Chinese technology giant Baidu have been making massive strides in Artificial Intelligence. This time, Baidu CEO Robin Li recently asserted that the company’s chatbot Ernie won’t make errors on important or sensitive topics. The chatbot, according to Li, will be tailored to meet China’s core socialist values, as indicated by drafts made public by the country’s cyberspace authority.
Chatbot Ernie’s development follows the creation of AI models that generate outputs based on probabilistic models, a framework mentioned by Li while referring to AI models used by rivals like ChatGPT. The so-called “large-language model” is an algorithm of language composition, based on user input. This could be used to answer questions about topics such as politics. However, as it is not yet a done deal, Baidu is still to file for government approval before the official launch of Ernie.
Apart from Ernie, Baidu and its CEO Robin Li have been quite involved in the sphere of Artificial Intelligence. Li has had a rather strong focus on AI and technology, even taking note of the vague regulations as of yet, and how these draft measures could benefit the firm. Additionally, Li has remarked that companies lacking a track of record with the regulators will face certain challenges.
It’s no surprise that Baidu has been operating a search engine in China for the past 20 years – given its success in mastering Chinese culture and regulatory circumstances. Nevertheless, the company’s CEO is still wary of possible “hallucinations” of AI models, emphasizing the need for caution while providing the technology with appropriate online content.
It remains to be seen if the effectiveness of the Ernie bot in terms of meeting China’s vision for a socially acceptable AI model. Nevertheless, with Baidu’s experience with Chinese regulation, it could very well be at the forefront of developing a such a chatbot.
