Zhipu AI (智谱华章), a Chinese large language model (LLM) developer, announced on October 20 that the firm had received 2.5 billion yuan (US$340 million) in venture funds from a consortium of well-known investors year to date.
The backers included Social Security Fund ZGC Innovation Fund, Meituan, Ant Group, Alibaba, Tencent, Xiaomi, Kingsoft, Shunwei Capital, TAL, HongShan, Hillhouse Capital, Legend Capital and other existing shareholders.
The money raised will be used to facilitate R&D on its pre-trained LLM, GLM-130B, which reportedly was built upon hundreds of billions of parameters.
The funds will also go toward better supporting the industrial ecosystem and high-clip growth of partners, the Beijing-based AI company said in an official WeChat post.
Founded in 2019 and spun off from the Knowledge Engineering Group, an AI-focused lab, of Tsinghua University, Zhipu AI is chosen by The Information as one the five Chinese companies with the best hope of rivaling OpenAI, the AI powerhouse behind the phenomenal generative AI tool ChatGPT.
Zhang Peng, chief executive of Zhipu AI, believes that the next-gen technology is cognitive intelligence.
It is different from the previous generation of perceptive intelligence in that they are meant to tackle different issues.
In the age of cognitive intelligence, a larger amount of data and more human-like capapbilities are required under a multimodal environment.
Besides, what’s also critical is the artificial general intelligence catering to multiple tasks and scenarios, Zhang was once quoted as saying.
Scores of research organizations, startups and entrenched internet companies have leaped into the race to produce a Chinese equivalent of ChatGPT or Llama from Meta. But observers predict that a majority of them will fail eventually.