When Boring is a Survival Strategy
[Opinion] China tech's old guard are pushing into AI. But now's not the time to be famous for it
Good Evening from Taipei,
Robin Li, Eddie Wu and Pony Ma are three of the longest-serving founders in technology history. All three men started building their empires at a time when Mark Zuckerberg was still a schoolboy.
Li and Ma’s quarter century leading Baidu and Tencent respectively are particularly impressive. Wu, a co-founder of Alibaba but only a recent entrant to the CEO club, has also seen a lot.
Today, their businesses are boring. In the US, boring would be a corporate death sentence. In Xi Jinping’s China, boring may be a survival strategy.
They weren’t always so dull. Baidu is the most staid of them all, but its long history has plenty of highlights. It pioneered web search in China and many of its innovations in ad sales predate Google. Robin Li has since ventured into streaming video and autonomous driving.
Alibaba, synonymous with its founding front man Jack Ma, started in web commerce and morphed into a fintech and logistics giant. Tencent has been an exciting company for most of its life. It started off with the instant messenger QQ, a Chinese copy of ICQ, but was quick to innovate in all directions. Weixin (aka WeChat) created the playbook for what we now call a super app. Along the way it’s moved into payments, bought global games & video companies, and added logistics.
In the past few years though, not much seems to have changed among China’s old guard. Across the other side of the world, there’s been a lot happening. Facebook got a new name, Amazon got a new chief, and Microsoft got a facelift. (Google is still Google, but at least YouTube makes it fun).
Sure, there’s new players setting the sector alight. ByteDance is the biggest and most famous, but Temu (part of PDD) and Shein are also gaining traction, especially in foreign markets.
Ma, Li and Wu — MaLiWu — ought to be forgiven, if not praised, for the lack of excitement inside their businesses.
Innovation Culture
At the heart of the challenge is the suppression of China’s innovation culture. That’s not to say Chinese entrepreneurs and engineers have stopped innovating, the country’s advances in artificial intelligence and global e-commerce show that there’s still a spark. But it says a lot that entire executive teams from the nation’s top startups have decamped to friendly locales like Singapore and the US.
Xi’s sweeping crackdown across sectors like education, finance, healthcare, content, and games has disincentivized an entire generation of fresh graduates from pushing boundaries. What many have learnt is that falling foul of the Party’s vision for China can result in their businesses being shuttered or, at worst, being blacklisted or black-jailed.
As the old guard of China’s tech industry, MaLiWu have seen this before. Crackdowns come in waves, usually targeting one subsector for a year or so before moving on to the next. They’ve learnt to hunker down and wait for the storm to pass.
This time it feels different, and you can see the results in the dullness of their recent investor conference calls. Executives talked about ad inventory and operational efficiency, not exciting product launches or fantastic opportunities in new realms. Among the reasons is the cloudy outlook for the world’s second-largest economy. Seeking to stem a meltdown caused by more than a decade of profligate real-estate construction, Chinese leaders are trying to boost the economy without stoking inflation. We see light-grey words like “optimism” coupled with dark-grey terms such as “subdued.”
It’s not just the economy. It’s Xi, the Party, and a China that’s vastly different to the go-go days of the early 2000s and 2010s. Regulators previously approached new industries with a strategy of allowed unless specifically prohibited. Alibaba and Tencent took this to heart, dabbling in money-market funds, payments, loans and all manner of financial products that definitely expanded beyond their playpen. This innovation often got curtailed, but there was a sense of “no harm, no foul” so they kept trying new things. Now it seems like there’s no room to make mistakes, especially with Party members installed inside every company to watch and chide.
Old Dogs Learn New Tricks
But they’re not being entirely held back. The hunger is still there. Innovation is still happening. These old dogs are learning new tricks. Alibaba is at the forefront of AI, not just in China but across the globe. Its open-source Qwen model is highly regarded and measures up against US counterparts like Meta’s Llama.
Baidu is also serious about AI with its Ernie bot, has been pursuing autonomous driving for a decade, and recently debuted AR smart glasses. There’s also reports it’ll be Apple’s Chinese partner for on-device AI, allowing the US company to skirt Chinese restrictions.
AI is Not Smart
My concern is that while Baidu, Alibaba and Tencent make inroads into AI, they may be doing so too well and attracting too much attention. A decade ago this kind of kudos was sought after because it attracted partners and investors. In today’s China, the tallest nail gets hammered.
They’re not the only ones making progress. Startup DeepSeek has quietly crept up the global AI charts. Jordan Schneider and his team at ChinaTalk have a great piece on its CEO Liang Wenfeng, including an English translation of an interview Waves did with him in July. Then there’s Zhipu AI and Bytedance.
But as major incumbents — and among the nation’s biggest companies — Baidu, Alibaba and Tencent have the most to lose by winning in AI.
Right now, Beijing seems fine with the progress that local tech companies are making in this emerging field. Anything which can show off China’s technological prowess is worth celebrating. Except AI is also pushing the boundaries of state control in an environment where the state very much wants to be in control.
That concern might explain why Chinese generative-AI developers, big and small, haven’t pushed much into consumer products in the same way that OpenAI and Google are doing in the West. Put simply, they may be courting trouble if they let users loose on generative AI products which can’t be controlled — even if they’re likely built on “safe” training data which has already passed censorship. Imagine the hell which would break loose if Chinese generative-AI were to hallucinate text and images of Xi Jinping or Winnie the Pooh (or both).
That said, I predict that in a few years the Party will decide enough is enough and start reining-in the nation’s AI leaders. Doing so could either kill the sector or reorient it towards the Party’s needs in military development, domestic propaganda, and foreign disinformation.
Tencent may be taking the smarter, more cautious route. The company doubled its capex last quarter, almost exclusively to fund the purchase of graphics-processing units which are at the heart of AI computing. One reason for the surge is that US-China tensions mean that many companies are rushing to buy high-end American chips before more bans are imposed.
Interestingly though, the games and social-media company doesn’t seem to be pushing hard in AI despite being a key player in the cloud-services market. Instead, executives said Tencent was using AI chiefly for content recommendations & ad targeting. That’s like flying at Mach 5 just to slice five minutes off a boba tea delivery.
Tencent’s Pony Ma isn’t stupid. He knows the company needs to be in AI if it’s to stay relevant, but perhaps Ma also gets that staying low-key about it is the trick to long-term survival. He has, after all, survived a quarter century and understands the value of being boring.
More Articles of Interest: