AI-Driven Digital Human Sector Heats Up, Experts Widely Optimistic
-
With the emergence of ChatGPT and breakthroughs in generative AI algorithms, AIGC (AI-Generated Content) has entered an explosive phase of application, bringing new possibilities to industries like artificial intelligence and the metaverse.
Recently, DeepBlue Technology officially announced the recruitment of nationwide partners for its 'Silicon-Based Brain SaaS Platform,' aiming to drive diversified enterprise applications of virtual digital humans through its generative AI-powered digital human creation platform.
IDC's report 'China's AI Digital Human Market Status and Opportunity Analysis' categorizes digital humans into five levels. Currently, digital humans have evolved from manual creation to AI modeling, possessing a basic human-like appearance but limited to simple interactions. At L4 and L5 levels, AI-driven digital humans can handle most scenario-based decisions and support real-time multimodal interactions. GPT, already impressive in text generation, enhances digital humans' 'IQ,' pushing them toward higher developmental stages.
NLP large models serve as the technical foundation for AI-driven virtual humans, enabling them to communicate fluently and reduce the cost of standardized content production. These models are trained for roles like smart customer service, hosts, and tour guides. In the long term, as capabilities in personalization and emotional understanding improve, digital humans could provide companionship for the elderly or serve as 'private tutors' for children.
As China's first digital human generation platform with AI interaction capabilities, DeepBlue's 'Silicon-Based Brain SaaS Platform' was launched in June. Built on DeepBlue's silicon-based knowledge model, it enables customized training for domain knowledge and personalized traits, offering zero-code, one-click generation, and low-cost features for full customization from appearance to knowledge systems. By importing product manuals, FAQs, and other documents, the platform automatically generates knowledge graphs, allowing digital humans to conduct intelligent one-on-one interactions, replacing rigid machine customer service. DeepBlue's digital humans are now available as SaaS, enabling businesses to create AI-driven digital humans at a low cost with minimal training.
Professor Sha Wenhao, Director of the Policy Research Office at Shanghai Jiading Metaverse Industry Promotion Association and member of the Shanghai AI Technology Association Expert Committee, explained that digital humans can be driven by human operators or AI. AI-driven digital humans, powered by machine learning and data training, represent a technological evolution under large models, offering advantages in intelligent interaction, personalized services, presentation, and commercial value. Whether providing tailored advice, answering questions, or guiding purchases, generative AI digital humans can generate appropriate content based on user needs.
Currently, digital humans are applied in enterprise scenarios like media broadcasting, financial services, live commerce, and virtual customer service, as well as consumer-facing roles like virtual idols and streamers. Their market value continues to rise as their applications prove increasingly valuable.
Li Dong, Managing Partner and President of Fortune Central Financial Holdings, noted that iMedia data shows China's virtual digital human industry reached RMB 107.47 billion in 2021 (core market: RMB 6.22 billion), with projections of RMB 640.27 billion (core market: RMB 48.06 billion) by 2025. As the industry grows rapidly, technological innovation becomes key to differentiation and competitiveness. AI digital humans, based on large-model generation technology, serve as a critical entry point for enterprise-level AI adoption, potentially reshaping human-machine collaboration and empowering business growth—a trend widely favored by investors.