SORA concept
SORA has three outstanding highlights, one is 60 seconds long**, SORA can maintain a high degree of fluency and stability between the subject and the background. The second is a single-** multi-angle lens, SORA realizes multi-angle lenses in one **, and the storyboard switching is logical and very smooth. The third is the ability to understand the real world, Sora does a very good job of handling details such as light and shadow reflection, movement mode, and lens movement, which greatly improves the sense of realism. Not only can it create realistic scenes according to text commands, generating up to 1 minute**, but also can realize multi-angle lens switching within one**, with good logic and smoothness.
Future Trends
According to incomplete statistics, more than a dozen A-share listed companies have previously deployed multi-modal large models or AI-related research and development.
Hikvision (002415) disclosed in response to investors' questions on April 20 last year that the company began to develop visual large models a few years ago, and has now reached the research and development stage of multimodal large models, including the fusion training and processing of multimodal signals such as vision, speech, and text.
In October 2023, Dahua Co., Ltd. (002236) released the "Xinghan Model", which integrates input such as point clouds, voice, and images to build a multi-modal fusion industry visual model.
Dahua Co., Ltd. introduced in the investor Q&A on January 15 that the company's multi-modal fusion industry visual model (Dahua Xinghan model) is based on the company's rich business experience in industries and subdivided scenarios in the past, facing the field of intelligent IoT with vision as the core, and built by fusing multimodal data such as images, point clouds, text, and voice, which greatly improves the visual analysis ability.
Wondershare Technology (300624) previously released the sound **multi** large model Wondershare "Tianmu". It is understood that Wondershare "Tianmu" is a multi-professional vertical model based on audio generative AI technology, which is composed of a large model, an audio model, a large model, and a language model, focusing on the creation scene of digital creative verticals. 
Focused lists.
1. Guangguang Media SZ300251
Now: 799, total market capitalization: 2344 billion.
Main business: film business.
Core competitiveness: 2017 annual film inventory of the company.
2. Shensi Electronics (SZ300479).
Now: 1528, total market capitalization: 301.1 billion.
Main business: intelligent identity authentication terminal and industry application software.
Core competitiveness: independent research and development and continuous technological innovation advantages.
3. Zhewen Film (SH601599).
Now: 312. Total market capitalization: 362.1 billion.
Main business: production and sales of various knitted woolen yarns and high-grade worsted woolen fabrics.
Core competitiveness: technology research and development advantages.
4. Xinhuanet (SH603888).
Now: 217 billion, total market capitalization: 1126.3 billion.
Main business: online advertising, information services, mobile Internet, network technology services and digital content.
Core competitiveness: brand advantage.
5. Wangda Software (SH603189).
Now: 991, total market capitalization: 267.1 billion.
Main business: high-tech technology platform, convergence ecosystem and AI big data platform.
Core competitiveness: deep cultivation in the field of technology for many years, all-round layout advantages of technology.