China Joins The Generative AI Race, But AI Chatbots Are Not Available To The Public

China is participating in the global race of generative artificial intelligence by having nearly 80 artificial intelligence models (photo: dock. pexels).

China has taken part in the global race for artificial intelligence, which has nearly 80 artificial intelligence models from companies such as Baidu and Alibaba, and received funding of nearly $14 billion in the last six months.

However, unlike in the West, where OpenAI's ChatGPT has attracted more than 100 million monthly users who provide input to help improve and refine its products, no Chinese AI chatbots are yet available to the general public.

This reflects a new reality in China's technology sector and the effectiveness of regulatory tightening carried out by the Beijing government over the years, which peaked with a large fine against Ant Group and Tencent last week.

It used to be known for its competitive "market-to-market" spirit, especially in the consumer internet sector, companies are now slowing down their steps to comply with the rules and follow government directives.

Previously free-to-roll industries now tend to obey, with companies starting to change their business, slowing expansion into new areas, and cutting off the employment of thousands of employees.

The new rules that Chinese companies must comply with include checking algorithms and monitoring security of data to be exported.

While Beijing now controls what it considers to be the advantages and inappropriate behavior that arises as a result of the industry's wild growth, some experts are starting to provide warnings about its impact on future competitiveness with Western competitors, given the balance these companies have to achieve.

Chinese regulators last week said most of the problems that exist in platform companies in the country have been resolved, a sign that regulatory tightening has finally ended.

However, Henry Gao, a professor of law at Singapore Management University, said that a strict regulatory environment, which is now stricter than any other country in many aspects, is expected to continue.

"Over the past few years, there has been so much what I call 'preventional regulation' in China. This is definitely hampering innovation and slowing down the ability of Chinese companies to catch up," Gao said.

Industry executives say these challenges are particularly felt in generative artificial intelligence, as global interest in this technology increases after the success of Microsoft-backed ChatGPT, while Chinese companies are trying to catch up with their competitors in the United States.

Although several widely discussed launches have been carried out in China, such as the Ernie Bot of Baidu and SenseChat of SenseTime, all are still in the beta test phase and are only available for selected users.

In May, Baidu CEO Robin Li said they were still waiting for government approval for Ernie Bot.

"As the official regulatory framework has not been established, many Chinese internet companies cannot conduct mass testing of their generative artificial intelligence products that are being developed or have been developed," said Jialong Shi, head of China's internet equity research at Nomura.

Having more users helping ChatGPT to improve its model to pay more attention to language patterns such as slang and idiom, detect errors, respond to unusual scenarios, and reduce cultural bias in its response, according to the chatbot.

One of the main reasons behind China's reluctance to release AI chatbots is because Beijing is worried that unsensored chatbots could influence people's views in potentially subversive directions, said Mark Natkin, director of research firm Marbridge Consulting.

In April, China detailed a series of draft steps for a generative AI service stating that companies should submit security assessments to authorities before launching their products.

Analysts have criticized some of these rules as too burdensome, such as the obligation for AI service providers to verify truth and accuracy not only from what AI models generate, but also from the materials used to train them.

The measures are expected to be completed and enforced by the end of this year. The Chinese government is also preparing a special AI law but has not provided details.

Of course, China is considered more advanced in terms of regulation compared to other countries in the world struggling to establish rules governing this very popular technology, taking safety and copyright into account while maintaining a conducive environment for innovation.

However, Heatherm Huang, co-founder of Hong Kong-based technology company Measurable AI, said that this puts Chinese companies in an unfavorable position, as products from the United States such as Google Bard and Microsoft Bing go straight into the market.

"It's like limiting the speed of racing cars - it's slowing everything down," he said. "While the United States is racing forward with AI, China is slowing it down with "more" rules."

Business Focus Most Chinese technology companies are currently more focused on seeking applications for industrial use, which is in line with Beijing's priorities and the main goal of a new regulatory framework.

Chinese President Xi Jinping has stressed the importance of directing more energy on "hard" technologies such as semiconductors and industrial artificial intelligence applications, as well as breakthroughs that could help China reduce its dependence on Western technology.

Huawei Technologies said last week that their AI Pangu model would serve company customers more for applications such as goods train security inspections and weather forecasts.

Baidu announced last week that more than 150,000 companies had signed up to test Ernie Bot since its launch in March, with more than 300 companies testing it in the field such as increased office efficiency, customer service, and marketing.