A Chinese startup named “01.AI” has garnered the world’s attention by becoming a unicorn in less than eight months, a phenomenal thing to achieve in a short span of time.
Leveraging the artificial intelligence technology’s exponential growth in recent times, Lee Kai-Fu – who is a computer scientist and author of “AI Superpowers”, founded this 01.AI assembling experts of every field and from top firms. The firm founded in March 2023, and started operations in June, has thrived enough with tech enthusiasts to innovate an AI model within six months and getting valued at over $1 billion as a unicorn in eight months.
Lee Kai Fu, the 61-yr-old chief executive officer of venture firm Sinovation Ventures will also be the CEO of the new startup. 01.AI announced its open-sourced, foundational large language model (LLM) – Yi-34B, that is available in Chinese and English, SCMP reports.
Origin of 01.AI
The onset of generative AI brought into play by OpenAI’s ChatGPT in November, 2022 urged every other big-techs to delve into the field. Alphabet, Microsoft, Meta, Samsung, China’s Baidu rolled out their AI models for maintaining their competency in the evolving technology. Silicon Valley becomes filled with large language models and its subsequent AI models. The latest being Elon Musk’s Grok.
However, the Silicon Valley models aren’t available in China and this pose Chinese firms a great opportunity to develop their own models for China, and will also ease the abroad business with its Chinese customers.
Lee Kai-Fu saw this bloom of AI and leveraged the opportunity by gathering people to start an AI firm “01.AI”. The group includes not just AI specialists, but also experienced businesspeople who can help with everything from mergers and acquisitions to an initial public offering.
Ultimately, the team came up with the Yi-34B large language model. Large language models are deep-learning algorithms trained on large quantities of data to read, understand and generate humanlike text, images and codes.
Yi-34B LLM Model
The 34-B in Yi-34B name represents the 34 billion parameters used in training the model. Though the Yi-34B’s parameters is much lower than Meta Llama 2’s 70 billion and Falcon’s 180 billion parameters, the LLM could possibly outperform Llama 2 and other leading open-source models on the market, Hugging Face ranks.
Hugging Face runs leaderboards for the best-performing LLMs in various categories. It posted evaluations citing Yi-34B as the best in pre-trained base LLMs category.
“Llama 2 has been the gold standard and a big contribution to the open-source community,” Lee said in an interview. “We want to provide a superior alternative not just for China but for the global market.”
The startup focusses on to build a state-of-the-art proprietary model that serves as a foundation for a diverse range of commercial products. The LLM can also be run on computers that are not prohibitively expensive. 01.AI also works on 100-bn and a 6-bn model to appeal to a broader swathe of developers. The model is now available in GitHub and hugging face.
Man behind 01.AI – Lee Kai-Fu
Lee Kai-Fu had worked in Google, Microsoft and Apple before moving into venture capital. Being a CEO of Sinovation Ventures, he backs over 10 unicorns and venture-built seven companies and is one of the most well-connected investors and entrepreneurs in China.
Lee authored the “AI Superpowers” and “AI 2041: Ten Visions for our Future”. He had been involved in AI researches before venturing 01.AI. While Biden’s administration banned the sale of Nvidia’s most advanced AI semiconductors to Chinese customers, Lee Kai Fu foreseen the future and stockpiled the GPU chips, which would be sufficiently enough for the next 12-18 months for powering the startup.
01.AI is working on to release an app within this calendar year, said Lee. The ultimate goal, according to Lee, is to become an ecosystem where outside developers can build applications easily. “The duty is not just to push out good research models, but even more importantly to make application development easy so that there can be compelling applications,” he said. “At the end of the day, it is an ecosystem play.”
The firm got funded recently by Alibaba Group Holding’s cloud unit, valuing it over $1 billion. Though achieving unicorn status for a firm in less than eight months is sensational, sustainment of the model and firm will be quoted by time.
Kindly add ‘The Inner Detail’ to your Google News Feed by following us!