17 October 2026 Weibo

Virtual Idols and VTubers : Blogger “KizunaAICN” posted on Weibo on behalf of the virtual character “Kizuna AI China,” inviting netizens to do sit-ups together and engage with the hashtag “#VirtualCharactersPlayingWithWeibo#”; account “VTuber _ 2434 Qingbaozhan (VTuber_2434情报站)” reported that virtual streamer “Kuramoto Meru” (Sophia Valentine), under the Japanese virtual artist project “に じ さ ん じ (にじさんじ),” released a dance cover video “BoKuraNoyoru”; netizen “AI Emotion Notes” posted that the AI virtual girl group “BABYQUEEN” announced that its virtual member “Mity” is pregnant and will suspend group activities, sparking widespread online discussion; the creator subsequently explained that this was a fictional storyline borrowed from the common “pregnancy and maternity leave” narrative in the real idol industry, used to temporarily take the virtual character offline to reduce the team’s workload in maintaining multiple AI characters; netizen “Coconut Milk Frozen o” called on fans to continue voting for the virtual idol group “PLAVE” and reminded them that subsequent votes including MV view counts and music streaming volumes require continued effort; virtual character/IP “HAT HuoAoTian” posted on its official account that it is attempting new visual styles in the process of “collecting emotional keywords,” and has added multiple new skins and costumes, with related updates being reposted and discussed by fans.

Controversy : A netizen with the handle “What Should I Do on Rainy Days” published a post criticizing malicious comments online directed at virtual streamers, such as remarks suggesting they are “old” or “washed up.” The netizen argued that virtual streamers have difficulty defending themselves against such evaluations, and that repeatedly bringing up these jibes, even in jest, serves to reinforce stereotypical impressions of virtual streamers among audiences.

Enterprise : Guangzhou Xuni Dongli (广州虚拟动力) (real-time motion capture technology) has launched a comprehensive digital human training room solution for higher education institutions, claiming that the complete solution can bridge the entire chain from “software proficiency to production capability” in digital human development and application. The solution is equipped with VD - face Scanner (VD-face Scanner) multi-camera facial scanning devices that can perform high-precision 3D reconstruction of student faces and rapidly generate digital human models. Once models are imported into UE5, the system can automatically identify and complete topology dressing for body and hairstyle. The motion capture segment is equipped with proprietary full-body inertial capture devices DreamsCap X1, multi-node motion capture gloves MHandPro (covering 31 high-precision nodes), and AH - D (AH-D) dual-lens facial capture helmets (supporting ARKit 52 BS binding standards and MetaHuman real-time facial expression driving). Through the motion capture engine DreamsCap Studio, recording, preview, and data export are completed, supporting up to five simultaneous captures, with export capability to Maya, Blender and other software for secondary refinement. At the application level, the solution includes photorealistic digital human software system MHTP and digital human short-form video rendering platform RenderHare, applicable for virtual host incubation, virtual live teaching, and virtual IP interactive displays. The offering includes 11 chapters with 35 course hours of digital human modeling and motion capture resources, plus 6 chapters with 16 course hours of ultra-photorealistic digital human scanning and modeling curriculum, forming a complete teaching closed-loop from content production to operation. Lansu Keji （ larxu ） (蓝速科技（larxu）) mentioned in its A5 Xilie Zhuomian AI Shuangping Fanyiji (A5系列桌面AI双屏翻译机) introduction targeting business negotiation scenarios that the device supports multi-language real-time translation in addition to AI digital human customer service and intelligent meeting minutes functions. The company simultaneously released technical specifications regarding “digital human cabin interactive experience hardware selection,” noting that many poor digital human interactive experiences actually stem from insufficient front-end audio-video hardware specifications rather than software defects. It recommends prioritizing wide-angle high-precision facial detection cameras, array noise-cancellation microphones, and backlit recognition configurations with hardware ISP engines based on specific scenarios, and states that its entire series of digital human devices are equipped with industrial-grade interactive hardware to enhance interaction naturalness and fluidity. Another account published a headline article titled “Ranking Analysis of Leading Digital Human Companies: AI Large-Screen Mobile All-in-One Holographic Interactive Product Analysis,” providing a systematic review of major digital human manufacturers in the industry and their holographic interactive products.

Cultural : Internet user “Zhong Zhenglvxing (中正履行)” published a series of articles on research and study of the Doctrine of the Mean, marking the work as a research and study theoretical achievement co-created and refined through collaboration between themselves and the AI digital avatar of scholar Lu Qiyuan (卢麒元), intended for classical Chinese studies exchange and discussion.