BN
|
TechAI Desk•2 views

Alibaba Leads $290M Investment for General World AI Model

Alibaba Cloud has spearheaded a $290 million investment into ShengShu, a startup known for its AI video generation tool, Vidu. This move signals a major industry pivot in AI development, moving beyond the limitations of traditional Large Language Models (LLMs) which are primarily text-based. The core objective is to build a 'general world model' capable of replicating the real world's complexity. This new model will be trained on multimodal data, including video and physical scenarios. Ultimately, ShengShu aims to use this technology to bridge the gap between the digital realm (like gaming and AI video) and the physical world (such as autonomous vehicles and robotics).

Ad slot
Alibaba Leads $290M Investment for General World AI Model

Alibaba Cloud has led a significant investment into ShengShu, a startup developing advanced AI video tools, signaling a major industry shift away from traditional text-based models toward comprehensive 'world models.'

The Shift from LLMs to World Models

The artificial intelligence sector is undergoing a rapid evolution, with developers recognizing the inherent limitations of Large Language Models (LLMs). These traditional models, such as those trained primarily on text, struggle to fully replicate the complexity of the real world.

Instead, the industry focus is shifting toward 'world models'—AI systems built on diverse, multimodal data that includes video, physical scenarios, and real-life interactions.

Alibaba's Strategic Investment

To capitalize on this trend, Alibaba Cloud announced it led a 2 billion yuan ($290 million) investment in ShengShu, the company behind the popular AI video generation tool, Vidu. The funding round also saw participation from TAL Education and Baidu Ventures.

Ad slot

This investment follows a previous funding round where ShengShu raised 600 million yuan from Qiming Venture Partners and other backers.

Building a General World Model

ShengShu stated that the capital infusion will support the development of a 'general world model.' This advanced AI aims to bridge two previously separate domains, connecting the digital and physical worlds.

Key areas the model intends to connect include:

  • The Digital World: AI-generated video and gaming environments.
  • The Physical World: Autonomous driving systems and robotics.

According to ShengShu, a general world model built on multimodal data—such as vision, audio, and touch—is better equipped to capture the natural mechanics of the physical world than current LLMs.

Ad slot