Alibaba Releases Chatbot As China Proposes Measures For Online Ads, AI
As China publishes proposed rules around advertising and artificial intelligence (AI), one of the country’s largest and most successful tech giants and marketplaces on Tuesday unveiled Tongyi Qianwen, the company’s artificial intelligence-based chatbot rival to ChatGPT.
Alibaba’s Tongyi Qianwen, which means “truth from a thousand questions,” can answer questions in English and Chinese. The company will initially implement the chatbot across its messaging app DingTalk and its smart digital assistant Tmall Genie.
The technology “will bring about big changes to the way we produce, the way we work and the way we live our lives,” CEO Daniel Zhang said at the livestreamed event.
Similar to OpenAI’s ChatGPT-4, Alibaba plans to open Tongyi Qianwen to clients and developers, allowing them to build custom large language models. Registration begins Friday.
The news follows the release of AI generative chatbots by Google, Microsoft, Baidu, and others.
China’s proposed rules — open for public comment through May 10 — come as China looks for ways to regulate advertising and generative AI technology has sparked concern about ethical implications as well as its impact on national security, jobs and education.
The draft rules, published by the Cyberspace Administration of China, state that the country supports the technology and innovation, but content generated must adhere to “core socialist values” as well as to laws on data security and personal information protection, Reuters reports.
China is not the only country attempting to put measures in place to regulate AI. The Biden administration is trying to determine whether the U.S. government needs to put in place checks and balances for AI tools such as ChatGPT, as concerns mount over the technology that could be used for misinformation and to spread harmful information.
The U.S. Commerce Department Tuesday published a formal public request for comment on accountability measures. Those measures include whether or not potentially risky AI models should go through a certification process before being released.
“It is amazing to see what these tools can do even in their relative infancy, Alan Davidson, leader of the National Telecommunications and Information Administration — the Commerce Department agency that put out the request for comment — told The Wall Street Journal. “We know that we need to put some guardrails in place to make sure that they are being used responsibly.