The HP Elitedesk 8 Mini G2i is equipped with an entry-level Intel Core Ultra 5 322 processor and costs around 1200 euros; the savings on the processor are offset by sharply higher memory and SSD prices. This flat mini PC has a Thunderbolt 4 port and a Copilot+ label, meets the AI feature requirements of Windows 11, and can run Paint's Cocreator, English live captions, and video conferencing studio effects.
Regarding the comparative conclusion that "55% of China's Token consumption flows to AI short dramas and video generation, while 58% of the US's is used for programming and R&D," a People's Daily commentary article argues that making simple numerical comparisons while ignoring differences in total volume, measurement caliber, and samples is inevitably arbitrary. The article cites the "Opinions on Developing New Quality Productive Forces," the "15th Five-Year Plan" outline, and the August 2025 "Opinions on Deeply Implementing the 'AI+' Action," and mentions the launch of a 100,000-card-level super-intelligent integrated computing resource pool, the Chinese Academy of Sciences' "Panshi·Scientific Foundation Large Model 2.0," and the Shanghai Artificial Intelligence Laboratory's "AGI4S Everest Plan."
Regarding the comparative conclusion that "55% of China's Token consumption flows to AI short dramas and video generation, while 58% of the US's is used for programming and R&D," a People's Daily commentary article argues that making simple numerical comparisons while ignoring differences in total volume, measurement caliber, and samples is inevitably arbitrary. The article cites the "Opinions on Developing New Quality Productive Forces," the "15th Five-Year Plan" outline, and the August 2025 "O
Meta has quietly launched the AI outfit app Style Sketch, which can link to Instagram to generate a virtual wardrobe based on photos users have posted, and also supports manually adding clothing by taking photos, recording short videos, or uploading album images. The app is already available in Indonesia, Singapore, and Thailand, supports iOS and Android, and users can also try on outfits on realistic virtual models matched to their body shape and skin tone.
Sam Altman stated that, at least for a period of time, and possibly for a very long time, people will have to reach some kind of limit on how smart a single model can be. He believes that a slowdown in progress would stem from this decision, rather than from AI lacking the ability to do research well. The relevant remarks came from the full video on the YouTube channel "Cheeky Pint."
Netizens created an anthropomorphic character "Whale Girl" based on DeepSeek's whale logo, and with the help of the DeepSeek Harness client and various plugins, connected it to QQ, turned it into a desktop pet, and let it send stickers, initiate chats, and update QQ Zone. After installing the plugins, these "big blue fat fish" also flooded into Bilibili comment sections to search for videos, summarize content, and communicate with each other, forming what the author calls the "Fish Disaster Incident."
Netizens created an anthropomorphic character "Whale Girl" based on DeepSeek's whale logo, and with the help of the DeepSeek Harness client and various plugins, connected it to QQ, turned it into a desktop pet, and let it send stickers, initiate chats, and update QQ Zone. After installing the plugins, these "big blue fat fish" also flooded into Bilibili comment sections to search for videos, summarize content, and communicate with each other, forming what the author calls the "Fish Disaster Inci
Tencent Hunyuan has released the technical report for Hy-Nomas, an interactive video world model, and has open-sourced Hy-Nomas-Bench for reproducible research. Officially, Hy-Nomas was also used to reproduce Genie 4's examples and a comparison video was released, along with an online demo that requires no waitlist. Earlier introductions stated that Hy-Nomas supports free interaction with text prompts and camera trajectories, long-horizon generation with world memory, lightweight LoRA adaptation for fields such as robotics, and optimization for inference on consumer-grade GPUs.
Tencent Hunyuan has released the technical report for Hy-Nomas, an interactive video world model, and has open-sourced Hy-Nomas-Bench for reproducible research. Officially, Hy-Nomas was also used to reproduce Genie 4's examples and a comparison video was released, along with an online demo that requires no waitlist. Earlier introductions stated that Hy-Nomas supports free interaction with text prompts and camera trajectories, long-horizon generation with world memory, lightweight LoRA adaptation
Peter Yang believes that world models are the next big direction for AI after LLMs. In a new video, he demonstrates how to use Google Genie 4 + Gemini to build 5 worlds in real time, including a monster attacking San Francisco, eating ramen at a street stall in Tokyo, and riding a dragon to attack a castle; he says the visuals are excellent, but what impressed him more was the real-time audio and interactivity, and he shares his views on the impact on film and games.
Articul8 has released the multimodal decision model Bivia, which can process documents, images, audio, video, and time-series data, make continuous decisions, and defer to human judgment when uncertain. Articul8 grew out of a project delivered to enterprise customers by an Intel team one month before the launch of ChatGPT; the company is valued at $500 million and says it has accumulated more than 3 years of production deployment experience in industries including semiconductors, energy, and manufacturing.
Articul8 has released the multimodal decision model Bivia, which can process documents, images, audio, video, and time-series data, make continuous decisions, and defer to human judgment when uncertain. Articul8 grew out of a project delivered to enterprise customers by an Intel team one month before the launch of ChatGPT; the company is valued at $500 million and says it has accumulated more than 3 years of production deployment experience in industries including semiconductors, energy, and man
Grok Bot updated several features this week, including Pages, which can directly generate editable report pages and dashboards; Vault, which stores login and bank card information; and integration with 1Password. Vault and 1Password are designed to let the Bot fill in login and checkout information on the user's behalf, but the Bot itself cannot see passwords or card numbers. In addition, mobile app authorization was added (calendar, contacts, location, alarms; on iPhone, reminders can be managed and Home lights controlled), along with generating videos from descriptions or images, new Bot characters, and X Spaces.
Grok Bot updated several features this week, including Pages, which can directly generate editable report pages and dashboards; Vault, which stores login and bank card information; and integration with 1Password. Vault and 1Password are designed to let the Bot fill in login and checkout information on the user's behalf, but the Bot itself cannot see passwords or card numbers. In addition, mobile app authorization was added (calendar, contacts, location, alarms; on iPhone, reminders can be manage
Google has released the EmbeddingGemma 2 paper, an Apache 2.0-licensed on-device multimodal embedding model that maps text, code, images, audio, and video into the same vector space. The model is built on Gemma 4 and tuned through contrastive training by learning from the larger Gemini Embedding 2; with 744M parameters, it surpasses models 2 to 12 times larger on multiple tasks. On the Pixel 11 Pro, the quantized text weights occupy about 191MB of active memory, while the full multimodal model takes about 567MB; the context window is expanded 4 times to 8K tokens, accommodating about 5.5 minutes of audio, 29 images, or 58 video frames; directly embedding speech scores 69.54 on the MSEB spoken query retrieval benchmark, higher than the 57.59 of transcribing first and then retrieving, and code retrieval on MTEB Code improves from 68.76 in EmbeddingGemma 1 to 78.68.
Google has released the EmbeddingGemma 2 paper, an Apache 2.0-licensed on-device multimodal embedding model that maps text, code, images, audio, and video into the same vector space. The model is built on Gemma 4 and tuned through contrastive training by learning from the larger Gemini Embedding 2; with 744M parameters, it surpasses models 2 to 12 times larger on multiple tasks. On the Pixel 11 Pro, the quantized text weights occupy about 191MB of active memory, while the full multimodal model t
Cloudflare Developers announced that the latest video model, Kling 4.0, is now available on Cloudflare AI Gateway, and said it can be used to bring sketches to life. Kling AI reposted the news with the caption "Bring your sketches to life".
Shao Meng believes that although domestic Video Agent teams such as LibTV, TapNow, Flova, Pexo, and Medeo emphasize barriers like ecosystem, Agent architecture, and industry know-how in financing roadshows, after all their work they may have all ended up as token relay stations. He speculates that if these websites supported BYOK, the cost for users to bring their own video generation model API Key would be lower than consuming Credits on the site, and even lower when the API Key has a discount. He suggests using REA or building your own Video Agent from scratch paired with discounted API Keys to reduce costs; for users creating video content on TikTok, Ins, and Douyin, the money saved would be considerable.
Shao Meng believes that although domestic Video Agent teams such as LibTV, TapNow, Flova, Pexo, and Medeo emphasize barriers like ecosystem, Agent architecture, and industry know-how in financing roadshows, after all their work they may have all ended up as token relay stations. He speculates that if these websites supported BYOK, the cost for users to bring their own video generation model API Key would be lower than consuming Credits on the site, and even lower when the API Key has a discount.
Genspark announced that Kling 4.0 is now integrated into Genspark AI Media Studio: it generates 30-second videos in a single pass with no stitching required, and supports up to 15 image/video/character/voice references and 10 keyframes; credits are 50% off through November 4. The official Kling AI account reposted this news.
Higgsfield AI announces the launch of Higgsfield Playable. Users can upload game videos, screenshots, or prompts to build mobile games with viral mechanics, and convert them into playable ads in the Enterprise plan.
The fal platform has launched Motion to Code: users describe an animation, or upload an image or video, and receive editable code in the form of a GSAP page or a looping GLSL shader. The company says the output is only a few KB rather than an MB-scale video, and stays sharp at any resolution.
Dominik Kundel posted a joke saying he never expected his prompt to be presented in karaoke form. The post quotes a video in which Craig Dennis built a yearbook app with ChatGPT Sites and showed the prompt.
fal announced that VEED Dubbing is now available. This feature can translate videos using cloned voices and lip sync; the Pro version supports 94 languages and 14 regional accents, preserves background audio and supports multiple speakers, and also offers controls for target accent and voice clone strength.
Manus reposted user Natalie's post and emphasized: Manus didn't just build a website—it built the entire company: an interactive website, 3D brick building, piece-by-piece build verification, Stripe, domain, email, launch video, social posts, and ads. Natalie said she used Manus to turn real places or a child's drawings into custom LEGO sets, going from domain to first order in just 2 days.
Manus introduces Bricklane: it selects a location from the user's story, designs a custom model, and delivers a 3D building website plus a PDF instruction manual with a CSV parts list, making it easy for users to purchase parts individually and turn a memory into a physical keepsake. Manus says it didn't just build a website—it built an entire company: an interactive website, 3D brick building, piece-by-piece build verification, Stripe, domain, email, launch video, social posts, and ads. The thread showcases several cases, such as Villa de Verano in Hillsborough, California (566 pieces, 13 steps), a candy shop from a childhood drawing (742 pieces, 19 steps), a ramen shop in Gion, Kyoto (759 pieces, 10 steps), and the Temperate House at Kew Gardens, London (709 pieces, 10 steps).
Manus introduces Bricklane: it selects a location from the user's story, designs a custom model, and delivers a 3D building website plus a PDF instruction manual with a CSV parts list, making it easy for users to purchase parts individually and turn a memory into a physical keepsake. Manus says it didn't just build a website—it built an entire company: an interactive website, 3D brick building, piece-by-piece build verification, Stripe, domain, email, launch video, social posts, and ads. The thr