JD.com is heating up the market artificial intelligence (AI) races in China with major language model release, even as regulators move to make regulations to govern artificial intelligence service.
China’s second largest online shopping platform, JD.com, says its ChatRhino(or yanxi in Chinese) has been customized to support a number of verticals such as logistics, retail, healthcare, and finance. The big language model consists of 70% generic data and 30% “native intelligence” supply chain data, says the e-commerce company with a logistics division as well as a healthcare business unit. .
Also: Is temu legal? Things to know about this shopping app before you order
ChatRhino boasts a base of 100 billion parameters, up from the benchmark of 10 billion parameters achieved by the previous Vega model early last year. Vega led the Common Language Understanding (GLUE) list, outpacing models from Microsoft and Facebook, JD.com said in a statement Thursday.
Glue measures and ranked natural language processing based on nine tasks, covering a range of “data set sizes, text genres, and difficulty levels”. OpenAI GPT-4 Reportedly, though unconfirmed by the company, there are more than 1 trillion parameters based on eight models. Its previous model GPT-3 has over 175 billion parameters, while GPT-2 runs on 1.5 billion.
ChatRhino offers more than 100 inference optimization and training tools that JD.com says support domain-specific application development, allowing customers to build their own specialized models faster. For example, the vendor claims a general AI model for the healthcare industry can be built in “minutes” with two algorithm engineers, compared to a traditional approach that typically requires a week and as little as a few minutes. 10 scientists.
According to JD.com, JD Health’s own large language model, Jingyi Qianxun, is built on ChatRhino and has been trained in medical scenarios to automate the deployment of services, including telemedicine. .
Also: Google tests its new AI medical chatbot at Mayo Clinic
Ecommerce merchants can also tap ChatRhino to create a series of images, marketing posters, and product images from a single product image. The AI model can cut production costs by 90% for each visual asset, reducing the time it takes to complete a task from a week to half a day, JD.com says.
Interim legislation to guide general AI implementation
ChatRhino’s launch comes the same week that China introduced temporary regulations for joint asset management. Domestic AI service.
The Chinese government said that will take effect from August 15, the new laws are necessary to ensure the healthy development of technology and protect both national security and public interests.
Also: Innovative AI is coming to your work. Here are 4 reasons to get excited
in a joint declare promulgated by various agencies including the Cyberspace Administration of China (CAC) and the Ministry of Science and Technology, the government noted that although general AI has created economic development opportunities and new society, but it also brings challenges such as fake news and privacy and data safety risks.
The statement stated that the provisional law outlines many measures to facilitate the healthy development of technology, while protecting the national and public interests as well as the legal rights of citizens and enterprise.
For instance, innovative AI developers will have to ensure their pre-training and model optimization processes are carried out in compliance with the law. These include the use of data from legitimate sources subject to intellectual property rights. If personal data is used, the consent of the individual must be obtained or it must be done in accordance with applicable regulations.
Also: Leadership Alert: Dust will never settle and innovative AI can help
Measures must also be taken to improve the quality of the training data, including its accuracy, objectivity and diversity.
Under the interim law, general AI service providers are legally responsible for the information generated and for the security of that information. They will need to sign service level agreements with users of their services that clarify each party’s rights and obligations.
When illegal content is discovered, service providers must take various measures such as preventing transmission of such content and redressing its use in model training. Relevant authorities are also notified.
In addition, service providers will have to take necessary measures if users engage in illegal activities using generalized AI services. These include the functions of restricting, suspending or terminating services, maintaining relevant records and reporting to the relevant authority.
Service providers who violate the new law will face penalties set out in many relevant existing Chinese laws, including Network security Laws, Data Privacy Laws and Law on protection of personal information. In the absence of a provision for a violation, a warning will be issued with an order modification that must be made within a period of time. Failure to comply with such orders may result in service suspension.
Also: According to a new study, 6 harmful ways ChatGPT can be used by bad guys
China in April released a preview the draft lawsays that the development of synthetic AI technologies like ChatGPT can lead to abuse if left unchecked. A separate law that went into effect in January sets out the basis rules to prevent “deep synthesis” technology, including deepfakes and virtual reality, from abuse. Anyone using these services must properly label the images and not exploit the technology for activities that violate local regulations.
In May, the Chinese government announced plans to build AI industrial centers and technology platforms across the country to support research and development. To date, development plans have been launched for 18 national AI pilot regions and 32 innovation platforms, including in Beijing and Tianjin.
In addition to JD.com, local players like Tencent And alibaba also announced efforts to provide or integrate generalized AI into their products. Alibaba Cloud in April unveiled its big language AI platform, called Dong Nghi Thien Van, is currently available to customers in China for beta testing and as an API for developers. Chinese cloud provider also Introducing the partnership program to accelerate the development of AI applications for verticals, including finance and petrochemicals.
