Chinese start-up Moonshot AI raises US$1 billion in funding round led by Alibaba and VC HongShan amid strong interest for OpenAI-type firms

Chinese start-up Moonshot AI raises US$1 billion in funding round led by Alibaba and VC HongShan amid strong interest for OpenAI-type firms

Chinese start-up Moonshot AI raises US$1 billion in funding round led by Alibaba and VC HongShan amid strong interest for OpenAI-type firms

The start-up in October launched smart chatbot Kimi Chat, built on its self-developed Moonshot large language model (LLM), which can process as many as 200,000 Chinese characters in a context window – referring to the swathe of text that an AI model can process during conversation with users.

A screenshot of smart chatbot Kimi Chat. Photo: Moonshot AI

Alibaba, HongShan and Moonshot AI did not immediately respond to requests for comment on Tuesday.

Moonshot AI’s new funding round marked the largest single financing raised by a Chinese AI start-up since ChatGPT was released in November 2022.
The latest investment recorded by Moonshot AI showed that interest in generative AI start-ups remains strong in mainland China, which led global investments into such firms in the first half of 2023.

China had the most number of generative AI start-ups that received funding at the time, according to a report by AI-focused research firm Zhidongxi. The mainland accounted for 22 of the 51 such start-ups around the world that raised about US$13.8 billion in investments during the period.

China has the most generative AI start-ups to receive funding in first half: report

Moonshot AI was founded in April 2023 by Yang Zhilin, a computer science graduate from Beijing’s prestigious Tsinghua University. He later received his doctoral degree in the US from Carnegie Mellon University in Pittsburgh, Pennsylvania.

In an industry event last November, Yang said that the ability to process long texts is critical to the future development of existing AI models.

An LLM’s capability partly hinges on its number of parameters, a measure of the sophistication of an AI model. While ChatGPT was trained on 175 billion parameters, the Moonshot LLM has more than 100 billion parameters, according to local media reports.

Chinese AI start-up claims to beat US rivals in processing long text

Also based in Beijing, Baichuan claimed at the time that its Baichuan2-192k LLM had surpassed Claude 2, the AI model of Amazon.com-backed Anthropic, as the world’s most advanced LLM in terms of the number of words that users could include in their chat queries. Claude 2’s context window was said to be capable of scrutinising around 75,000 English words, equivalent to hundreds of pages of documents or a thick book.

Mainland China had 254 LLMs as of November, according to a white paper published by the Beijing Municipal Science and Technology Commission.

The launch of various domestic LLMs, however, has been criticised by Baidu founder, chairman and chief executive Robin Li Yanhong as a “huge of waste of resources”. At an event in Shenzhen last November, Li said there were too many LLMs in China, “but too few AI-native applications based on those models”.

Meanwhile, Moonshot AI might somehow have to contend with another start-up named Moonshot AI that is based in New York City. This US firm, according to its website, provides an AI platform that helps e-commerce stores fully automate their online operations to increase conversions, sales and revenue.

Source link