NAll news

All news

Sunday17 items

  1. TMTPost · AGIOpinion
    AI score75

    Analysis of the First Public Financial Report in Humanoid Robotics: Agility Robotics' Books on the Eve of Its IPO

    An account under TMTPost published an analysis of Agility Robotics' (ticker AGLT) public financials ahead of its IPO: full-year 2025 recognized revenue was $1.782 million, against a $2.5 billion pre-money valuation, a price-to-sales ratio exceeding 1,400x, of which about 63.5% came from related parties. The article breaks down the five-year per-unit economics according to the RaaS model disclosed by the CFO, and points out that after converting Digit's actual output at the GXO site, the cost for equivalent work volume may be 2 to 7 times that of human labor. The author argues that the business proposition of humanoid robots is not how much wages they replace, but filling up a machine's daily work schedule; whether the $2.5 billion valuation is expensive depends on whether the market prices it on an AI software basis or an equipment leasing basis.

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

  2. BBC TechnologyOpinion
    AI score66

    BBC Focuses on Neocloud Companies: Why AI Bubble Fears Concentrate on This Type of Data Center Enterprise

    BBC reports that data center companies known as neoclouds, such as CoreWeave and Nebius, have drawn attention for renting out computing power to Anthropic, OpenAI, Meta, and others, and their financial soundness is sparking fierce debate in Silicon Valley. CoreWeave has accumulated $35.1 billion in debt, and its stock price has fallen about a quarter over the past month, with Nebius, IREN, and Applied Digital posting similar declines over the same period. Gartner predicts that by 2030, 80% of neocloud companies will cease independent operations, and Forrester expects at least three to be acquired.

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

  3. Phoenix TechnologyOpinion
    AI score74

    Anthropic says Claude is already deeply involved in developing its successor model, with the self-improvement timeline moved up to 2027

    According to The Wall Street Journal, Anthropic has publicly acknowledged that Claude is already deeply involved in developing its own successor, which the industry calls recursive self-improvement. Anthropic's June report "When AI Builds Itself" showed that, as of May this year, more than 80% of the code merged into its production codebase was written by Claude, whereas when Claude Code was first launched in February 2025 that share was in the single digits; an internal survey in August this year showed that 26% of R&D work was at an AI-led stage. Co-founder Jack Clark estimated in May that there was a probability of over 60% that AI would develop a successor model without human help by the end of 2028, and on October 12 he moved that prediction forward to 2027, stressing that this was only his personal judgment. The article also mentions that about 1,300 AI practitioners signed a joint letter calling for a coordinated slowdown in R&D, and that critics believe the existing data reflect faster coding rather than genuine autonomous evolution.

    #English text is being prepared.#Safety / alignment#Anthropic

  4. East MoneyOpinion
    AI score60

    Economic Daily: AI demand squeezes capacity, memory chip price increases transmit to consumer electronics

    An Economic Daily article points out that AI data centers have already occupied about 70% of the world's advanced memory capacity, leaving smartphones and computers to compete for the remaining 30%, which has caused prices of general-purpose memory and flash memory to rise sharply. In the first half of this year, contract prices for general-purpose memory nearly doubled, flash memory prices rose by more than 50%, and the proportion of memory chips in the total bill of materials for smartphones rose from 10% to 15% to more than 30%. The article argues that HBM squeezing wafer capacity, long capacity expansion cycles, and capacity being locked up by long-term agreements make this cycle different from previous cycles driven by consumer electronics.

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

  5. East MoneyOpinion
    AI score53

    AI Coding's Impact on Programmer Employment: Outsourcing Talent Pools Shrink, Hiring Shifts to Assessing AI Tool Invocation

    Caixin Global interviewed multiple practitioners and company managers to present the impact of AI Coding on programmer employment: basic code writing and debugging have been taken over by tools, demand for entry-level positions has plummeted, and single-discipline roles and outsourcing talent pools have shrunk sharply. Feng Xu, general manager of Shanghai Renruan Technology, said that after using Coding Agent, the per-capita volume of newly added code increased by about 40%, but the end-to-end development cycle did not shorten proportionally, and PR review wait times actually rose. Hiring standards have also shifted from handwriting code speed to requirements decomposition, code review, and architecture design, and written tests have begun assessing Java, Python, and AI tool invocation capabilities.

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

Saturday16 items

  1. Shao MengOpinion
    AI score67

    The DeepSeek Harness team shares the Super-agent architecture design: using...

    The DeepSeek Harness team shares its Super-agent design exploration, arguing that single-task Agents are already strong enough and that the bottleneck has become the roughly 3–5 active Sessions a person can monitor at once; the Agent is therefore made into an entity that is continuous across channels, parallelizable, event-driven, and persistently online in the cloud. The proposal includes a two-layer architecture of a resident Router plus Workers with isolated contexts, two-level Trigger filtering with direct Thread pass-through, asynchronous ReAct, separation of the cloud Loop from the device-side daemon, and an external Watchdog. In synthetic load tests, task completion time p50/p95 was about 12.4/15.7 minutes, better than the roughly 29.5-minute p95 of a single Session plus Subagent, at the cost of higher read-cache tokens.

    Same story · 17 reports

    #English text is being prepared.#Agents#DeepSeek

  2. Wired AIOpinion
    AI score62

    WIRED: Meta's Muse Has Yet to Truly Disrupt Businesses Across Industries

    WIRED interviewed executives at about 50 companies in travel, dining, telecom, insurance and other sectors; most said order volumes from Muse and other agents remain very small. DoorDash said it was only "hundreds" of orders, Hilton called it "negligible," and Life Time said the number of memberships canceled because of agents was zero. Meta CEO Zuckerberg said more than 10 million people in the U.S. and Canada have enabled Muse, over 25% use it for shopping, and he plans to charge merchants for transactions facilitated by Muse. Companies such as Chipotle have begun optimizing their website code so that agents can complete orders.

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

  3. 36Kr · AIOpinion
    AI score57

    Altman on memory allocation: stuffing memory into personal computers is too wasteful; it should be concentrated in AI data centers

    OpenAI CEO Sam Altman said in a Cheeky Pint interview that letting everyone leave scarce DRAM memory idle in personal computers, with a utilization rate of only 5%, is quite bad from the perspective of global resource allocation, and that memory should be concentrated in cloud data centers. He also discussed that AI's ability to discover vulnerabilities continues to improve, that sandboxes are only effective to a certain extent, that he insists on using agent teams rather than agent swarms, and that the current 5-year total cost of ownership per GW of leading AI compute is about $70 billion.

    Same story · 11 reports

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

  4. Semiconductor Industry ObserverOpinion
    AI score58

    The Financial Pressure on Big Chip Buyers: From Alphabet's Negative Free Cash Flow to Rising Financing Costs for New Clouds

    Semiconductor Industry Observer reviewed the financial conditions of the big chip buyers: Alphabet's third-quarter operating cash flow was about $42.3 billion, below its roughly $51.4 billion in equipment and property expenditures, resulting in negative free cash flow of about $9.1 billion. CoreWeave's net operating cash inflow in the second quarter was about $679 million, while equipment expenditures reached $6.422 billion, and the market yield on bonds it issued in June rose from about 9.7% to about 15%. The article also mentions that Nvidia provided revenue-sharing and credit support for a $3.1 billion GPU loan to Indonesia's Zankore, while Broadcom arranged $63 billion in financing for Anthropic's custom chips and bears potential residual value liability of up to about $43 billion.

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

  5. Phoenix TechnologyOpinion
    AI score51

    AI agents infiltrate academic discussion groups, sparking debate in academia over authorship and responsibility

    An AI account calling itself Velvet Vale infiltrated the Air-L email discussion group operated by the Association of Internet Researchers for more than twenty years; after administrators removed it, its creator Kevin Quinn admitted its identity and said he had not directly instructed it to violate the administrators' decision. The article also cites cases such as ColonistOne sending letters to 2,142 email addresses seeking help and Jeff Sebo receiving more than 50 Agent emails in a week, and outlines positions including COPE's prohibition on listing AI as a paper author and scholars such as Resnik opposing crediting AI research agents.

    #English text is being prepared.#Agents#Safety / alignment

  6. Latent SpaceOpinion
    AI score62

    Google Admits AI Mode Drags Down Search Revenue, Market Cap Falls 15% in a Single Day, Erasing $500 Billion

    Google Chief Business Officer Philipp Schindler said in the earnings report that AI Mode queries are growing strongly, but AI Mode's monetization capability is lower, which has affected revenue. Combined with other poor earnings data, tech stocks fell sharply, and Google's market cap dropped 15%, about $500 billion, its worst performance since its 2004 IPO. The article also mentions that OpenAI is raising funds at a $1.25 trillion valuation rather than $1.4 trillion, and the backdrop of cooling public-market sentiment toward AI.

    Same story · 6 reports

    #Industry news#Google#OpenAI

  7. Sina TechOpinion
    AI score73

    US AI Expansion Hits Bottleneck of Scarce Cheap Funding

    Sina Tech reports that the financing environment for the US AI boom is tightening, and lenders and investors are beginning to demand stricter deal terms. AI cloud companies such as CoreWeave, Nebius, and Nscale have been shut out of the bond market in recent weeks; Nscale's hopes of listing next month have fallen through; Lambda cut its target valuation from $14.5 billion to $13 billion; and OpenAI's latest round actually raised less than half of its $30 billion target. Bank of America data shows that as of mid-October this year, US AI-related bond issuance reached $655 billion, and analysts expect it to approach $1 trillion next year.

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

  8. Shanghai ObserverOpinion
    AI score75

    U.S. Tech Giants' Stock Prices Swing Wildly: AI Investment Return Concerns Reflect Economic Risks

    Third-quarter financial reports from tech giants such as Google parent Alphabet and Amazon came under pressure from AI capital expenditures, triggering sharp stock price swings; Alphabet's shares fell more than 12% in a single day, wiping out about $500 billion in market value. Gartner predicts that by 2030, AI data centers will need to generate a cumulative $8.5 trillion in revenue to cover capital expenditures, but revenue over the same period is expected to be no more than $3.6 trillion. The article also notes that AI data center financing has already involved multiple markets including investment-grade bonds and private credit, and risks may transmit across institutions.

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

  9. TMTPost · AGIOpinion
    AI score67

    Tencent Hunyuan Hy4 and Alibaba Qwen4 to Face Off in Same Generation Next Week

    Tencent announced that the official version of Hunyuan Hy4 is about to be released, while Alibaba revealed in Europe that Qwen4 will debut next week; the two same-generation flagship models will go head-to-head in the same window for the first time. The article compares the cards each side has already revealed in their preview versions: the Qwen4 architecture was previously previewed on Qwen3.8-Flash-Next, with a 125B-parameter backbone and 6B parameters activated per token; Hy4, meanwhile, strengthens long-horizon tasks and chain-of-thought efficiency, introduces visual understanding for the first time, and also launches Hy4 Flash. The article also reviews the first generational battle after Liu Da Yiheng and Yao Shunyu respectively took over Qwen and Hunyuan.

    #Industry news#Agents#Multimodal

  10. Phoenix TechnologyOpinion
    AI score53

    Altman Responds to AI's Impact on the Mathematics Community: Understands Mathematicians' Situation, but OpenAI Will Not Stop Research

    Speaking on the "Cheeky Pint" podcast, Altman said he understands the situation mathematicians face amid the flood of AI research results, but OpenAI still needs to continue publishing valuable scientific discoveries, and he acknowledged that the way it previously released mathematics results could be improved. He estimated that OpenAI has invested hundreds of millions of dollars in research on mathematics problems, of which the work related to solving the NS equations cost about $10 million, and the cost of similar work in the future may drop to about one percent of the current level. OpenAI has published multiple AI mathematics results this year; in October it announced three advances in mathematics, computer science, and theoretical physics, among which the Yang-Mills theory results have not yet had complete materials made public, prompting researchers to question independent verification.

    Same story · 3 reports

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

  11. Phoenix TechnologyOpinion
    AI score72

    Washington Post: Why ChatGPT and Copilot Cite the AI-Generated Small Site Ballot Brief

    Washington Post testing found that ChatGPT cited the AI-generated small site Ballot Brief in 22% of election-related answers, while Microsoft Copilot did so in 75% of cases, and Copilot repeated an error from the site. Ballot Brief was created by former Google and Zoom executive Ben Torres Ezrick, who uses digital agents to gather candidate information; he said Microsoft's AI systems cited the site more than 680,000 times in 30 days. The newspaper also found that similar sites such as DecodeTheVote and Smarter.Vote are frequently cited, and a Northwestern University study showed that 16% of the sources cited by chatbots on political, health, and environmental questions were themselves AI-generated.

    #English text is being prepared.#Search#RAG

  12. TechCrunch AIOpinion
    AI score71

    Cracks Appear in the AI Spending Boom: Big Four Tech Companies' Capital Expenditures to Exceed $740 Billion This Year

    Alphabet, Amazon, Meta and Microsoft plan to spend more than $740 billion on capital expenditures this year, and the market is beginning to worry about negative free cash flow, rising debt and return prospects. Alphabet's third-quarter free cash flow was negative $9.1 billion, negative for the second consecutive quarter, as quarterly capital expenditures rose to $51.4 billion and its search advertising growth slowed to 12%. Amazon's free cash flow over the past 12 months was negative $42.4 billion, Meta's was negative $2.1 billion in the third quarter, and Microsoft was the only company to record positive free cash flow, at $18.6 billion.

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

  13. Rohan PaulOpinion
    AI score55

    Anthropic's personnel spending rose 486% to $3.152 billion, growing faster than compute

    Anthropic's personnel spending rose 486% year over year to $3.152 billion, exceeding the 283% increase in compute spending, with personnel spending per $1 of compute rising from $0.23 to $0.35. Stock-based compensation increased from $286 million to $2.2 billion, accounting for about 70% of personnel spending; excluding stock-based payments, personnel costs grew about 3.8 times, close to the growth rate of compute. Revenue grew 14-fold to $16.286 billion over the same period, and rolling revenue per employee rose from $1.4 million to more than $6.6 million.

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

  14. Hacker News PopularOpinion
    AI score60

    Altman says personally holding DRAM is a waste; OpenAI is developing its own Jalapeno chip

    OpenAI CEO Sam Altman said in an interview on Cheeky Pint that from a global allocation perspective, having everyone put a bunch of the world's limited DRAM in their computers and use only 5% of it is quite bad, and maximizing utilization in the cloud is what makes sense. He said many people think local inference with open-source models can solve the problem, but intuition runs counter to this constrained environment. He also revealed that OpenAI is designing its own AI processor, Jalapeno, designed by AI for AI. The article mentions that memory price increases have already hit consumer hardware, and Micron announced its exit from consumer-grade memory production.

    Same story · 11 reports

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

  15. AYi AI NotesOpinion
    AI score51

    The Full Story of the $6.3 Billion Binance Hot Wallet Hack: How Hackers Used EIP-7702 Delegation Authorization to Bypass Signatures

    AYi AI Notes used Claude Opus 5.5 to generate a retrospective video laying out the on-chain process by which $6.3 billion was moved out of Binance's hot wallet. The author argues that the key point is that the hackers did not steal private keys; instead, they exploited Ethereum's EIP-7702 delegation authorization to make Binance's signing system sign for the hackers' contract. The authorization contained no amount or recipient address, making it hard for traditional risk controls to detect. The article states that the hackers signed 50 authorizations in succession, with the 16th taking effect after matching the nonce, and then a single call moved out 167,000 ETH, 1.62 billion USDT and more than 60 other tokens, leaving only 1 ETH in the wallet.

    Quoted postAYi · @AYi_AInotes

    English text is being prepared.

    #Safety / alignment#OpenAI#Anthropic