-
Kai-Fu Lee: Zero-One Everything is Building Windows for the AI 2.0 Era Based on DeepSeek
March 30 afternoon news, in the 2025 Zhongguancun Forum annual meeting, zero one million things CEO, innovation workshop chairman Li Kaifu said, "the reasoning cost of large models is rapidly declining at a rate of ten times lower per year, which provides very important conditions for the outbreak of AI-First applications." "Models whose performance was not good enough two years ago are now good enough; models whose inference cost was too expensive two years ago are now 'cabbage price'." Li Kaifu believes that "AI-First applications will soon blow out, and 2025 will be the year when AI-First applications...- 600
-
Kai-Fu Lee's AI company Zero One Everything exposed to multiple changes: plans to split digital people business, shut down some projects
February 21 news, according to "intelligent emergence" yesterday cited "a number of independent sources" news, Lee Kai-Fu founded the AI unicorn zero one million things, recently internal changes occurred in a number of places: plans to split the digital human business, by the zero one million things digital human business R & D head Yang Changpeng led the team. Blue Yuchuan, head of the AI video mixing and editing project at Zero One Thousand Things, has a new project, SparkView, oriented as an AI video editing tool. To the above information, zero one million things replied to the media, said: zero one million things last year there is more than one hundred million income, this year there will be rapid growth. Zero One Thousand Things is not only continuously online more...- 1.8k
-
Kai-Fu Lee: The current economics of AI are 'extremely unhealthy,' except for NVIDIA
October 25, according to "AI Robot Tea House" reported on October 22, Innovation Works Chairman and CEO Kai-Fu Lee recently expressed his views on the current AI situation in a private forum: the current AI ecosystem may be gradually moving towards "unsustainable! In his opinion, the current AI ecosystem may be gradually becoming "unsustainable". In his opinion, at this stage, except for graphics card giant NVIDIA, which is reaping a full harvest financially, other players in the AI field, including application developers and enterprises, are struggling to make a meager profit, and the entire AI ecosystem shows a serious "imbalance". Li Kaifu pointed out that, in general, application...- 3.7k
-
Kai-Fu Lee: Zero-One Everything hasn't given up on pre-trained models, and new ones are coming soon
According to three words of science and technology reports, Li Kaifu on October 11, said in an article, in response to several online media said that the domestic big model company has a few give up the pre-training, and also said that zero one million things is one of them. Li Kaifu responded that: "One million things have been doing pre-training, last year and this year released Yi-34B, Yi-Large, Yi-Coder, Yi-VL, YiVision are released when the world's first echelon, the domestic leading pre-training models. And our new pre-training model will be released soon, which will also be the first tier in the world and the leading one in China. By then, the rumors will be dispelled. Build...- 5.1k
-
AI unicorn Zero One Everything founded by Kai-Fu Lee completes hundreds of millions of dollars in financing
According to the news from "Intelligence Emergence", the AI big model unicorn company Zero One Everything, founded by Kai-Fu Lee, has completed a new round of financing of hundreds of millions of dollars, with a valuation exceeding the 20 billion yuan mark. This round of financing was participated by many institutions such as international strategic investors and Southeast Asian consortiums, but Zero One Everything has not yet responded to this. Zero One Everything, founded in May 2023, has a core team from well-known companies such as Google and Microsoft. Today, it is known as the "Six Little Tigers of Big Models" together with Zhipu AI and Baichuan Intelligence, and is rising at an astonishing speed. In the field of AI big models, a valuation of 20 billion yuan has become a threshold. Internet strategic investors and state-owned platforms have entered...- 2.6k
-
OpenAI is expected to become a trillion-dollar company in two to three years
Chinese investor and serial entrepreneur Kai-Fu Lee recently said that he is bullish on OpenAI becoming a trillion-dollar company in two to three years. "OpenAI is likely to become a trillion-dollar company in the near future (two to three years)," Kai-Fu Lee said in a recent interview with Fortune magazine. "I am very optimistic about the future of OpenAI. They have indeed done an admirable and incredible job. Even today, GPT-4 is still the gold standard. You see, both Gemini Ultra and Claude3 claim to be...- 3.3k
-
Zero One Everything launches a one-stop AI work platform "Wanzhi", which supports Q&A, document interpretation, and PPT making
Kai-Fu Lee's AI company Zero One Everything announced on Tuesday the launch of a one-stop AI work platform "Wanzhi", which can be used to make meeting minutes, weekly reports, interpret financial reports or papers, make PPTs, etc. It can also be used as a "writing assistant". The general features of "Wanzhi" are as follows: According to the official introduction, "Wanzhi" can output higher-quality content in a variety of forms, including presenting answers in a variety of forms such as tables and mind maps. It supports real-time access and integration of Internet information to provide users with the latest data and insights. "Wanzhi" claims to be able to complete the reading of ultra-long texts of hundreds of thousands of words in 10 seconds...- 5.9k
-
Kai-Fu Lee's AI company Zero One Everything announced the open source Yi-9B model, claiming to have the strongest mathematical capabilities in the same series of codes
The official WeChat account of "Zero One Thousand Things 01AI" announced the open source Yi-9B model tonight. The official called it the "Science Champion" in the Yi series of models. Yi-9B is currently the model with the strongest code and mathematical capabilities in the Yi series of models, with an actual parameter of 8.8B and a default context length of 4K tokens. This model is based on Yi-6B (trained with 3.1T tokens) and uses 0.8T tokens for further training. The data is as of June 2023. According to the introduction,…- 3.9k