The Beijing startup founded by tech pioneer Kai-Fu Lee is unveiling its first consumer application of artificial intelligence, aiming to help China capitalize on the promising technology.
Lee’s company 01.AI is launching a free productivity assistant called Wanzhi, the latest in a series of AI products it is developing. Similar to Office 365 Copilot from Microsoft Corp. It helps users create spreadsheets, documents, and slide presentations faster – although it is primarily tailored to the Chinese market. It can interpret financial reports, take meeting minutes and quickly read books, up to the 600,000-word biography of Elon Musk for a quick overview. The app works in Chinese and English.
In an interview with Bloomberg, Lee said China needs its own ChatGPT — OpenAI’s chatbot, which was released in 2022 and banned in the country — to accelerate interest, adoption and investment.
“For Americans, this moment happened 17 months ago,” Lee said on a Zoom call from Beijing. “China’s users didn’t have a ChatGPT moment. So far, none of the Chinese chatbots or tools have been good enough.”
While U.S. firms like OpenAI, Meta Platforms Inc. and Alphabet Inc. have taken the lead in generative AI, Chinese players are pushing hard to catch up. In addition to 01.AI, technology companies such as Baidu Inc. and TikTok parent company ByteDance Ltd. are investing. money into developing their own AI models and chatbot services. Beijing has also provided financial and political support. Beijing blocks foreign AI models in part because of its strict censorship regime, but the so-called Great Firewall also ensures that domestic players have a huge local market without global competition.
The Taiwan-born 62-year-old, who worked for Apple Inc. and Google before starting his own venture capital firm more than a decade ago, became CEO at 01.AI last year. The startup reached a valuation of $1 billion Unicorn statuswithin eight months based on an open source AI model that outperformed Silicon Valley competitors in several key ways.
In addition to Wanzhi, the company is also introducing a larger, proprietary Large Language model – the technology that underlies AI chatbots – called Yi-Large, aimed at enterprise users.
Software developers can use the Yi-Large model at competitive prices. Lee says the model’s application programming interface (API) will cost $2.50 for 1 million input tokens and $12 for 1 million output tokens – about 1 million tokens allow a developer to make about 250 requests and send it here. That’s far less than OpenAI’s GPT-4 Turbo, he said.
Like many Chinese companies, 01.AI stockpiled Nvidia Corp.’s “graphics processor” semiconductors. came when it emerged that the US government was planning to ban the export of high-end chips like the H100s, which are used to train leading AI services. Alibaba Group Holding Ltd., the Chinese e-commerce giant that invested in Lee’s company, supplied an additional supply of H100 processors, and 01.AI supplemented its needs with slightly slower and less powerful Nvidia H800 processors.
“Our models were trained on H100 processors that were legally imported into China,” Lee said. “Necessity is the mother of invention; we squeeze everything out of the available computing power.”
Unlike many global AI startups, 01.AI is nearing profitability, according to Lee. After training the models on Chinese and universal datasets, Lee is rolling out the models and apps globally and acquiring both domestic and foreign customers to increase sales next year.
After a month and a half of testing with users, Lee’s company is launching a version of Wanzhi for PC browsers with more comprehensive features, as well as one for mobile phones that can be accessed via the messaging service WeChat. He appears in video tutorials on Douyin, the Chinese version of TikTok, to coach potential users.
“The year 2024 will be explosive for generative AI applications in China,” Lee said.
Lee said his startup is closing the second tranche of a $250 million pre-Series A round in a few weeks and will begin seeking investors for its Series A round by the end of the year. The company has also optimized its hardware and software processes to maximize efficiency and keep costs down.
“When GPT-5 comes, we will be one step behind,” Lee said, referring to OpenAI’s rumored next-generation AI model. But 01.AI is focused on making AI affordable rather than building huge, more expensive models. “You can build a huge, amazing spaceship, but can it take you from Sacramento to San Francisco?”