Alibaba Unveils Powerful AI Model Tongyi Qianwen with Hundreds of Billions of Parameters
Alibaba, the Chinese tech giant, has made an exciting announcement regarding its AI model Tongyi Qianwen. The latest update brings in hundreds of billions of parameters, making it one of the most powerful AI models in the world.
During Alibaba’s annual conference in Hangzhou, the company also revealed eight industry-specific AI models. These models are designed for the entertainment, finance, healthcare, and legal sectors, offering dedicated tools for various purposes such as generating graphics, writing code, analyzing financial data, and conducting legal paper searches. These tools have already gone online following the announcement.
What sets Tongyi Qianwen apart is its impressive number of parameters. Alibaba’s cloud computing arm stated that the new Tongyi Qianwen 2.0 boasts “hundreds of billions of” parameters, which is a benchmark used to measure the power and capabilities of an AI model. Comparatively, OpenAI’s ChatGPT-3 from Microsoft had 175 billion parameters in 2020, while Meta Platform’s Llama 2 model had 70 billion parameters this year, as reported by Reuters.
According to Forbes, Alibaba’s Tongyi Qianwen is not just an ordinary chatbot. It serves as a question-and-answer chatbot and can also function as a coding co-pilot by generating and automatically error-checking code in multiple programming languages. Similar to its competitors like ChatGPT, Tongyi Qianwen is open-sourced, meaning that the source code is freely available to the public and can be updated by anyone. This openness allows for integration of enhancements and changes by developers worldwide. This is in contrast to other models like OpenAI’s GPT-4, which is not open-sourced.
Alibaba has actively collaborated with various organizations, including OPPO, Taobao, DingTalk, and Zhejiang University, to train their own large language models or create language model applications based on Tongyi Qianwen, as reported by Reuters. During the recent announcement, Alibaba chairman Joe Tsai revealed that approximately half of China’s large-language AI models currently run on Alibaba Cloud. Tsai also mentioned that Alibaba’s AI community ModelScope has grown to include 2,300 models and 2.7 million developers.
Alibaba’s CEO, Eddie Wu, has previously emphasized the company’s focus on becoming more customer-centric and AI-driven.
China’s tech sector is currently witnessing a “War of a Hundred Models,” a term coined by Tencent vice president Jiang Jie. This refers to the intense competition among Chinese companies to design and deploy generative AI technologies. The success of ChatGPT has triggered a flurry of product announcements from both startups and industry giants, leading to a race among Chinese tech giants like Alibaba and Huawei to showcase their AI offerings.
Tencent, Alibaba’s competitor, has also entered the race with its Hunyuan AI model, which boasts more than 100 billion parameters and surpasses OpenAI’s GPT-4 model in Chinese language processing capabilities.
Data from brokerage CLSA indicates that China now has at least 130 large language models, accounting for 40% of the global total, behind the United States with 50%.
In conclusion, Alibaba’s latest update to its AI model Tongyi Qianwen, with its vast number of parameters, solidifies its position as one of the most powerful AI models globally. The company’s focus on customer-centricity and AI-driven solutions aligns with the current trends in the tech industry. The “War of a Hundred Models” in China highlights the fierce competition among Chinese companies to lead in the AI sector and develop innovative generative AI technologies.
I have over 10 years of experience in the cryptocurrency industry and I have been on the list of the top authors on LinkedIn for the past 5 years. I have a wealth of knowledge to share with my readers, and my goal is to help them navigate the ever-changing world of cryptocurrencies.
