NAll news

All news

Friday1456 items

  1. t3n
    AI score61

    Anthropic Becomes SpaceX's Largest Single Customer, Contributing About One-Third of Revenue

    According to Bloomberg, citing anonymous sources, Anthropic has become SpaceX's largest single customer, contributing about one-third of revenue and replacing the U.S. government in its former position. This relationship stems from a multibillion-dollar partnership Anthropic reached with SpaceX this spring, under which it obtained most of the AI compute capacity originally belonging to xAI. The report also mentions that SpaceX plans to move AI data centers to space on a large scale, but the technical feasibility remains disputed.

    #Industry news#English text is being prepared.#Anthropic

  2. cnBeta
    AI score58

    Wikimedia proposes a compromise position in AI copyright case: training may constitute fair use, but reproducing original text should require authorization

    Wikimedia has proposed a compromise position in the AI copyright lawsuit, arguing that using copyrighted works without permission for large language model pre-training may qualify as fair use if the model's output differs from the original text, but that when chatbots directly output article content and replace readers' access to the original works, AI companies should comply with licensing requirements. Wikimedia points out that models reproduce Wikipedia text published under open licenses but do not necessarily comply with conditions such as attribution and share-alike, and it worries that if courts broadly rule that AI training is entirely unrestricted, this could open the door to ignoring licenses. The backdrop to this position is that The New York Times sued OpenAI and Microsoft in 2023 for scraping large numbers of news articles without permission to train AI systems, with the two companies arguing that such training constitutes fair use.

    Same story · 2 reports

    #English text is being prepared.#OpenAI#Microsoft

  3. cnBeta
    AI score68

    Anthropic plans to raise up to $100 billion in IPO, scale may exceed SpaceX record

    Reuters reports that Anthropic, the company behind Claude, plans an IPO with a fundraising cap expected to reach $100 billion, benchmarked against the $86 billion record set by SpaceX in June. The article notes that Meta has just completed a $38 billion offering of common and preferred stock and plans to raise another $60 billion through targeted additional issuance starting next year, showing that equity investors are taking over from the credit market in funding AI capital expenditures. But the article also cautions that blue-chip secondary offerings differ from IPOs, and that the U.S. 10-year Treasury yield has risen from 4.5% to 5%, while most U.S. IPOs this year have averaged gains of only 2%.

    #Industry news#OpenAI#Anthropic

  4. cnBeta
    AI score41

    AI models "conquer" Double 11: technology is not guilty, fraud is

    AI model images cost as little as 6 yuan each, while a real-life model, venue, and shooting package costs at least two to three hundred yuan; this cost advantage has driven the rapid expansion of the AI model industry chain during Double 11. It is not in itself illegal for merchants to use AI to generate virtual models to display products, but if the rendered images differ greatly from the actual product, the AI labeling is deliberately removed, and consumers are misled, this may be deemed consumer fraud, with "refund plus triple compensation" liability under Article 55 of the Consumer Rights Protection Law. The Measures for the Labeling of AI-Generated Synthetic Content require that AI-generated content carry clear and perceptible labels, and platforms also bear verification obligations.

    #English text is being prepared.#Image generation

  5. LeitechResearch
    AI score69

    AgiBot, together with Tsinghua University and Infinigence AI, releases the CEL collective experience learning paradigm

    AgiBot, together with Tsinghua University and Infinigence AI, has released the collective experience learning paradigm CEL, which enables robots to use their peers' real operational trajectories as learning material, with humans only responsible for judging outcomes. On five real-machine tasks—flower arranging, clothes folding, phone charging, stool assembly, and phone packaging—the policies trained with CEL all exceeded the average level of 6 human teleoperators, leading by up to 16.7 percentage points, with an overall success rate of 97.5%; in the phone packaging task, the number of robots participating in training expanded from 16 to 96, and the success rate rose from 86.7% to 98.3%. The accompanying FERT algorithm uses failure trajectories to identify weaknesses and retrieves successful segments for prioritized sampling, while the FATS fault-tolerant asynchronous training system completed a cumulative 9,834 rounds of task collection in the 96-robot experiment, with the machine-time loss caused by 30 failures accounting for only 1.3% of the cluster's total operating time.

    Same story · 8 reports

    #Papers/Research#Embodied AI#English text is being prepared.

  6. Leikeji
    AI score35

    D-Robotics and Unitree Robotics jointly launch an embodied intelligence education solution, with the Sunrise S100 strengthening the computing foundation

    D-Robotics and Unitree Robotics exhibited together for the first time at the Changsha Education Equipment Exhibition and announced a strategic partnership, launching an embodied intelligence education solution with the Sunrise S100 computing box as the core computing foundation, combined with Unitree's Go2 quadruped robot and R1 humanoid robot. The solution comes with 15 themed courses totaling 180 lessons, covering elementary, middle, and high school, and offers a two-tier progression path: desktop-level and real-robot-level. The Sunrise S100 integrates a CPU, BPU, and MCU within a single SoC, achieving integrated computing and control; education is the first implementation scenario of the two parties' collaboration.

    Same story · 3 reports

    #Industry news#Embodied AI#English text is being prepared.

  7. Shanghai Observer
    AI score17

    X-GAME 2026 Intelligent New Energy Vehicle Big Data Competition Finals Held in Jiading, Shanghai

    The finals of the X-GAME 2026 Intelligent New Energy Vehicle Big Data Competition were held on October 30 in Jiading, Shanghai. Twenty teams stood out from more than 100 entries and presented their defenses using data models and AI algorithms around real-world scenarios such as sanitation vehicle range, road condition prediction, and fleet charging. Across the three tracks, this year's competition attracted nearly 1,000 registrations in total, covering 29 provinces, autonomous regions, and municipalities across the country, 181 domestic universities, and 6 overseas universities. The "X-SEEK Smart Talent Direct Program" was also launched on site, with companies including GAC, BAIC, IM Motors, NIO, and Leapmotor setting up booths to recruit.

    #Industry news#English text is being prepared.

  8. Testing Catalog
    AI score28

    October 30 AI Daily Brief: Mistral Large 4 Tops Open-Source Rankings, Google Shuts Down Jules

    Mistral released the open-weight model Mistral Large 4, taking the top spot among open-source models on FrontierCode. Google shut down Jules and merged its features into Antigravity, which launched a plugin marketplace including Jira, Notion, Stripe, and Figma. OpenAI doubled the usage quotas for Pro and Business Premium, while Claude's mobile app added multi-account login, CarPlay, and 24 languages.

    Quoted post🚨 AI News | TestingCatalog · @testingcatalog

    DAILY AI BRIEF 🗞 — Oct 29 GOOGLE 🔥: - Gemini 4 Carbon tops the AA Intelligence Index at 59, rolling out first to cyber defenders. - Alphabet Q3 revenue rose 25% to $127.6B, with Cloud up 103% and Gemini at 1B users. - Connected Apps Platform is in preview for Search and Gemini, with Adobe, Calm and Peloton. - Gemini Notebook will soon get Connectors, starting with Google Drive. - XREAL AURA, the first Android XR glasses with Gemini, launched from $1,099. OPENAI 🔥: - API docs reference an unannounced GPT Image 2.6 with markup-guided editing. - ChatGPT Pro and Business users can claim up to $500/month in API credits. - US Pro users can now message Dots via iMessage or RCS. - ChatGPT and ChatGPT Work will merge later this year, per an early pop-up. - OpenAI API, Codex and ChatGPT Work are now on Oracle Marketplace in the US. MICROSOFT 🔥: - GitHub launched Copilot Fusion, nearly matching Claude Code at about half the cost. - Copilot Projects orchestrates multiple agents across repos. - Copilot Automations runs jobs on demand, on a schedule or on GitHub activity. - GitHub added a Copilot Security extension and Copilot dreaming. - Microsoft FY27 Q1 revenue hit $92.3B, with Azure up 48%. SPACEXSI 🔥: - Grok API now accepts guidelines to steer handling of content categories. - Grok Bot launched referrals with a two-week free Premium trial for friends. ANTHROPIC 🔥: - Claude for Legal added ready-made artifacts for contract review and cite checks. PERPLEXITY 🔥: - Voice mode now runs on OpenAI's gpt-live-1, and Computer supports multiple sessions. MISTRAL 🔥: - Mistral's Responses API is in public preview for building agents. NVIDIA 🔥: - Nemotron 3.5 Transcribe is an open-weight model for transcription in 40 languages. AMAZON 🔥: - Amazon unveiled five new Echo devices powered by Alexa+. * Used Grok to compose this brief, cherry-picking the news and doing some post-editing.

    #Industry news#Google#OpenAI

  10. AYi AI Notes
    AI score20

    Altman Talks About OpenAI's New Model Cracking Math Problems After Drinking

    OpenAI CEO Altman revealed in an interview after drinking that ChatGPT's new model, over an internal "Hold My Beer" weekend, successively cracked multiple unsolved interdisciplinary problems, with the proof related to the Navier-Stokes equations alone consuming $10 million in compute. He said that the actual scientific research findings produced far exceed the several hundred results publicly disclosed, and admitted that the excitement was accompanied by more sadness and melancholy, feeling that "something is coming to an end."

    Quoted postAYi · @AYi_AInotes

    English text is being prepared.

    Same story · 3 reports

    #English text is being prepared.#OpenAI

  11. Leiphone
    AI score35

    ZKTeco's ZKBrainScape Brain Center Opens, Jointly Releases Four New Non-Invasive Brain-Computer Interface Products

    At its Global Partner Conference in Dongguan, ZKTeco inaugurated the over 3,500-square-meter ZKBrainScape Brain Center and, together with WAGAWAGA, NeuroSky, and Xinling Zhihe, released four new brain-computer interface products: the XF10 and XM20 wearable EEG interaction terminals, a complete BCI experiment kit, and the BDE brain-driven exoskeleton. The Brain Center showcases the NSPU neural computing chip, a liquid-helium-free magnetoencephalography system, and neural data governance technologies such as federated learning and zero-knowledge proofs, covering scenarios in health management, smart education, industrial safety, and rehabilitation assistive devices. At the same conference, the MarsLume AI full-domain solution was released globally.

    #Industry news#Embodied AI#English text is being prepared.

  12. Leiphone
    AI score28

    Doubao Launches Check-in Credit Campaign: 8,800 Credits for 30 Cumulative Days of Check-ins, New Users Get a Free 30-Day Subscription

    Doubao has launched a daily check-in credit campaign for all users. Users can earn 8,800 credits after 30 cumulative days of check-ins, which exceeds the 8,400 credits of a one-month standard plan. New users who log in to Doubao Desktop or Doubao Work Desktop for the first time can claim a free 30-day trial plan. Check-ins earn 200 credits per day, doubled on Mondays. The gifted credits are valid for 30 days and can be used for all capabilities except video generation and Pro models. This is the third time since Doubao Work launched that credit benefits have been distributed to all users.

    Same story · 2 reports

    #Product updates#Agents

  13. LeiphoneProduct release
    AI score67

    NVIDIA and MIT Team Open-Source a 4K Video Enhancement Solution Based on MiniMax H3

    The Efficient-Large-Model team, composed of researchers from NVIDIA and the Massachusetts Institute of Technology, has released the model weights for the video enhancement project Open-H3-Regenerate-2K on Hugging Face, and the inference code has been made available in the NVlabs/Sana repository. The solution is based on MiniMax H3 and supports 2560×1440 2K output, and can achieve native 4K regeneration at up to 3840×2160. During training, the H3 Transformer backbone and the video VAE are frozen, and only the newly added refinement LoRA is trained. Team tests show that on a single NVIDIA B200 GPU, the combined time for upscaling, three-step refinement, and 2K decoding of an approximately 5-second video dropped from 87.26 seconds to 22.45 seconds; on RefinerBench-150, the solution's UniPercept average score is 60.0, compared with 53.4 for the official H3-Regenerate-2K API.

    Same story · 2 reports

    #English text is being prepared.#Video#Open-source ecosystem

  14. LeiphoneOpinion
    AI score54

    Rokid Releases Full-Color AI Glasses Rokid Color, Misa Says Losing If It Can't Reach Ten Million Units

    Rokid founder and CEO Misa said after the debut of its first full-color AI glasses Rokid Color that next year Rokid's shipments will be in the millions, the year after that in the high millions, and the year after that ten million will definitely be reached, no later than 2030. Rokid Color was initiated in November 2024; through a partnership with JBD to customize a 0.16cc Micro-LED full-color optical engine, it retains complete functions such as a camera and speakers within a 55-gram body, falling short of Misa's previously stated goal of under 45 grams. The article also sorts out the unresolved issues of full-color solutions in areas such as color cast, yield and cost, field of view, power consumption and heat dissipation, and the lack of killer native applications.

    #Industry news#Multimodal#English text is being prepared.

  15. Leiphone
    AI score57

    Unitree Technology's First Quarterly Report After Going Public: Revenue of 1.817 Billion Yuan, Non-Recurring Profit Down 18.8%

    In the first three quarters of 2026, Unitree Technology's revenue reached 1.817 billion yuan, up 55.7% year-on-year and already exceeding its full-year 2025 revenue of 1.699 billion yuan, but its gross margin fell from 60.4% in 2025 to 54.0%, and non-recurring net profit attributable to parent-company shareholders was 349 million yuan, down about 18.8% year-on-year. R&D expenses reached 263 million yuan in the first three quarters, up 191.5% year-on-year, and the R&D expense ratio rose from 8.8%–9% to 14.47%, with the increase mainly directed toward embodied large-model R&D and computing power. The company listed on the STAR Market on August 19, 2026, at an issue price of 150.80 yuan; its share price once surged to 1,100 yuan before pulling back more than 60%, and recovered to around 670 yuan around the time of the quarterly report.

    Same story · 5 reports

    #Industry news#Embodied AI#English text is being prepared.

  16. LeiphoneProduct release
    AI score60

    Xingyuan Intelligence Releases Cross-Embodiment Data Engine Chorus, Untangling the Challenges of the Embodied Data Pipeline

    Xingyuan Intelligence has released Chorus (Zhongxing), a cross-embodiment data engine for robot pre-training, which can convert human manipulation data collected by devices such as smart glasses, chest-mounted cameras, wrist cameras, and UMI into training data adapted to different robot embodiments. Xingyuan claims that the average usable data conversion rate has risen to 85%, that it adapts to 16 robot embodiments under the current test configuration, and that the processing cost per hour of effective robot trajectories has dropped to 15 yuan; combined with 3x instruction augmentation, this can achieve roughly 40x expansion of effective training data. The article also outlines the difficulties in the five stages of embodied data—filtering, annotation, motion conversion, validation, and feedback iteration—as well as the respective shortcomings of four types of data sources: real robots, simulation, UMI, and human first-person videos.

    #Industry news#Embodied AI#English text is being prepared.

  17. Leiphone
    AI score38

    Boya Intelligence's Liu Xin: After GPT-6, Hardware Will Hold Robots Back

    Liu Xin, founder of Boya Intelligence, judges that models may reach the capability inflection point earlier than hardware, and by then what limits robot operation will be the body rather than the brain. Boya's first dexterous hand, Gaoshan S1, has 22 degrees of freedom and uses a hybrid drive that is primarily tendon-driven, supplemented by direct gear drive and screw-rod linkages, and the company claims to be the world's first dexterous hand company to use a tendon-driven + screw parallel wrist configuration. Liu Xin also mentioned that domestic dexterous hand shipments last year were about 20,000 units, most of which were left idle after being used in humanoid robot performances.

    #English text is being prepared.#Embodied AI#Reasoning

  18. Leiphone
    AI score40

    IROS 2026 Observation: How Yuansheng Intelligence's Apex Hand Takes Dexterous Hands from Demo to Productivity Tool

    At IROS 2026, Yuansheng Intelligence showcased the five-fingered dexterous hand Apex Hand, defining the product based on real-world deployment needs with 21 degrees of freedom, a minimum grasp diameter of about 5 mm, and full-palm tactile sensing. In live demonstrations, the Apex Hand continuously grasped objects weighing over 10 kg for 8 hours a day with stable temperature, achieved a maximum single-hand active payload of 30 kg, and completed high-impact operations such as hammering nails and splitting wooden boards. Yuansheng also announced the "Yuanchuang Program," offering free loaner Apex Hand units to universities, research institutions, and developer teams worldwide.

    #Industry news#Embodied AI

  19. OpenBMBProduct release
    AI score58

    OpenBMB Reposts Real-World Test Case of MiniCPM-OCR on Handwritten Notes

    OpenBMB reposted a user's real-world test case of using MiniCPM-OCR to recognize actual handwritten notes, and thanked them for sharing. The quoted tweet explains that MiniCPM-OCR consists of one model responsible for reading documents and another model that evaluates the output and flags parts that may be wrong, and is open-sourced under the Apache 2.0 license.

    Quoted postHasan Toor · @hasantoxr

    OpenBMB just released an OCR parser alongside a separate model that evaluates its outputs. It’s called MiniCPM-OCR. One model reads your documents. A second model tells you which parts it probably got wrong. 100% open source. Apache 2.0. ↓ https://t.co/ihbtDjZaUy

    #Industry news#Multimodal#Product updates

  20. Ma Dongxi NLP
    AI score33

    Dongxi's Pick: How System-1 Determines the Scaling Direction of AGI

    English text is being prepared.

    Quoted postDr. Cheems Wang 🏡 · @AlbertW24045555

    Every rollout, reasoning step and tool call is paid for before you know if it helped. Most don’t. We argue the next leap toward AGI comes not from computing more, but from a cheap System-1 that decides what a costly System-2 should compute at all. https://t.co/6GIRws2YNi

    #Papers/Research