Chinese multinational technology company Baidu on Tuesday claimed that its AI (artificial intelligence) chatbot ‘Ernie’ will not make mistakes on important and sensitive issues like its competitors.
Baidu CEO Robin Lee told JEE News the company was awaiting government approval before launching its Ernie bot.
Its chatbot Ernie does not respond to questions on topics such as politics, particularly those related to Chinese government leaders, JEE News reported.
“For important and sensitive topics, we need to make sure that the artificial intelligence will not cheat,” Lee said, using the industry term for when AI models produce different results than expected.
Referring to the AI models used by many chatbots such as ChatGPT, he added: “Given that the LLM [large language model] is more or less a probabilistic model, the task is not trivial at all.”
“The industry regulation was not yet final, and the company will continue to update its strategy as it evolves,” CEO Lee said.
“Beidou has been conducting search operations in China for more than 20 years and has extensive experience of the Chinese culture and regulatory environment,” he said.
Baidu’s CEO also said: “In contrast, companies that do not have extensive experience in providing appropriate online content or a track record of working with regulators will face significant challenges.”
Last month, draft measures were made public by China’s cyberspace regulator, which is responsible for regulating services powered by generative AI technology such as Baidu’s Ernie bot. It said: “The material produced by this frontier technology had to be aligned with the country’s core socialist values.”
Lee said these initiatives will benefit Baidu.
“We believe that regulators’ active engagement in early-stage generative AI will raise the bar for entry, and we are well-positioned for that,” Lee added.
