China is focusing on large language models (LLMs) in the field of artificial intelligence.
BlackdoFx | iStock | Getty Images
China’s attempts to dominate the AI world may be paying off, as industry insiders and technology analysts told CNBC that Chinese AI models are already very popular and are keeping pace with — and even surpassing — U.S. models in terms of performance.
Artificial intelligence has become the latest battleground between the United States and China, with both sides viewing it as a strategic technology. Washington continues to restrict China’s access to cutting-edge chips designed to help power artificial intelligence amid concerns that the technology could threaten US national security.
This has prompted China to take its own approach to enhancing the appeal and performance of its AI models, including relying on open source technology and developing its own ultra-fast software and chips.
China is creating popular LLMs
Like some of the leading American companies in the field, Chinese AI companies are developing so-called large language models, or LLMs, which are trained on huge amounts of data and support applications such as chatbots.
Unlike the OpenAI models that run the hugely popular ChatGPT, many of these Chinese companies do Develop open source or open weight LLMs Which developers can download and build on for free and without strict licensing requirements from the inventor.
At Hugging Face, a repository for LLMs, Chinese LLM degrees are the most downloaded, according to Tizhen Wang, a machine learning engineer at the company. Qwen, a family of AI models created by the Chinese e-commerce giant AlibabaHe said he is most popular on Hugging Face.
“Qwen is rapidly gaining popularity due to its outstanding performance against competitive benchmarks,” Wang told CNBC via email.
He added that Qwen has a “very convenient licensing model” which means companies can use it without needing “extensive legal reviews”.
Qwen comes in different sizes or parameters, as it is known in the world of LLMs. Large parameter models are more powerful but have higher computational costs, while smaller parameter models are cheaper to run.
“No matter which size you choose, Qwen is likely to be one of the best-performing models available right now,” Wang added.
The startup DeepSeek also recently made waves with a model called the DeepSeek-R1. DeepSeek said last month that its R1 model competes with OpenAI’s o1 model, a model designed to infer or solve more complex tasks.
These companies claim that their models can compete with other open source offerings such as deadLlama, as well as closed LLM courses like OpenAI, across different functions.
“In the past year, we have seen a rise in Chinese open source contributions to AI with really strong performance, low cost of service and high productivity,” Grace Esford, partner at Lux Capital, told CNBC via email.
China pushes open source global
Open source technology serves a number of purposes, including stimulating innovation as more developers gain access to it, as well as building a community around the product.
Chinese companies were not the only ones that launched open source master’s programs in management. Meta, the parent company of Facebook, as well as European startup Mistral, have open source versions of their AI models.
But with the tech industry caught in the geopolitical battle between Washington and Beijing, holding open source masters gives Chinese companies another advantage: enabling their models to be used globally.
“Chinese companies want to see their models used outside of China, so this is an ultimate way for companies to become global players in AI,” Paul Triolo, a partner at global consulting firm DGA Group, told CNBC via email.
While the focus is on AI models at the moment, there is also debate about what applications will be built on top of them – and who will dominate the global internet landscape from now on.
“If you assume that these basic frontier AI models are bets on the table, then it comes down to what you’re using these models for, like accelerating frontier science and engineering technology,” Lux Capital’s Isford said.
Today’s AI models have been compared to operating systems, e.g Microsoft windows, GoogleAndroid and appleiOS, with the ability to dominate the market, just as these companies do on mobile phones and computers.
If true, this makes the stakes for building a dominant MBA higher.
“They (Chinese companies) view the MBA as a hub for future technology ecosystems,” Shen Sun, senior lecturer in Chinese and East Asian business at King’s College London, told CNBC via email.
“Their future business models will depend on developers joining their ecosystems, developing new applications based on LLMs, and attracting users and data from which they can subsequently monetize through various means, including but beyond directing users to use their cloud services.” The sun added.
Restrictions on chips cast doubt on the future of artificial intelligence in China
AI models are trained on massive amounts of data, requiring massive amounts of computing power. currently, Nvidia He is the leading designer of the chips required for this, known as graphics processing units (GPUs).
Most leading AI companies train their systems on high-performance Nvidia chips, but not in China.
Over the past year or so, the United States has intensified export restrictions on advanced semiconductor and chip manufacturing equipment to China. I mean NvidiaThe company’s advanced chips cannot be exported to the country, and the company has been forced to create sanctions-compliant semiconductors to export them.
Despite these limitations, Chinese companies are still able to launch advanced AI models.
“Major Chinese technology platforms currently have sufficient access to computing power to continue improving models. This is because they have stored large numbers of Nvidia GPUs and are also leveraging domestic GPUs from Huawei and other companies,” DGA Group’s Triolo said.
Indeed, so were Chinese companies Strengthening efforts to create viable alternatives to Nvidia. Huawei has been one of the leading companies pursuing this goal in China, and companies love it Baidu Alibaba is also investing in semiconductor design.
“However, the gap in terms of edge computing will become larger over time, especially next year as Nvidia rolls out its own Blackwell-based systems that are limited to export to China,” Triolo said.
Lux Capital’s Isford noted that China is “systematically investing in and growing the entire domestic AI infrastructure outside of Nvidia using high-performance AI chips from companies like Baidu.”
“Whether or not Nvidia chips are banned in China will not prevent China from investing and building its own infrastructure to build and train AI models,” she added.
https://image.cnbcfm.com/api/v1/image/108075021-1734011164475-gettyimages-1456167226-gitech84.jpeg?v=1734011751&w=1920&h=1080
Source link