Large models to be key infrastructure in future
What will the final landscape look like after the fierce competition for developing large language models?
Huang Tiejun, head of the Beijing Academy of Artificial Intelligence, gives us a glimpse.
"In the long run, large models will become the infrastructure of the whole society, just like water, electricity and coal, and all parties entering this market will gradually develop in the direction of mutual cooperation, and finally form an ecosystem of systematic cooperation," Huang told China Daily in an exclusive interview.
A large model is in itself never a competition barrier, Huang said.
"Only through holding an open attitude during the dynamic competition can one find its own comparative advantage and survive."
The academy, also known as Zhiyuan Institute, is poised to be among the front-runners in the generative AI revolution, along with Open AI's ChatGPT and Google Inc's DeepMind, Microsoft Vice-Chairman and President Brad Smith has been quoted as saying.
Zhiyuan Institute was established in 2018 to develop LLMs, and is supported by the Ministry of Science and Technology, the Beijing municipal government, and leading companies and universities.
Huang, citing the example of power generation and electricity consumption, said that large models are also on their way to continuous iteration in the coming decades.
"Thus, there won't be a closed monopoly or so-called barrier in this field. The government is expected to play its overall planning role while letting the market be the major driving force. Cooperation will be the main way of development in the end," he said.
Huang's remarks echo what OpenAI CEO Sam Altman said at the annual conference of BAAI in Beijing in June.
Altman emphasized the importance of collaboration between US and Chinese researchers to mitigate the risks of AI systems, against a backdrop of escalating competition between Washington and Beijing in technology.
"China has some of the best AI talent in the world and, fundamentally, given the difficulties in solving alignment for advanced AI systems, this requires the best minds from around the world," Altman told participants via a video message.
As of May, China had developed at least 79 AI large language models, or rivals of ChatGPT, according to a report released by the Institute of Scientific and Technical Information of China.
(Editor:Wang Su)