As China introduces interim law, another player heats up the generic AI race

Too: Is TEMU legit? What you need to know about this shopping app before placing your order

ChatRhino boasts a base of 100 billion parameters, surpassing the 10 billion-parameter benchmark clocked by its previous model, Vega, early last year. JD.com said in a statement on Thursday that Vega topped the General Language Understanding Evaluation (GLUE) list, beating models from Microsoft and Facebook.

glue solution and ranks natural language processing based on nine tasks, which span a range of “dataset sizes, text styles, and degree of difficulty”. OpenAI’s GPT-4, although unconfirmed by the company, reportedly contains over 1 trillion parameters based on eight models. Its previous GPT-3 model has over 175 billion parameters, while the GPT-2 runs at 1.5 billion.

ChatRhino offers more than 100 training and inference optimization tools, which JD.com says supports domain-specific application development, helping customers build their own specialized models more quickly . The vendor claims that a generic AI model for the healthcare industry, for example, can be built in “minutes” with two algorithm engineers, whereas a traditional method would typically require a week and at least 10 scientists. it occurs.

According to JD.com, JD Health’s own big language model, Jingyi Qianxun, is built on ChatRhino and trained on medical scenarios to automatically deploy services, including telemedicine.

Too: Google tests its new AI medical chatbot at the Mayo Clinic

E-commerce merchants can also tap ChatRhino to create visuals, marketing posters and a series of product images from a single product image. JD.com said the AI ​​model could cut production costs for each visual asset by 90%, reducing the time required to complete a task from a week to half a day.

Interim legislation to guide generic AI rollout

ChatRino’s launch comes in the same week that China introduced interim regulations to govern generic AI services in the country.

The Chinese government said the new laws, which came into effect on August 15, are necessary to ensure the healthy development of technology and protect both national security and public interests.

Too: Generative AI is coming for your job. Here are 4 reasons to be excited

in a joint statement Issued by various agencies, including the Cyberspace Administration of China (CAC) and the Ministry of Science and Technology, the government noted that while generative AI creates new economic and social development opportunities, it also poses fake news and data privacy and security risks. It has brought challenges as well. ,

The interim law outlines various measures, the statement said, aimed at facilitating the sound development of technology while protecting the national and public interest and the legal rights of citizens and businesses.

For example, generative AI developers must ensure that their pre-training and model optimization processes are carried out in compliance with the law. These include using data from legitimate sources that comply with intellectual property rights. Should the personal data be used, the consent of the individual should be obtained or it should be done as per the existing rules.

Too: Leadership Alert: The Dust Will Never Settle And Generator AI Can Help

Measures should also be taken to improve the quality of the training data, including its accuracy, fairness and diversity.

Under interim laws, generic AI service providers bear legal responsibility for the information generated and its security. They will be required to sign service-level agreements with users of their service, thereby clarifying the rights and obligations of each party.

When illegal content is exposed, the service provider should take various measures, such as preventing its transmission and improving its use in model training. The concerned authority should also be informed.

Further, the service providers shall take necessary measures if any user engages in illegal activities by using the Generic AI service. These include banning actions, suspending or terminating service, maintaining relevant records and reporting to the relevant authority.

Service providers violating the new laws will face penalties set out in various existing relevant laws in China, including the Network Security Law, the Data Security Law and the Personal Information Protection Law. In cases where there are no provisions for contravention, a warning will be issued along with order rectification which must be completed within a specified period. Failure to comply with such orders may result in suspension of services.

Too: 6 harmful ways ChatGPT can be used by bad actors, according to a new study

China released a draft preview of the law in April, saying that the development of generic AI technologies like ChatGPT could be abused if left unregulated. A separate law went into effect in January, laying down basic rules to prevent misuse of “deep synthesis” technology, including deepfakes and virtual reality. Anyone using these services must label images accordingly and refrain from using the technology for activities that violate local regulations.

In May, the Chinese government unveiled plans to build AI industrial centers and technology platforms across the country to support research and development operations. To date, development plans have been launched for 18 national AI pilot zones and 32 innovation platforms, including Beijing and Tianjin.

In addition to JD.com, local players such as Tencent and Alibaba have also announced efforts to introduce or integrate Generative AI into their products. Alibaba Cloud unveiled its big language AI platform in April, called Tongyi Qianwen, which is currently available to customers in China for beta testing and as an API for developers. The Chinese cloud vendor also introduced a partnership program to promote the development of AI applications for verticals including finance and petrochemicals.