Table of Contents
Generative AI is a new frontier of artificial intelligence that enables machines to create content and data, such as images, text, speech, and music. This technology has the potential to improve enterprise productivity and deliver an immersive experience to users across various domains.
China’s tech giants are racing to develop their own generative artificial intelligence capabilities to compete with US companies, launching their own large language models (LLM) to support a wide array of uses, ranging from weather forecasts and financial analysis to 3D e-commerce.
Generative AI Growth in China
According to a report by global research firm McKinsey, China has a tremendous opportunity in the coming decade for generative AI growth in new sectors, including areas where innovation and R&D spending have traditionally lagged behind global counterparts, such as automotive, transport, manufacturing, healthcare, and life sciences.
The research firm said it identified clusters of use cases in these sectors that would allow generative artificial intelligence to create upwards of US$600 billion in economic value annually.
One of the key players in China’s generative artificial intelligence development is Huawei, the leading telecom equipment and smartphone maker. Speaking at the Huawei Connect 2023 event held in September, Sabrina Meng, deputy chairwoman of the firm, said Huawei continued to invest in core AI technologies and the development of a robust computing foundation for China and the world to support a wide range of AI models and applications for all industries.
She said Huawei’s “All Intelligence” strategy is designed to help all industries make the most of new strategic opportunities presented by generative AI.
The strategy provides a massive amount of computing power that is required to train foundation models for different industries.
Intelligent Economy Powered by Generative AI
According to Huawei, generative AI is a key driver of the intelligent economy, which generates value through the seamless integration of next-generation technologies including 5G+, AI, and the Internet of Things (IoT) to progress single-point intelligence systems working in isolation towards multiple intelligence systems that operate in tandem.
This effort has great potential to create new and innovative solutions to enhance productivity gains, social well-being, and benefits to the environment, said the company.
The intelligent economy is expected to be worth $18.8 trillion by 2030, said Huawei.
One of the examples of generative artificial intelligence applications in the intelligent economy is Huawei’s Pangu Weather Model, an AI model used for forecasting the weather worldwide. Thailand’s Meteorological Department plans to use this model to provide forecasts in Thailand.
The company claims the model is 10,000 times quicker at predicting the weather than other methods, producing a result in a matter of seconds.
Pangu models have been used in vehicles, for medicine, mining, and R&D. They have been used by governments and by virtual humans, said Huawei.
For example, the Pangu Vehicle Model automatically generates samples to create complex driving scenarios. Using this model, the amount of time an autonomous driving system requires to learn a new corner case has been reduced from more than two weeks to less than two days.
Generative AI Expansion Overseas
Another major player in China’s generative AI development is Alibaba Cloud, the cloud computing arm of e-commerce giant Alibaba Group. Selina Yuan, president of international business at Alibaba Cloud Intelligence Group, said the company’s generative AI solutions are designed to assist customers in creating their own generative artificial intelligence applications in a more efficient, security-focused, and cost-effective manner.
Alibaba Cloud introduced the PAI-Lingjun Intelligent Computing Service for high-performance computing tasks, such as foundation model training and inference.
The platform, already in use in China, will be available in Singapore from early next year, followed by more general availability across Asia throughout 2024, she said.
To simplify the building of an AI-based search service, Alibaba Cloud unveiled OpenSearch LLM-Based Conversational Search, its proprietary large-scale distributed search engine that enables intelligent search services in e-commerce, multimedia, social media, and big data queries in enterprises.
Alibaba Cloud recently showcased an array of industry-specific generative AI models at the Apsara Conference, its annual flagship tech event.
These advancements are built upon Tongyi Qianwen, the company’s proprietary foundation model, and are designed to streamline business operations and enhance user experiences.
Generative AI Innovations for Digital Content Creation
The cloud pioneer also unveiled a multitude of generative AI innovations, including a digital avatar creation tool and AI text-to-image tools, aimed at simplifying the process of digital content creation for businesses across sectors.
For example, its character creation and AI chat model (Tongyi Xingchen) facilitates engaging, human-like interactions with virtual characters.
The model enables users to create their own characters and interact with them for personality-driven companionship, emotional support, and entertainment.
Alibaba Pictures Group and T&B Media Global forged a multi-tiered strategic partnership in October to develop immersive entertainment technologies to bring new experiences to audiences worldwide.
Jwanwat Ahriyavraromp, chief executive of T&B Media Global, said the strategic cooperation focuses on three domains, including generative AI and virtual human development.
The partners plan to work together to create virtual humans for use in film and TV productions, as well as using generative AI to generate content.
The goal is to be at the forefront of generative AI development and virtual human creation for the entertainment industry, said T&B Media Global.
Generative AI Competition with US Companies
Chinese tech company Tencent recently launched its proprietary foundation generative AI model dubbed Hunyuan as it joined the generative AI race, bidding to automate human work and increase human productivity.
Tencent also offers the LLM Hunyuan for enterprise applications.
Enterprises in China can access Hunyuan via Tencent’s public cloud platform and fine-tune it to their specific needs.
The platform features strong Chinese language processing abilities, and advanced logical reasoning, and comes with reliable task execution abilities, according to the company.
Tencent’s foundation model supports a wide array of functions spanning the creation of images, copywriting, text recognition, and customer service.
These should be instrumental in key industries such as finance, public services, social media, e-commerce, transport, and gaming, said the company.
Tencent recently held its Global Digital Ecosystem Summit in China to showcase the company’s latest advanced technology, including a prototype of a 3D camera that uses generative AI to enable sellers to live-stream products that potential buyers will see in 3D without wearing 3D eyeglasses.
China’s tech sector is racing to gain ground in the development of generative AI technology, but Google DeepMind co-founder Mustafa Suleyman predicts the US will still be in the lead in a decade.
“I’m pretty sure of that. The fundamentals in the US innovation ecosystem are unquestionably the best in the world,” said Mr Suleyman, whose latest venture is Inflection AI, a startup valued at $4 billion.
“Without doubt, Silicon Valley is leading the pack, and I think we’ll continue to. I think we worry too much that we’re going to be dislodged from our podium position.”
US officials are particularly concerned about China’s ability to leverage generative AI for military purposes.
In August, the Biden administration imposed restrictions on US investments in Chinese semiconductor, quantum computing and generative AI firms, citing national security concerns. Last month, the US announced additional measures to curb China’s access to advanced semiconductors capable of training generative AI algorithms.
Chinese startup 01.AI, which was founded by computer scientist Kai-Fu Lee, developed an open-source LLM that outperformed some of Silicon Valley’s leading systems on certain metrics.
Baidu Inc. also recently claimed its LLM is as good as OpenAI’s GPT-4.
While Mr Suleyman is convinced the US isn’t at risk of losing its lead, he refuses to write off the competition.
“I think it’s better to take an approach of humility, acknowledging China is going to be a really significant player in this field over the next few decades,” he said.
