(Yicai Global) June 27 -- The chairmen of Alibaba’s Cloud Intelligence Group, Baidu, and 360 Security Group spoke about how to build safe, controllable, and easy-to-use artificial intelligence at a global internet forum in China yesterday.

“We three pals see eye to eye,” 360’s founder Zhou Hongyi wrote in a social media post accompanied by a photo of himself with Baidu’s Robin Li and AliCloud’s Zhang Yong.

The three attended the World Internet Conference Nishan Dialogue on Digital Civilization held in Shandong province, giving speeches that also elaborated on the progress their respective companies have made in developing large language models.

LLM developers need to build a flaw tolerance and error correction mechanism to achieve a dynamic balance between regulation and development, Li said.

Such ChatGPT-like models need to be secure, reliable, controllable, easy to use, and universal, Zhou said, adding that they should provide help as assistants.

Building safe and reliable AI has become the industry consensus, according to Zhang said, who added that the legal and regulatory frameworks are being improved, creating a sound environment for the sustainable development of the technology and its industry.

More than 200,000 enterprise users have applied to access Tongyi Qianwen, AliCloud’s LLM, said Zhang, who will step down as Alibaba Group Holding’s chairman on Sept. 10 to concentrate on leading its AliCloud unit.

The Tongyi Qianwen Partnership Program covers the sectors of oil and gas, electricity, transportation, finance, tourism and hospitality, enterprise services, and communications, and it will build more customized models for enterprises, he added.

The digital economy, whose key driver is LLMs, should be integrated with the real economy, Baidu’s Li said. Such models can go deep into core business scenarios, innovate on intelligent customer service, supply chains, and systems governance, as well as promote digital transformation, especially in auto production, energy, and transportation, he added.

The main scenarios for future LLM development are enterprises, industries, governments, and cities, 360’s Zhou said, adding that LLMs should advance industrial upgrading from digitalization to intelligence.

Editor: Emmi Laine