On January 31, at the 2026 Xiong'an New Area "AI+" Innovation Ecosystem series event, the latest achievements and strategic plan for the "Jishu" data large model (LimiX), jointly developed by Tsinghua University and Wenzhun Intelligent (Xiong'an) Technology Co., Ltd., were unveiled. This model is the first and currently the only general-purpose large model in China designed for processing structured data. The rise of data large models signifies a crucial shift in the AI paradigm, with its technical essence being the construction of a 'general world model' capable of understanding the intrinsic causal mechanisms within data, thereby truly enabling 'one model to empower thousands of industries,' stated Academician Zhang Bo of the Chinese Academy of Sciences. The success of the "Jishu" data large model (LimiX) is achieving a leap from 'following' in language large models to 'leading' in data large models. Xiong'an is attempting to define a new paradigm for AI development for a city of the future by pioneering new tracks and building a new ecosystem. The core of this new paradigm lies in its precise targeting of the real bottlenecks in industrial development. Currently, the AI wave, represented by large language models (LLMs), is widely applied on the consumer end but faces fundamental limitations when integrating into the core of the real economy, such as industrial manufacturing and energy power. The "Jishu" data large model (LimiX) directly addresses this pain point; it can understand the internal causality of structured data like tables, time series, and graphs, constructing a 'general world model' for the data space. This represents a track that is deeply aligned with the transformation and upgrading needs of the real economy, offering greater foundational potential. The support for this new paradigm lies in building a self-driven, closed-loop, collaborative innovation ecosystem, rather than relying on isolated breakthroughs. Xiong'an's layout reflects systematic thinking, with the "Jishu" data large model (LimiX) as the technological spearhead, supported by a complete ecological chain integrating 'research, production, and training.' On the R&D front, the Xiong'an Artificial Intelligence Institute tackles cutting-edge technologies through the "1+6 Public R&D Platform" and promotes the commercialization of results. On the industry front, simultaneously held seminars and venture capital meetings aim to accelerate technology adoption with real-world scenarios and financial support. On the talent front, the newly launched "Xiong'an New Area Artificial Intelligence Training Ecosystem" is guided by demand across five major job categories, offering modular courses comprising over 4,000 class hours across 11 core directions, providing a continuous talent pipeline for the industry. This model, which coordinates frontier research, industrial application, capital support, and talent cultivation within a single framework, will generate sustainable, endogenous innovation momentum. The ultimate goal of this new paradigm is to explore systemic solutions for an 'AI-native city' and to export a replicable model for urban development. As a future city built from scratch, Xiong'an has the opportunity to elevate artificial intelligence from a single industry to the underlying logic reshaping urban governance, public services, and industrial organization. The series of events showcased precisely this kind of systematic practice: aggregating global resources, leveraging solid application scenarios, and nurturing industrial clusters, with the ultimate goal of forming a set of 'benchmarkable, governable, and iterable' standards for an 'AI+ city' system. From pioneering the new track of 'data intelligence' to building a new collaborative ecosystem of 'research, production, and training,' the Xiong'an New Area is using the future city as a vehicle to construct an 'AI-native city with global influence.'
Comments