — he separately describes a friend, also a founder, whose signed offers to two strong candidates were declined at the last minute, partially because the candidates didn’t trust the company to honor them. He’s unambiguous that Xingyu’s conduct was wrong (“clearly not ethical, not a professional thing to do”), but he’s struck by how effectively the affected graduates organized: escalating to the company’s biggest overseas clients was, in his words, “really smart,” and the incident became national news specifically because of how well-organized the response was. The bigger pattern, per Robert, is that Chinese companies now selling into global supply chains carry global reputational exposure that most of them have not internalized. “You can see that young people today are very innovative and creative in how they seek their own rights. Like this going-to-the-EU thing — it’s really smart, I have to admit.” — Robert Wu Moonshot’s IPO and a Rare Truce Among China’s AI Labs Competition looks different these days Moonshot AI, maker of the Kimi model family, is reportedly preparing a dual Shanghai an… This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.baiguan.news/subscribe

    Xi's Busy Travel Schedule, Long-Form AI Video, Dating Scam Bust by Anthropic, Moonshot IPO - Baiguan Radio #42

  5. Sep 4

    Alibaba's $10B AI Equity Raise, Niu Lai, and Beijing's ‘Going Global’ Support Plan - Baiguan Radio #41

    Notice: Early bird pricing for Baiguan China Tour will end on September 13. Check the full program, and the Q&A on top questions. Also read this interview transcript with Mr. Gao Erji, executive president of Caixin Media, on the AI & Robotics leg of the tour. Episode in Brief * Alibaba’s $10B Hong Kong share placement (closed Aug 26) buys AI runway but reads as a strain signal, not a strength signal: the company is raising equity for AI while still funding a costly local-delivery fight, and founders’ post-news insider buying landed as weak reassurance because it was priced below what public investors had just paid. * Robert sees ByteDance, not Tencent, as Alibaba’s most underrated AI rival: it is privately held, so it can absorb losses without market pressure; it has a profitable core business funding its AI bets; and its short-video AI products are already generating meaningful revenue. * A crudely animated, five-year, zero-marketing-budget independent film (‘Niu Lai’) became the breakout hit of the summer precisely because it is not AI-polished — this is seen as a genuine hunger for authenticity that a pre-AI production timeline made possible, and that they expect will be very hard to repeat. * Beijing’s new 2026–2030 “going global” action plan for cyberspace enterprises (issued Aug 21) is read as the government formalizing and promoting a trend companies already started on their own — partly in response to unregulated conduct abroad, such as the Didi/Meituan food-delivery fight with iFood in Brazil. * China’s new rules restricting emotionally manipulative AI “companion” features (effective July) pushed ByteDance, Alibaba, and Tencent to pull related features even though nothing was formally banned. * Justin Sun’s public dispute with actress Jing Tian — including his own admission that he uses Claude for major daily business decisions — has turned into a genuine reputational liability for him. Alibaba’s AI Investment and the Competitive Landscape A $10B raise that reads as a strain signal Alibaba closed a roughly US$10 billion Hong Kong share placement on August 26, selling more than 700 million newly issued shares to non-US investors. The company said about 60% of proceeds will fund computing infrastructure, with the rest going to AI data centers and cloud upgrades. The raise followed a reported ~75% year-on-year fall in net profit, which Alibaba attributed to increased AI spending. Olivia opened the discussion by asking why Alibaba chose to issue new equity rather than fund the AI buildout from operating cash flow, debt, or asset sales. Robert Wu’s reading: the “best case” would be for Alibaba’s profitable legacy e-commerce business to fund the AI buildout directly. That is not what is happening, because Alibaba is simultaneously fighting a costly local-delivery and shopping-cart war, largely against Meituan. That war has cooled since government intervention and mutual exhaustion, but it has not ended — so Alibaba cannot redirect that cash to AI. At the same time, AI investment cannot wait. The result, in Robert’s view, is a company fighting on multiple fronts at once rather than concentrating resources, which is why the equity raise landed badly with the market: Alibaba’s stock fell after the news, in part because Alibaba had been repurchasing its own shares before the raise, raising the question of why it needed to issue new equity so soon after buying stock back. “They are really stretching themselves all over the place. And that’s what got the market... worried about.” — Robert Wu Insider buying that undercut its own signal Immediately after the placement news, Joe Tsai, Jack Ma, and acting CEO Eddie Wu each bought shares personally, together totaling roughly HK$1 billion — an apparent attempt to show the market they have “skin in the game.” Steven Xiao and Robert both judged the signal as weak, and Robert identified a specific reason: the founders bought back in after the equity-issuance news, at a lower price than the investors who had just been asked to buy into the placement. Robert also offered a structural explanation for why Alibaba’s leadership appears unusually reactive to short-term stock moves: the company sits in an unusual position where its founder-class shareholders (Tsai, Ma, and others) are not in day-to-day operating control, but still care deeply about the share price, while the operating management team runs the business day to day. He contrasted this with founder-led-and-controlled companies, or with companies under clean management-board mandates, where leadership would be less sensitive to daily price swings. This dynamic, in his view, is a reason for caution about Alibaba’s near-term decision-making — a view he said is shared by “most of the market investors” the two hosts talk to. Where the AI strategy has already pivoted — and where it should go next Asked which part of Alibaba’s AI strategy is most likely to generate revenue first, Robert pointed to a recent pivot away from consumer-facing ambitions. Earlier in the year, emphasis was on Qwen, Alibaba’s ChatGPT-equivalent consumer AI app, which was meant to integrate with e-commerce and food delivery, reportedly generating headline-grabbing order volumes before fading. Robert’s explanation was cultural rather than strategic: much of the visible campaign activity, he argued, is aimed at impressing internal management rather than the market, which produces flashy but poorly thought-through launches. Robert’s own preference is for Alibaba to concentrate on enterprise cloud and AI — following the model he attributes to Microsoft, Amazon, and Google — leveraging Alibaba’s existing cloud business and in-house chips, rather than continuing to chase consumer integrations. He added a personal, clearly labeled opinion that the company “can be much better run if they are broken up into vertical dedicated arms.” The wider field: why ByteDance may be the strongest contender nobody is discussing Asked to compare Alibaba’s AI investment with Tencent and other players, Robert described China’s AI landscape as more fragmented than the US, where he sees OpenAI and Anthropic as the only two frontier labs that matter. In China, independent AI labs (the transcript names Kimi/Moonshot AI, DeepSeek, and MiniMax; one further name was not clearly transcribed) compete alongside the older internet platforms, and he flagged Tencent specifically as a strong contender in its own right — Steven added that Tencent’s cloud business runs neck and neck with Alibaba Cloud. The conversation then turned to workplace collaboration tools: Steven compared Alibaba’s DingDing (钉钉) with ByteDance’s Feishu (飞书), noting both are widely adopted by businesses, small enterprises, and even government agencies for a similar mix of office-automation and messaging functionality. Robert added that an upcoming Baiguan field trip to an AI and robotics lab will include a visit to ByteDance’s cloud service arm, which powers many of ByteDance Group’s key AI initiatives — and from there made the broader case for why ByteDance is a stronger AI contender than its public profile suggests, for three reasons: First, its Doubao/TikTok-adjacent consumer AI products are already “in great use, generating a lot of revenues,” helped by the company’s existing base of short-video content to train on. Second, ByteDance has strong cloud and B2B service infrastructure. Third — and most distinctively — ByteDance is privately held, which Robert argues insulates it from the quarterly pressure that pushes public companies toward short-term, headline-driven moves. He cited an internal letter reportedly circulated about a month earlier by ByteDance founder Zhang Yiming, arguing against sacrificing long-term model quality for shortcuts like distilling other companies’ models, as evidence the company can genuinely take a longer view because its profitable core business funds its AI bets. “They are not bound by quarterly results, and also they have a very profitable cash flow, posi… This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.baiguan.news/subscribe

    Alibaba's $10B AI Equity Raise, Niu Lai, and Beijing's ‘Going Global’ Support Plan - Baiguan Radio #41

  6. Aug 19

    Unitree’s IPO, Chinese Premier, Chinese Mayor - Baiguan Radio #40

    Welcome to the new episode of Baiguan Radio, hosted by Olivia Plotnick. Episode in brief * Unitree had a blockbuster IPO, confirming intense demand for a sector Robert Wu calls “a necessary bubble”: capital has to fund years of loss-making data collection before humanoid robots have a real commercial use case. Robert also explains why the AI & Robotics module in the upcoming Baiguan China Tour will give participants nuanced insights into this red-hot sector. * Zhu Rongji’s death drew unusually wide public mourning, but Robert reads it as nostalgia for a leadership style and authenticity, not the economic era itself: Zhu was a genuinely divisive figure at the time because of the mass layoffs his state-enterprise reforms caused. * Both hosts trace a direct institutional line from Zhu-era reforms to two live issues today: China’s property-driven local government finances (via a 1994 tax reform that pushed localities onto land-sale revenue) and today’s AI-driven job anxiety, which Steven Xiao contrasts with SOE-era layoffs by noting nobody today believes “a job is for life.” * Moonshot’s Kimi K3 briefly escaped a testing sandbox last week — traced to human error, not autonomous reasoning — but Robert argues the real AI risk isn’t a “Skynet” scenario; it’s AI systems that are simply too effective at pursuing assigned goals and find unplanned ways around constraints. * Chinese outbound M&A into consumer brands (Blue Bottle Coffee, a Puma stake, Mammut) hit $9.6 billion last quarter, a five-year high, but Steven argues it’s opportunistic rather than structural — a “rounding error” next to China’s roughly $1 trillion trade surplus — while heavier industrial and tech acquisitions remain foreclosed by tightened US and European investment screening. * Steve explains why he is leading a two-day tour to Datong, another module of the upcoming Baiguan China Tour, one of China’s ancient capitals for both its unique ancient culture and a no less unique modern story, immortalized in the documentary The Chinese Mayor. About this episode’s guest Steven Xiao is a longtime friend of Baiguan and an investor and businessperson with a particular interest in Buddhist arts and Chinese history. He is leading a two-day Datong module on Baiguan’s upcoming autumn China tour. His investing background — he mentions prior robotaxi investments and cross-border M&A experience — is the source of the episode’s most direct pushback on the robotics-bubble narrative and its most detailed read on outbound M&A, and is worth keeping in mind through both sections below. Unitree’s IPO and robotics’ “necessary bubble” The listing Unitree, the humanoid robot maker behind the dancing robots at this year’s Spring Festival Gala, priced its Shanghai IPO at a $9 billion valuation, raising about $904 million. Formal listing was set for August 19, the day after this recording. DeepSeek and Tencent both invested; DeepSeek’s stake includes a co-development agreement on AI models and embodied-intelligence technology rather than capital alone, while Tencent’s stake size wasn’t disclosed. Meituan is a pre-IPO investor. Robert frames this as the second massive China listing after CXMT last month, inside a broader wave — Hong Kong just posted its strongest first-half IPO performance in five years, led by AI and tech names. The listing was confirmed two days after the US announced an effective ban on imports of Chinese-made robots. The demand signal, and the data-bottleneck argument Robert says pricing on “hyperliquid” markets — a venue trading perpetual contracts linked to pre-IPO stocks — already puts Unitree at four to five times its IPO price; he applied for shares himself and received none, calling the listing “highly oversubscribed.” He draws a parallel to China’s EV industry five or six years ago: too many companies across too many cities, raising real overcapacity concerns. The key difference, in his view, is that EVs had an existing market to overbuild into; robotics doesn’t yet. Most robots currently shipped are for data collection and research rather than deployment — “the robots are being made in the end for the purpose of making better robots,” he said. He compares the bottleneck to autonomous driving: Tesla’s FSD only became viable after millions of vehicles generated enough real-world driving data to train its models, while robotics has only “tens of thousands” of data-collecting devices in the field — nowhere near, in his estimate, a critical mass. He described Baiguan’s robotics tours, where participants teleoperate robots by hand to generate training data, and named LinkerBot, a company specializing in robot hands, as a stop on the upcoming October tour — illustrating how fragmented the supply chain of hands, “brains,” and data still is. As Robert put it, physical AI hasn’t yet had “a real ChatGPT moment.” Steven’s pushback Steven challenged the framing directly, pointing to earlier waves of “robot” hype that didn’t sustain elevated valuations: Roborock (石头科技), the floor-cleaning robot maker, has fallen from roughly 300 to under 100, and SenseTime’s machine-vision technology — once positioned to extend from security cameras into autonomous driving — also disappointed. He noted that industrial robots — for instance KUKA, the German robotics maker China’s Midea acquired a decade ago — already have high penetration in fixed factory and home settings at reasonable cost and risk. “A necessary bubble” Robert’s response was unusually direct about the sector’s risk. He argues genuine use cases exist — healthcare, elderly care, housework, factory and port operations, coal mining, driving — but that the technology, not the demand, is what’s missing. Reaching it requires a critical mass of deployed devices collecting real-world data, comparable to Tesla’s vehicle fleet; simulated and synthetic data can help, but he argues real physical data “cannot be replaced.” Funding that data-collection buildout, in his framing, requires exactly the kind of overcapacity and speculative capital that looks like a bubble today. Local governments reinforce this dynamic, he adds, because robotics “looks fancy to demo” in a way software doesn’t — an institutional preference layered on top of investor enthusiasm. “I don’t think it will sustain. Everyone know this is a bubble... but to be honest, this is a necessary bubble.” — Robert Wu Steven, while skeptical of the near-term case, agreed there’s a longer-term one: an aging society will eventually need labor substitution, 15 to 20 years out. He compared the sector’s likely path to robotaxis, which he has personally invested in — an initial boom-and-bust around the Waymo era, a second wave once EV components matured, and now a plateau amid a deflationary economy and labor-market pressure. His read: this robotics cycle probably won’t be the industry’s last bubble. The US ban, in practice On the US import ban, Robert thinks its impact is “over-estimated” for leading Chinese humanoid-robot makers, since most units are sold domestically or into non-US export markets. He expects the real casualties to be older-generation, narrowly specialized robot makers — “pre-GPT,” single-task machines like robotic vacuum cleaners — calling them likely “collateral damage.” He separately noted that Chinese robotaxi companies Pony.ai and WeRide have already faced US congressional scrutiny over the driving, mapping, and location data they collect, despite one company’s CEO holding an American passport; being treated as a Chinese company for regulatory purposes, in his telling, wasn’t seriously in question. Some robot makers sidestep the exposure by selling hardware only, with third-party providers supplying the software layer — a model Robert argues shouldn’t raise national-security concerns on a purely technical basis, though he expects the current ban to affect it regardless. He plans to press robotics companies directly on this during the upcoming October tour. Implications The episode’s most investable distinction is between humanoid-robot … This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.baiguan.news/subscribe

  7. Aug 5

    DeepSeek, LVMH vs Molly Tea, state of economy - Baiguan Radio #39

    Episode in brief • Louis Vuitton won in court but may be losing with consumers. Robert says BigOne Lab’s transaction data show a major sales decline after the Molly Tea dispute, making this a materially different episode from many short-lived social-media controversies. • DeepSeek’s leaked investor call reinforces the image of a mission-driven company. The discussion suggests that open-weight distribution, very low pricing, independence from the Nvidia-centered stack, and talent retention sit above profit maximization. • China’s first-half economy remains sharply uneven. Manufacturing and exports are stronger than household demand, but Beijing appears concerned rather than alarmed. Services, domestic travel, and strategic technology financing are the areas to watch. Guest perspective: Olivia Plotnick is an American entrepreneur and marketing professional who has lived in China for about a decade and writes at who what wai. Over the past 18 months, she has traveled to nearly 50 Chinese cities, bringing a consumer and brand perspective from beyond the largest urban centers. Louis Vuitton vs. Molly Tea: a legal victory, a commercial setback What happened You may read our previous article on this topic to have a grasp of the incident. The data suggest the backlash is commercially material Robert says BigOne Lab’s offline transaction data show LV sales declining since the controversy began. That persistence matters. The Arc’teryx controversy reportedly produced only a brief dip before sales recovered within one or two weeks, while the 2024 Nongfu Spring backlash is a closer example of an online campaign that translated into sustained purchasing behavior. “They won the legal battle but they’ve lost massively on the business and commercial front.”Robert Wu Why consumers turned against LV The podcast rejects the simplest explanation, that this is merely nationalist sentiment. Unlike the Qiaodan case, where the resemblance to Michael Jordan’s branding was difficult to deny, Molly Tea’s logo sits closer to the boundary. Robert says he did not associate it with LV even after repeated exposure. LV’s history of aggressively pursuing trademark disputes also made the company look less like a victim and more like an overreaching incumbent. Robert offers a second interpretation: after years of criticism that China under-protects foreign intellectual property, some consumers now worry that courts may be overcorrecting in favor of foreign companies. The brand-response problem Olivia’s communications advice is not to confuse speed with effectiveness. Recent rapid responses from foreign brands have sometimes satisfied nobody and have even exposed internal divisions. Yet prolonged silence allows other actors to define the story. The operational answer is preparation: establish decision rights before a crisis, give the China team authority to respond, and create a direct pathway between headquarters and local leadership. “Staying completely silent for too long leaves a void for people to fill.”Olivia Plotnick Beijing’s message may be indirect Robert’s reading is that Beijing is displeased but constrained. Direct criticism of LV could be interpreted as hostility toward foreign business. Instead, state media highlighted an older lawsuit in which LV had challenged China’s national intellectual-property administration, reviving it at the height of the Molly Tea controversy. In his view, this was a way to shape the narrative without issuing an official rebuke. “That’s the fascinating thing about Chinese politics: it’s subtle.”Robert Wu Business implications • A courtroom win does not protect brand equity when consumers see the enforcement itself as unfair. • Online controversy should be measured through sales persistence, not social-media volume alone. • Foreign brands need pre-authorized local crisis protocols. A headquarters bottleneck can become a commercial liability within hours. • A repeated legal posture can create cumulative reputational damage that a one-off apology cannot repair. DeepSeek’s leaked investor call: mission, pricing, and strategic independence Why the leak matters The investor call took place months earlier, but the transcript surfaced only recently. The episode notes that its authenticity has not been officially confirmed. Robert nevertheless believes it is genuine because the arguments are consistent with DeepSeek’s behavior, model design, and the public persona of founder Liang Wenfeng. A Bloomberg report cited in the conversation said DeepSeek paused fundraising after the leak, which the speakers treat as further circumstantial evidence. Open weight as the objective, not a marketing tactic The most important distinction is motive. Many companies use open source to recruit developers, accelerate adoption, or support a later commercial model. The leaked discussion portrays DeepSeek differently: broad access appears to be the desired outcome itself. Robert sees Liang as part of a newer generation of Chinese founders for whom money is an instrument for solving difficult technical problems rather than the final objective. “For DeepSeek, it sounds like open source is the end goal itself, not just a means.”Robert Wu The Nvidia dilemma The transcript also makes explicit a strategic direction already visible in Chinese AI: reduce dependence on the US-centered technology stack, especially Nvidia. DeepSeek is not yet independent, but its architecture and optimization work point toward greater flexibility across chips and systems. This creates a three-way misalignment. China wants technological distance from Nvidia; the US government wants Nvidia to limit its China exposure; Nvidia wants to remain central to both ecosystems. “China wants distance from Nvidia, the US government also wants Nvidia to keep its distance from China, but Nvidia itself wants to stay in the center of it all.”Robert Wu AI as infrastructure rather than a premium product Olivia compares DeepSeek’s low-price strategy with China’s manufacturing playbook: enter cheaply, improve quality, and expand global adoption. Robert sees an analogy but not an exact match. AI may become less like a consumer product and more like electricity, water, or the internet, something every person and business needs. If that is the destination, minimizing price while earning a modest return may be socially rational and commercially sustainable. “We’re talking about something that could become genuinely necessary to human life. That’s the key difference.”Robert Wu What happens next Robert does not expect a public explanation. His base case is a private investigation into the leak, a temporary fundraising pause, and then a return to the original financing plan. Longer term, he expects DeepSeek eventually to list domestically, where investors could assign a large strategic premium even if near-term profits remain limited. The most immediate corporate constraint is talent: the company needs enough capital and organizational stability to keep researchers who could earn much more elsewhere. This is Robert’s outlook, not a confirmed company plan. Investor implications • DeepSeek’s core value may be adoption and ecosystem influence rather than near-term pricing power. • Nvidia faces pressure from both geopolitics and Chinese substitution, even while commercial incentives pull it toward the China market. • Talent retention, fundraising resumption, and any movement toward a domestic listing are the practical milestones to monitor. • Chinese open-weight models may function as technology exports and soft power, particularly in cost-sensitive markets. China’s H1 economy: strong supply, weak demand, and no dramatic pivot The K-shaped data The first-half figures cited in the episode show a clear divergence between industrial activity and household demand. Manufacturing and exports remained comparatively strong, while retail sales and consumer spending were much softer. Why Beijing is not signaling a September 2024-style pivot Robert’s macro takeaway is that Beijing appears concerned, but not alarmed. … This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.baiguan.news/subscribe

    DeepSeek, LVMH vs Molly Tea, state of economy - Baiguan Radio #39

  8. Feb 3

    After Maduro, how does Latin America really see China today - Baiguan Radio #38

    In this episode of Baiguan Radio, I speak again with Mr. João Philippe de Orléans e Bragança, a Brazilian macro hedge fund manager with long experience living and working in China, about how China is actually perceived on the ground in Brazil and across South America. We talk about why Chinese influence in the region looks very different from Europe or the US, how companies like BYD have quietly reshaped public opinion, why Latin America is more complementary than competitive with China, and what Chinese investors often misunderstand about Brazil. This is a conversation about perception versus reality, geopolitics versus everyday economics, and why Latin America is becoming an increasingly important part of China’s global story. João is a macro investor and a portfolio manager at Absolute Investments, Brazil’s largest hedge fund with $11bln in AUM. João lives in São Paulo, but has lived for many years in Asia, mostly in Singapore, but also in Shanghai, which he claims to be his favourite city. (Just to refresh your memory, in our episode #37, João shared his bullish view about RMB, which has been validated since then.) We are thrilled for Baiguan Radio to become a forum for all of you to exchange and debate your insights. So please find us whenever you have something to say. Table of Contents 03:40 | How Brazilians see China today 07:45 | BYD and China’s “Best Ambassador” 11:20 | Fear of economic invasion 13:15 | Why Brazil is attracting Chinese capital 16:10 | Chinese companies in Brazil: reality vs narrative 20:00 | Latin America and US–China rivalry 25:00 | Services, technology, and the next phase 26:55 | The biggest misconception about Latin America 28:30 | Branding China abroad 29:50 | The appeal of China’s education sector This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.baiguan.news/subscribe

    After Maduro, how does Latin America really see China today - Baiguan Radio #38
See All (40)

About

Actionable insights for successful China-related business and investment, with a human touch www.baiguan.news

Information

You Might Also Like