In a recent turn of events, Chinese AI firm iFlyTek witnessed a sharp decline in its shares, plummeting by a significant 10 per cent on Tuesday. The company, renowned for its advancements in artificial intelligence, faced a backlash as social media users raised concerns about its AI-powered study assistant tablet. According to reports, the device allegedly generated an essay that criticized Chairman Mao Zedong, a revered figure in China’s history.
Several individuals on platforms such as Baidu’s Baijiahao shared accounts of their grievances, detailing how the iFlyTek study assistant tablet reportedly labeled Mao Zedong as “narrow-minded” and “intolerant” for initiating the notorious Cultural Revolution in China. However, Reuters was unable to independently verify these claims, and iFlyTek has remained unresponsive to requests for official comments.
A customer service representative from iFlyTek’s website, when prompted about the incident, confirmed that the matter had been addressed. Additionally, Chinese media outlet Cailianshe cited iFlyTek’s founder and Chairman Liu Qingfeng, who disclosed that a supplier had provided the contentious content for trial purposes. Upon the detection of the issue, both the supplier and iFlyTek personnel faced punitive measures, as mentioned by Liu.
The company also reportedly reinforced the content review mechanism of the study assistant tablet, striving to prevent such occurrences in the future. It is imperative to note that China’s strict censorship protocols closely monitor content that might contradict government policies or exhibit dissenting views toward political figures, including historical leaders like Mao Zedong, who remains officially revered by the ruling Communist Party as the founding father of modern China.
The iFlyTek incident sheds light on the unpredictable nature of generative AI technology, which has the potential to subvert the government’s stringent censorship tools. Notably, China has been experiencing a burgeoning trend in the field of generative AI, prompting the authorities to introduce comprehensive regulations for companies intending to introduce AI-driven products to the market. Among these initiatives, the proposal for a blacklist of unsuitable sources for training Chinese AI models stands out as a significant step towards ensuring compliance with established norms.
iFlyTek’s future endeavors and collaborations
Despite the recent controversy, iFlyTek continues to pursue advancements in the AI sector, revealing its latest Spark AI model, touted to rival the capabilities of OpenAI’s ChatGPT across various crucial aspects. Liu Qingfeng emphasized that iFlyTek’s model is gearing up to compete with OpenAI’s anticipated GPT-4 as early as the following year. Furthermore, the company announced its collaboration with Huawei Technologies on an AI model training platform, leveraging the capabilities of Huawei’s flagship AI chips to enhance its AI development endeavors.
As the situation unfolds, industry observers remain watchful of iFlyTek’s response to the recent crisis, while the company strives to reinforce its position in the competitive landscape of artificial intelligence.
