NAll news

Saturday46 items

  1. Shao MengOpinion
    AI score53

    QodoAI founder: Agentic SDLC is not equal to a software factory; teams must build and own a bounded L4 system

    QodoAI founder Itamar Friedman believes that having Agents at every stage does not mean a software factory already exists; the factory's goal is a bounded L4 system that, on defined classes of work, autonomously completes the authorized lifecycle, produces verifiable evidence, and safely falls back on failure. Borrowing from autonomous driving levels, he devised an L0-to-L5 framework, pointing out that L4 does not mean omnipotence—operational boundaries are themselves part of the capability. Teams can first aim for L4 on routine changes to mature internal services, while still tightly controlling things like payment authorization. He emphasizes System > Agents > LLMs: a team's advantage lies in building and owning the boundaries, evidence, permissions, and feedback, rather than stacking models or buying black boxes.

    Agents across the SDLC are a step toward a software factory. Then the harder work is engineering the system around them. My take on the path to bounded L4 software engineering, and why engineering teams need to BUILD and OWN that SYSTEM. https://t.co/HfYgwUYd2Y

    #English text is being prepared.#Agents#English text is being prepared.

  2. 36Kr · AIOpinion
    AI score65

    SoftBank CDS rises to 432 basis points, Masayoshi Son continues to increase AI investment

    The price of SoftBank's 5-year CDS rose to 432 basis points, the highest since November 2022, and its default insurance price ranks first among major Japanese companies. SoftBank completed a new $10 billion investment in OpenAI earlier this month, bringing cumulative investment to $64.6 billion, and issued bonds with coupons as high as 9.75% to raise funds. The article also mentions the rising scale of CDS trading by AI-related companies, as well as the debt and computing power expenditure pressure of companies such as Anthropic and CoreWeave.

    #Industry news#OpenAI#Anthropic

  3. gihyo.jpOpinion
    AI score65

    AI accelerates cyberattacks; Anthropic, OpenAI, Microsoft and others disclose defense and agent control developments

    gihyo.jp reviews multiple releases and reports from late October, organizing AI-era defense challenges along two lines: vulnerability remediation implementation and permission management. Anthropic Deputy CISO Jason D. Clinton says that within about 18 months, high cyber capability will spread to small, cheap models, but fixes still need to be applied to running systems. The article cites OpenAI's discovery of Brotli's CVE-2026-111497 during testing in an isolated environment, and introduces remediation and credential control solutions such as AWS Continuum, Google Agent Compute Engine, and 1Password for Agents, as well as Microsoft's report on Storm-0940 using AI to automate attacks.

    #Industry news#Safety / alignment#Agents

  4. HuxiuOpinion
    AI score63

    Unprecedented trillion-scale investment—is NVIDIA the only winner?

    Huxiu reviews the global AIDC construction wave, noting that combined computing power investment is expected to reach $830 billion in 2026, up 79% year over year, but as of July 2026 more than 95% of NVIDIA's Grace Blackwell capacity has not yet been put into use. Industry consensus is that about 60% or more of giant capital expenditure corresponds to clear customers, while the remaining 30-40% is preemptive capacity grabbing. The article cites cases such as CoreWeave's Q2 revenue of $2.575 billion and net loss of $626 million to show that the three conditions—single-park revenue covering costs, stable utilization and rental prices, and follow-on financing continuing—are currently not broadly verified, and the certain winner may only be equipment seller NVIDIA.

    Same story · 20 reports

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

  5. YicaiOpinion
    AI score61

    When Meta Muse Collides with Amazon: AI Is Leaving the Chatbox to Contest the New Internet Gateway

    Meta launched its personal AI Agent Muse last month. It has an independent virtual computer and browser and can send emails, book travel, and shop on behalf of users. It reached about 2.8 million downloads within two weeks of launch, and at the October 21 earnings call it was announced that more than 10 million people had activated it. Less than two weeks after launch, Muse was blocked by Amazon from accessing its shopping site. Amazon said AI Agents should identify themselves and respect whether the service provider permits entry. The article argues that the two sides are fighting over who controls the subsequent search, comparison, decision-making, and payment after consumers form purchase intent, while Google keeps intent-to-execution within its own system through its own ecosystem such as Universal Cart.

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

  6. WallstreetcnOpinion
    AI score55

    Goldman Sachs trading head interprets institutions' significant AI position reduction: power gap and cost of capital in focus

    Rich Privorotsky, Goldman Sachs' head of European fixed income and equity trading, revealed in an interview on the "The Markets" podcast that last week the Goldman Sachs prime brokerage business book saw the largest technology stock selloff in its history. He noted that the market expects AI capital expenditure to reach 1.1 trillion US dollars next year, but investors demand to see backlog orders converted into revenue, while a power gap of about 40 to 50 gigawatts next year casts doubt on computing power delivery. He said the macro environment is shifting from a savings glut to a spending glut, and believes spot market computing power prices are still hitting new highs, with demand showing no sign of weakening.

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

  7. WallstreetcnOpinion
    AI score57

    DeepSeek did the math: $38 per GPU-hour — can the inference business make money?

    When DeepSeek open-sourced its inference framework DeepInfer on October 20, it announced that a single cluster can process 155 trillion tokens per day, which, converted at DeepSeek API off-peak pricing, corresponds to an average of $38 in revenue per GPU-hour. The article notes that this is theoretical billed revenue; DeepSeek has not disclosed the cluster's hardware composition, cost bills, or the share of paid traffic, and DeepInfer currently mainly targets DeepSeek-V4.1 Flash. The article further compares the R&D and inference expenditure structures of OpenAI and Anthropic, as well as cases such as CoreWeave, Cursor, and Character.AI, to show that high output value per GPU does not mean inference services are necessarily profitable.

    Same story · 4 reports

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

  8. East MoneyOpinion
    AI score53

    Xiong Bingqi: In the AI era, employment-friendly development must move beyond replacement thinking

    In the commentary, Xiong Bingqi analyzes the Opinions on Adapting to the Development of Artificial Intelligence Technology to Help Build an Employment-Friendly Development Model issued by the State Council. The document proposes tapping the AI employment potential of traditional sectors, mitigating the employment substitution impact of AI applications, and encouraging enterprises to establish unemployment buffer periods, notify workers in advance, and provide support for job transfers and transitions. The article argues that when an enterprise replaces an old position, it should simultaneously create a new one, rather than only cutting jobs, and suggests exploring employment models that improve efficiency without reducing staff and raise quality while shortening working hours.

    Same story · 15 reports

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

  9. HuxiuOpinion
    AI score77

    Compute companies expand with high-interest debt: Anthropic borrows $60 billion to buy TPUs, CoreWeave default probability exceeds 60%

    Anthropic borrows $60 billion at a coupon above 10% to purchase about 1.5 gigawatts of Google TPUs, with a Blackstone fund investing $9 billion; CoreWeave bond yields exceed 15%, and the CDS market prices its five-year default probability above 60%. High-interest debt has spread to compute players such as Volta, xAI and Hongxin Electronics; Nvidia holds about 10% of CoreWeave and has committed to buying unsold compute through 2032, and Goldman Sachs warns that the semiconductor supply chain may face an order cliff in 2027-2028.

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

  10. HuxiuOpinion
    AI score62

    SoftBank's 5-year CDS rose to 432 basis points, and Masayoshi Son is still increasing investment in OpenAI and AI infrastructure

    SoftBank's 5-year CDS rose to 432 basis points, the highest since November 2022 and the highest among major Japanese companies, with its stock price down more than half from its June high. SoftBank has invested a cumulative $64.6 billion in OpenAI, holding about 13%, with the highest coupon on new bonds reaching 9.75%, and is still negotiating participation in a new funding round and the acquisition of Allied Universal. As of the end of June, SoftBank's LTV was 13%, and Bloomberg expects net cash outflow of at least $6 billion in the fiscal year ending March 2027; in the same period, the average weekly trading volume of CDS for AI-related companies such as Oracle and Nvidia rose from less than $2 billion in July to more than $9 billion in September.

    #Industry news#OpenAI#Anthropic

  11. HuxiuOpinion
    AI score53

    In-depth discussion of Personal Agent: Muse costs $10 to $100 per month, and the industry awaits the iPhone moment

    Huxiu reposted a discussion from the overseas unicorn Best Ideas community, outlining the shift of Personal Agent from a chatbot to a cloud assistant that proactively handles tasks. Meta's Muse has been online for over a month with about 3 million DAU, but its day-14 open rate is lower than ChatGPT's, and about 90% of interactions are still initiated by users; the estimated monthly cost per user is in the $10 to $100 range, with early teams estimating about $50 to $100 including tokens and compute resources. The article argues that shopping and personal finance are the scenarios where value will be validated first; big companies have advantages in compute, identity, and entry points, while China is constrained by the closed ecosystems of super apps.

    #English text is being prepared.#Agents#English text is being prepared.

  12. 36Kr · AIOpinion
    AI score56

    Are Wang Hong and Wei Dongyi and their peers also about to be "slain" by AI?

    A 36Kr-affiliated account reviewed how, after OpenAI released 722 mathematics papers at once and claimed progress on problems such as the zero-free region of the Riemann hypothesis and the irrationality of Catalan's constant, Fields Medalist Hugo Duminil-Copin said AI was bulldozing the problems he was working on as if it were a bulldozer. The article also mentioned that an Anthropic researcher had previously claimed to have disproved the Jacobian conjecture, and that a doctoral student in operator algebras adjusted their paper plans after their results were scooped. The author thereby discusses mathematics becoming a battlefield for large models to demonstrate reasoning ability, and mentions the alliance of mathematics and artificial intelligence laboratories led by Yau Shing-Tung.

    #English text is being prepared.#Reasoning#OpenAI

  13. 36Kr · AIOpinion
    AI score72

    Double 11 E-commerce AI Shadow War: The Different Calculations of Qwen, Doubao, JD, and Pinduoduo

    36Kr reviews the competitive landscape of e-commerce AI during this year's Double 11: Alibaba launched the personal intelligent assistant "Qwen Go!" on October 13, and released the Qwen Office e-commerce edition the next day, extending from user-side shopping agents to merchant-side cross-platform operations. The Doubao mobile assistant, however, encountered resistance due to GUI automated operations; as of October 14, more than 150 apps had refused its Screen Automation Operation Declaration Protocol (SAEP), including Taobao, Alipay, Meituan, and Pinduoduo, while JD has not yet appeared. The article argues that what platforms resist is external Agents' unauthorized operations on their own apps, and the core disagreement lies in who controls the traffic entrance and the advertising monetization mechanism.

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

  14. TMTPost · AGIOpinion
    AI score53

    Tencent QClaw announced service shutdown after only 188 days of public beta, as big-tech AI products accelerate reshuffling

    Tencent QClaw announced on September 24 that due to business adjustments it will cease operations on December 24, 2026, and guide users to migrate to WorkBuddy; from its public beta on March 20 to the shutdown announcement was only 188 days. The article reviews the product moves of eight companies—Tencent, Alibaba, Baidu, ByteDance, and Moonshot, Zhipu, MiniMax, DeepSeek—across ten AI tracks over four years, noting that competition cycles are shortening and product lifecycles are getting shorter, with standalone products continuously converging toward general assistants and underlying platforms. Alibaba's Qwen announced it will soon launch a personal assistant product "Qwen Go!", and ByteDance was also reported to be advancing a personal assistant project codenamed "Spell".

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

  15. TMTPost · AGIOpinion
    AI score71

    Let AI spend for you: how long can the personal Agent wave fueled by Meta Muse last

    Meta's personal agent Muse accumulated 10.9 million downloads in its first month and a half, including 10.2 million in the US; about 25% of users use it daily, over 25% shop through Muse, and the average order value exceeds $200. The article reviews domestic and international offerings such as ByteDance Xiaodou, Tencent LightVela, Qwen Personal Agent, and Today AI, and notes that Muse's day-7 retention falls to just over 30% and day-14 retention is only 22.5%, as well as ecosystem and security challenges such as Amazon banning Muse from accessing its shopping platform permissions.

    #English text is being prepared.#Agents#OpenAI

  16. HuxiuOpinion
    AI score67

    Tesla's Q3 operating profit halves to $720 million; Musk leans on SpaceX merger to support valuation

    Tesla delivered 486,500 vehicles in Q3 2026, down 2.1% year-on-year, with revenue flat at $28.05 billion, operating profit halved to $720 million, and R&D expenses up 69% to $2.76 billion. Musk confirmed that the Tesla and SpaceX boards have each established special committees to evaluate a merger, with the two companies valued at a combined approximately $3.5 trillion; the deal requires approval from more than half of Tesla shareholders. BYD's pure electric sales in the first three quarters were 1.63 million vehicles, leading Tesla by more than 300,000, with automotive gross margin of 22.3%, higher than Tesla's 18.1% excluding carbon credits.

    Same story · 11 reports

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

  17. HuxiuOpinion
    AI score65

    The AI virtual character Xiaohai uses a tragic persona to attract middle-aged and elderly women and sell products, with over 600 million views on a single platform

    The AI virtual character Xiaohai uses a tragic yet sunny persona to attract middle-aged and elderly women, and after accumulating followers shifts to high-commission product sales; according to CCTV News, related videos have accumulated over 600 million views on a single platform. A Beijing News investigation states that such videos are cheap to produce and can be automated, earning two to three thousand yuan a day, with product-sales commissions reaching 52%. After reports by multiple media outlets, the related account stopped updating and deleted its works in bulk, but the author believes similar scams may recur.

    #English text is being prepared.#Video#Agents

  18. Cailian PressOpinion
    AI score60

    AI Shortage of Money Reshapes Industry Narrative; Market No Longer Believes Burning Cash Wins

    Cailian Press reported that AI investment continues to siphon market liquidity; Meta raised its full-year capital expenditure to $135 billion to $145 billion and newly established non-voting Class C shares to pave the way for financing; SpaceX's superintelligence business posted quarterly revenue of about $7.6 billion and capital expenditure of $16.4 billion. JPMorgan estimates that this year global listed tech companies will have operating cash flow of about $2.2 trillion and capital expenditure plus R&D spending of about $1.8 trillion, leaving a surplus of about $400 billion, but new cloud vendors are more dependent on external financing. Meta's stock fell nearly 7% in a single day, wiping out about $124 billion in market value; among the Magnificent Seven, Apple hit a record high thanks to its asset-light approach.

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

  19. East MoneyOpinion
    AI score56

    AI Funding Shortage Reshapes Industry Narrative: Market No Longer Believes Burning Money Wins

    A Cailianshe article points out that the AI industry is not facing an overall funding drought, but rising financing costs and thresholds are changing market expectations about burning money. JPMorgan estimates that global listed tech companies will have about $2.2 trillion in operating cash flow this year, with capital expenditure and R&D spending totaling about $1.8 trillion, leaving a surplus of about $400 billion; while new cloud vendors have about $30 billion in operating cash flow against about $80 billion in corresponding spending, making them more dependent on external financing. After Meta raised its full-year capital expenditure to $135 billion to $145 billion, its stock fell nearly 7% in a single day; SpaceX shares fell more than 9% the day after its earnings report; Apple hit a record high on its asset-light approach.

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

  20. A Yi AI NotesOpinion
    AI score71

    A Yi claims Kimi K3.1 Pro surpasses Opus 5.5 and GPT-6 Astra, and announces open-sourcing of weights

    A Yi AI Notes posted that Kimi released the 2.8T flagship K3.1 Pro, topping the SWE Marathon with 51.7 to take first place, surpassing Opus 5.5 and GPT-6 Astra, and comes standard with a 1M native context window. The post also claims the official team promised to open-source the full weights before November 4, and the author also expressed dissatisfaction with Claude account bans.

    English text is being prepared.

    Same story · 34 reports

    #Industry news#Model releases#Open-source ecosystem