Written by 
AI Technology Review has exclusively learned that ByteDance is preparing to establish an AI research center in Europe.
Informed sources revealed that ByteDance has already begun searching for top talents in the fields of LLM and AI in Europe, actively recruiting personnel.
Looking at the level of AI technology across various European countries, Switzerland, the United Kingdom, and France are relatively advanced.
Among them, both Switzerland and the UK have many renowned institutions. The strong technical capabilities and talent density of ETH Zurich and EPFL in Lausanne make Switzerland a top-notch robotics innovation center worldwide. Moreover, Switzerland is where Google has established the most offices overseas apart from India, with over 5,000 employees covering various businesses such as Google Assistant, YouTube, Cloud, Commerce Research Geo Core Ads etc.
The UK and France have been fiercely competing for leadership in establishing an AI hub in Europe since last year. The former hopes to position the AI technology industry as a core economic development sector while the latter is ‘crazily’ investing in training and researching artificial intelligence.
In fact, as early as the beginning of September this year, there were reports in the media that ByteDance’s board of directors had added a new European director, namely French tycoon and founder and chairman of French Internet service provider and mobile operator Iliad Group Xavier Niel, indicating an obvious intention to expand into the European market.
Not limited to Europe, in June this year there was news that ByteDance plans to invest around 10 billion ringgit (2.13 billion US dollars) to establish an AI center in Malaysia. In addition, ByteDance also intends to invest an additional 1.5 billion ringgit to expand its data center facilities in Malaysia.
SEE ALSO: ByteDance Releases ‘Doubao MarsCode’ Tool with Code Completion Function
Now, as global large-scale narrative enters the second half, ByteDance is also accelerating its layout.
Actually, from an overall perspective, in the first half of the year, ByteDance was not among the earliest to focus on large models, but currently they are making frequent moves.
In January 2023, ByteDance started to build a large model team.
In August, ByteDance launched its self-developed underlying large model “Yunque,” followed by the official release of AI dialogue product “Doubao.”
In November, ByteDance established the Flow department focusing on research and development at the AI application level. The project is led by Vice President of Technology Hong Dingkun and Zhu Wenjia from the ByteDance Large Model Team.
By the end of December, another product called Coze was launched overseas besides Doubao. Its core service is to provide users with a platform for quickly developing AI applications that support generating AI bots without code within 30 seconds. It also integrates a plugin toolset to meet personalized user needs.
And since the beginning of this year, the pace of progress and talent recruitment in large-scale related work at ByteDance has clearly accelerated.
At the beginning of the year, Coze’s corresponding domestic version product ‘Kouzi’ was officially launched. The overall product form and functional positioning of both overseas and domestic versions are basically similar, allowing for the rapid creation, debugging, and optimization of AI chatbot applications.
In September, ByteDance officially released two large models: Doubao Video Generation – PixelDance and Doubao Video Generation – Seaweed.
Earlier this month, ByteDance established a joint research center for scalable large models intelligent technology (SIA Lab) with Tsinghua AIR to promote industry-academia-research cooperation in large models.
A few days ago, there was news that ByteDance has added another AI expert to its team: Zhou Chang, technical director of Alibaba‘s Thousand Questions Large Model.
Zhou Chang graduated from Fudan University with a bachelor’s degree in Computer Science and Technology in 2012. In 2017 he obtained his Ph.D. in Computer Software and Theory from Peking University before joining Alibaba through the ‘Alibaba Stars’ program. During his tenure at Alibaba, Zhou Chang led a team to design and implement M6, an ultra-large-scale multimodal pre-training model with up to 100 billion parameters – making it the largest model ever created in multimodal pre-training field history.
Now ByteDance is setting its sights on the European market once again as AI has become a crucial part of ByteDance’s global strategy.