Advertisement
Advertisement
Artificial intelligence
Get more with myNEWS
A personalised news feed of stories that matter to you
Learn more
China’s tech CEOs are taking different approaches towards ChatGPT and related AI technologies. Photo: Reuters

China’s Big Tech bosses take different business approaches towards generative AI like ChatGPT

  • Tencent CEO Pony Ma Huateng and Sohu CEO Charles Zhang Chaoyang have taken a more measured approach
  • In contrast, Baidu CEO Robin Li Yanhong and Alibaba CEO Daniel Zhang Yong are quickly pushing new products through the pipeline

China’s Big Tech bosses are displaying conflicting views towards artificial intelligence (AI), as some eagerly embrace the advancing technology while others warn against hasty adoption amid the frenzy around ChatGPT.

Pony Ma Huateng, founder and CEO of social media and video gaming giant Tencent Holdings, said his company would take a more measured approach in launching generative AI products, despite the potential opportunities.

“I think a lot of companies are in too much of a rush, trying to boost their share prices. That has not been our style all along,” Ma was quoted as saying at the company’s shareholder meeting on Wednesday in a report by Chinese media outlet Thepaper.cn.

Pony Ma Huateng, founder and CEO of Tencent, delivers a speech during the World Artificial Intelligence Conference in Shanghai in 2018. Photo: STR/AFP
Tencent, which reported an 11 per cent jump in revenue for the first quarter, said in its earnings report that it was investing in AI capabilities and cloud infrastructure, and expects AI to be a “growth multiplier”. The company is currently working on its own foundational AI model called Hunyuan.

However, Ma said Tencent was in no rush to launch unfinished products.

“[AI] is a once-in-a-century opportunity like the invention of electricity during the industrial revolution,” he said. “In the grand scheme of things, introducing the light bulb a month earlier wasn’t that important. The key [for us now] is to build a solid foundation of algorithms, computing power, data and more importantly, use cases.”

Ma’s caution was echoed by Charles Zhang Chaoyang, founder and CEO of news portal Sohu.com, who also warned against blindly following trends.

Charles Zhang Chaoyang, Founder and CEO of Sohu, gives a speech at the China Internet Conference in Beijing in 2014. Photo: Shutterstock

Sohu is not building its own AI model, choosing instead to take a “wait-and-see” approach because the sector is “overhyped” in China, Zhang said at the firm’s tech forum on Thursday, according to a report by Chinese news outlet Jiemian.

“ChatGPT has been under development for years in a process that involves a series of things from servers and computing power to knowledge base and tagging. Companies that have no such capabilities but choose to follow suit will squander a lot of their resources,” Zhang said, adding that “even for companies with AI and search capacity, they will need to be well prepared”.

While Tencent and Sohu are taking their time, their local peers, including search giant Baidu and Alibaba Group Holding, owner of the Post, have gone ahead and launched their respective alternatives to ChatGPT, which is officially unavailable on the mainland.
Baidu in March became the first Chinese tech company to launch its ChatGPT rival. At the launch event in Beijing, founder and CEO Robin Li Yanhong acknowledged that the new product, called Ernie Bot, was not perfect, but his company would not wait “because there is market demand”.
Baidu founder and CEO Robin Li Yanhong demonstrates Ernie Bot, the company’s answer to ChatGPT, at a launch event in Beijing in March. Photo: Handout
This week, Li addressed some concerns over the impact of AI advancement, such as job security.

“For humanity, the most egregious peril and the largest contributor to unsustainability are not the uncertainty resulting from innovation. On the contrary, if we stop innovating … it would bring all kinds of unpredictable risks, which would be the biggest menace to humanity,” he said on Thursday at the 7th World Intelligence Conference held in China’s northern port city of Tianjin.

Li said that as AI development transitions from cognitive to generative, productivity will improve significantly. While some people worry that this might lead to the loss of many jobs, Li countered that “while we can see what jobs are disappearing, we don’t know what new roles will be created”.

Li said at the event that Baidu had developed its own chips, named Kunlun, to power its AI services. The company has already deployed tens of thousands of the first- and second-generation Kunlun chips, with the third iteration to be rolled out early next year.
Alibaba Group CEO Daniel Zhang Yong says the company plans to integrate AI capabilities into its products. Photo: Reuters
Meanwhile, Alibaba chairman and CEO Daniel Zhang Yong said on a conference call with analysts on Thursday that the company had started to work with partners to develop industry-specific AI models.

Zhang said Alibaba, whose ChatGPT-like service Tongyi Qianwen had received more than 200,000 beta testing applications from corporate clients, was planning to launch cloud products and enterprise solutions based on its AI model, and integrate AI capabilities into various products, including its workplace collaboration tool DingTalk.

“We believe that all of our consumer-facing businesses can be reinvented with large models to offer a new AI-based service experience for our users,” Zhang said.

Liu Qingfeng, president of iFlyTek, speaks during the 2019 World Artificial Intelligence Conference in Shanghai. Photo: Getty Images

Liu Qingfeng, founder and chairman of Chinese language recognition company iFlyTek, said at the World Intelligence Conference in Tianjin that cognitive large language models (LLMs) have brought about a fundamental revolution to how humans interact with machines, as well as new opportunities in scientific research, businesses and industrial internet.

He said the company’s LLM Spark Desk now leads the Chinese market in text generation, language understanding and mathematics. By October, it will exceed ChatGPT in Chinese capability and get closer to its English capability, he said.

“Traditional business models that rely heavily on human capital will be fully disrupted in the next two to three years,” Liu said.

1