All news

WWallstreetcn15:46Opinion

AI score76

Anthropic loses $80.7 billion in half a year, compute commitments up to $750 billion, Polymarket gives 17% bankruptcy probability

Data from prediction market platform Polymarket shows that Anthropic has a 17% probability of bankruptcy before the end of 2028. According to its draft S-1 prospectus, the company's net loss in the first half of 2026 exceeded $80.7 billion, of which about $72 billion was non-cash accounting adjustments, with actual operating loss of about $8.7 billion; compute infrastructure committed spending could reach up to $750 billion, and most contracts specify payment regardless of usage. First-half 2026 revenue was $16.3 billion, of which about 20% came from two customers; third-quarter revenue growth slowed quarter-on-quarter from about 140% to 45% to 46%, and some institutions cut the reference valuation from $2 trillion to about $1.5 trillion.

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

WWallstreetcn15:46Opinion

AI score52

JPMorgan: New Cloud Providers' Financing Pressure Won't Shake AI Infrastructure Cycle; Tech Sector Still Has $400 Billion Surplus This Year

JPMorgan said in its report that the risks of new cloud providers' financing models are not enough to shake the overall AI capital expenditure cycle. Global listed tech companies are expected to generate 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 after offsetting the two. The report said that even if new cloud providers run into financing difficulties, their construction gap can be filled by companies with more robust balance sheets, such as hyperscale cloud service providers; under a stress scenario, even if capital expenditure rises 25% per year through 2030 and cash flow grows 20%, the tech sector will still maintain a small surplus.

Same story · 2 reports

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

WWallstreetcn14:45Opinion

AI score72

Ramp Data: 1% of Companies Account for 80% of OpenAI and Anthropic Commercial Spending, Token Usage Rises While Payments Stall

After tracking payment data from more than 70,000 U.S. companies, Ramp found that just 1% of companies accounted for 80% of OpenAI and Anthropic commercial spending; token spending showed almost zero growth in August and September, and in early October was only 1.5% above the August peak. Ramp chief economist Ara Kharazian attributed the stagnation to price deflation. Ornn's Anthropic Token price index has fallen from more than $1.70 in mid-September to $0.83, while the corresponding OpenAI index stands at $0.45. In September, the paid AI adoption rate across the corporate network rose only 0.62 percentage points to 55.9%. Goldman Sachs and Apollo chief economist Torsten Slok both questioned whether compute demand can support sustained investment.

#Industry news#OpenAI#Anthropic

AAlibaba Cloud Developer Community AI14:36Opinion

AI score58

Google Open-Sources AQuA Quality Agent: What Should Test Teams Hand Over to It

On October 8, the Google Developers Blog published a reference implementation of AQuA (Ambient Quality Agent), which runs a quality agent alongside production agents, responsible for clustering failed conversations, reviewing issues, and anchoring diagnoses to deployed source code snapshots. The article argues that it will not make testing disappear, but rather hands over to test teams the tasks of defining anomalies, calibrating evidence, and bearing release responsibility, and presents three validation methods: injecting known faults, blind review controls, and root-cause replay.

#Industry news#Agents#Open-source ecosystem

AAlibaba Cloud Developer Community AI14:36Opinion

AI score55

Google Open-Sources AQuA Quality Agent: What Should Test Teams Hand Over to It?

On October 8, the Google Developers Blog published the AQuA (Ambient Quality Agent) reference implementation, which places a quality agent alongside production agents, responsible for clustering failed conversations, reviewing issues, and anchoring diagnoses to deployed source code snapshots. The article argues that this type of outer-loop quality inspection may still share blind spots with production agents in models, logs, and business definitions, and that test teams need to shift toward designing observability surfaces, calibrating anomaly labels, maintaining evidence chains, and defining escalation rules. The article also presents three validation methods—injecting known faults, blind-review controls, and root-cause replay—as well as a phased adoption path for small teams.

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

336Kr · AI14:36Opinion

AI score56

Anthropic adopts a PBC plus long-term trust governance structure in its bid for IPO

When Anthropic pursues an IPO, it adopts a governance structure combining a public benefit corporation (PBC) with a long-term independent trust (LTBT). Control of the company is handed to external trustees who hold no shares and receive no dividends, and former Federal Reserve Chair Ben Bernanke is among the trustees. The trustees have the power to replace a majority of the board, and major transactions such as large-scale unconventional borrowing, the sale of more than half of the equity, or the transfer of core AI models reaching a threshold must be reported five days in advance; scholars believe this creates a soft veto power. The article compares OpenAI, which is controlled by a nonprofit foundation, and xAI, which abandoned the PBC and reverted to an ordinary commercial company, and points out that Anthropic's transfer of some shareholder rights may lead to a valuation discount.

#Industry news#Anthropic#OpenAI

336Kr · AI14:36Opinion

AI score56

Why large model vendors cannot decide their own position in embodied intelligence

The article compares DeepSeek's investment in Unitree and its planned lead investment in Winverse AI with OpenAI's in-house development of humanoid robots, analyzing the different positions of large model vendors in embodied intelligence. After SenseTime connected GPT-6 Astra to the RoboDojo testing platform, 9 tasks that had previously failed 5 times in a row became 5 out of 5 successes once more steps were allowed, for a success rate of 81.9%; however, the bowl-stacking task took about 131 minutes and $19, and the pen-into-pen-holder task took about 263 minutes and $55. On this basis, the author argues that the bottleneck in physical interaction lies in cost and speed, and classifies market players into four categories according to who bears the fixed costs.

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

WWallstreetcn12:44Opinion

AI score52

Jensen Huang responds to CoreWeave debt, GPU depreciation period, and DeepSeek inference gross margin

In an interview with Yahoo Finance after the GTC Berlin conference, Jensen Huang responded to three major questions on AI investors' minds: the debt-financing risk of new cloud computing companies, GPU service life, and the inference gross margin data disclosed by DeepSeek. He said Nvidia GPUs can last seven to eight years, far longer than the four years commonly assumed by the market, and that the company has already provided guarantees for the residual value of chip resale in some financing projects; he also introduced the DSX Pod solution, which prefabricates AI factories in containers, and said it is expected to compress the deployment cycle for gigawatt-scale data centers from four to six months to three to four months or even less. Gavin Baker, citing DeepSeek data, said the monetized price of GPU inference for its open-source models reaches $38 per hour, far above market rental prices, while Jensen Huang stressed that strong demand is the fundamental prerequisite for solving financing and supply problems.

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

336Kr · AI11:35Opinion

AI score56

Yuanchuan Research Institute: Following the Personal Agent Trend Is Not Advisable

Yuanchuan Research Institute analyzes the phenomenon of major Chinese tech companies following the Personal Agent trend, arguing that China's super apps are highly concentrated and that user time is the basis for monetization; cross-platform Agent calls would directly conflict with existing business models. The article mentions that one month after the consumer version of the Doubao phone launched, it claimed to have been blocked by 151 apps, and cites the "Agentic Walled Gardens" hypothesis proposed in Microsoft Research's "Agent Economy," pointing out that closed ecosystems operated by various companies may prevent digital assistants from interoperating with one another.

#English text is being prepared.#Agents#OpenAI

336Kr · AI11:35Opinion

AI score52

Who Is Feeding Anthropic: $4.6 Billion in Revenue and Over $8 Billion in Losses

36Kr's Brain Polar Body outlet analyzes Anthropic's business structure: according to Reuters, its 2025 revenue approached $4.6 billion, up about 12-fold year-over-year, with more than 80% coming from B2B services billed by API and token, and the two major cloud platforms Amazon and Google contributing about 47% of sales. The article also notes that its 2025 operating loss exceeded $8 billion, including $7.3 billion in compute and infrastructure spending, and that Anthropic has committed to paying Google at least $111.1 billion, Amazon $110 billion, and Microsoft $31.4 billion over the next seven to ten years. The author argues that rumors and account-selling businesses surrounding Claude on the Chinese internet cannot explain its real revenue; what truly determines its value are the B2B customers who continuously call the models and the cloud platform channels.

#Industry news#OpenAI#Google

336Kr · AI11:35Opinion

AI score60

The Personal Agent battle spreads to hardware: the different paths of Qwen, Shokz, viaim and Bee

36Kr's All-Weather Technology reviewed the trend of Personal Agents entering AI glasses, earbuds and recording devices: on October 13, Qwen released the AI glasses N1 and N1 Pro and AI clip-on earbuds, with the glasses starting at 2,218 yuan and the earbuds at a first-sale price starting at 1,099 yuan, and said the goal is for the same Qwen to understand users and continue tasks across devices. The article compares different divisions of labor, such as Shokz OpenFit 2 AI calling on Qwen for transcription summaries, viaim using its own Agent to orchestrate other Agents, and Amazon's Bee integrating with Alexa+, and points out that hardware manufacturers are seeking their own positions around context management, task invocation and subscription services.

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

336Kr · AI11:35Opinion

AI score58

Hot Money in Embodied Intelligence Shifts: Unitree's Stock Price Rebounds as Yinghe and a Wave of Robotics Companies Head Toward Bankruptcy Liquidation

Over the past three trading days, Unitree Technology's stock price has risen nearly 30% cumulatively, with its market value returning above 200 billion yuan. The company announced that operations are normal and cautioned that its rolling price-to-earnings ratio has reached 388.25 times. In the same period, Yinghe Robotics held its first creditors' meeting after bankruptcy liquidation on October 19, and a wave of robotics companies including Zhicheng Dynamics and CloudMinds Robotics have also successively gone bankrupt. The article notes that in the first half of 2026, domestic embodied intelligence financing totaled more than 93 billion yuan, a fivefold year-on-year increase, but in global humanoid robot shipments in the first half of the year, intelligent manufacturing and warehousing logistics accounted for only 13% and 5% respectively, while entertainment performances, scientific research and data collection still accounted for more than 60%.

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

CCNET11:34Opinion

AI score53

Experts Say AI and Data Centers Will Influence US Midterm Election Voting

CNET reports that a poll by ABC News, The Washington Post, and Ipsos shows that 57% of Americans believe AI is harmful to society, 66% oppose building new data centers in the US, and opposition rises to 72% if one is built locally. Political and polling experts say opposition to AI and data centers has become bipartisan and could affect results in key districts, with candidates who previously voted in favor of building data centers under particular pressure. 67% of respondents believe the government has not done enough to regulate AI.

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

LLydia Hallie10:46Opinion

AI score60

Developer shares experience of using Claude Duet to monitor Slack, email, and calendar

Lydia Hallie says Claude Duet makes focused work easier, as she no longer has to be online at all times and context switching is noticeably reduced. Her approach is to have Claude Duet monitor Slack, email, and calendar, alerting her only when something is important, including threads she is not part of, often with relevant documents or designs attached as context. Claude officially announced the opening of the Claude Duet research preview earlier, describing it as a proactive agent app that alerts users when needed, completes tasks they assign, keeps working while they are away, and checks with them before acting on their behalf.

Quoted postClaude · @claudeai

We’re opening a research preview of Claude Duet, a new app where Claude works as a proactive agent. It comes to you with what needs your attention, completes tasks you ask of it, keeps working when you step away, and checks with you before it acts on your behalf. https://t.co/kt8gAbWuX8

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

SShao Meng10:46Opinion

AI score62

Microsoft CEO Satya Nadella's 2026 Annual Letter to Shareholders: Mapping Token Costs to Business Outcomes, Building a Frontier Ecosystem

In his 2026 annual letter to shareholders, Microsoft CEO Satya Nadella proposed building a frontier ecosystem and introduced cost-to-outcome—the business results obtained per unit of token spend—as a new optimization metric. In fiscal year 2026, Microsoft's revenue reached $331.8 billion, up 18% year over year; Azure grew 41%, surpassing $100 billion for the first time; M365 Copilot paid seats exceeded 30 million, and GitHub Copilot users reached 50 million. The letter states that interaction is shifting from synchronous chat to asynchronous agents, with the harness separated from the model; Foundry offers more than 11,000 models and has over 100,000 customers, and Microsoft Frontier Co. was established, deploying 6,000 experts on-site.

Quoted postSatya Nadella · @satyanadella

In my annual letter, I share how we’re building a frontier ecosystem to turn technological progress into real-world outcomes and economic opportunity for people, organizations, and communities everywhere. https://t.co/Zcrb96sCC3

#English text is being prepared.#Agents#Reasoning

SShao Meng10:46Opinion

AI score60

Nathan Lambert Harvard lecture: Chinese open-source models are fully ahead; US compute advantage is about 16x

In a guest lecture for Harvard CS 2881R "AI Security," Nathan Lambert judged that open-source models lag the closed-source frontier by about 3 to 5 months and the gap is narrowing; Chinese open-source models are fully ahead in performance, downloads, and paper citations, while US open-source models lag China by another 3 to 4 months. He believes treating distillation as a core issue overestimates it; the US's truly durable advantage is compute, with the combined compute of the top five labs averaging 16 times that of China's top five, and by the end of 2026 Anthropic and OpenAI each having more than 5 GW of compute, while China's leading labs have at most 100 to 200 MW. He also pointed out that banning open source is not feasible in reality and would only suppress economic activity, and suggested assessing additional risk on a per-model basis with unified national standards.

Quoted postNathan Lambert · @natolambert

Finally got to put together a complete deck on the state of open models. What they are, licenses, how competitive they are, signs people are using them, distillation's impact, and how we balance the risks and benefits of openness! Will record a video lecture of these soon, and hopefully re-do them a few times a year.

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

SShao Meng10:46Opinion

AI score67

Claude Mods In-Depth: Harness Shifts from Managing Machines to Managing Human Attention

Shao Meng relays Dan McAteer's article on latentspacepod, arguing that Claude Mods lets users and Claude work together to use code to transform the human-facing layer of Claude Code—deciding when to interrupt, what to display, and which decisions to let users make—and that it is an early form of an attention interface. Mods are essentially TS/JS plugins that listen to prompt, tool-call, and interface-rendering events, and work in both the CLI and the desktop app. The official release already includes 6 built-in Mods: agents-md, diff, plugin-authoring, sec-default, telemetry, and you-should-know. Dan also points out the noise problem: after installing it on October 15, Supervisor was triggered in 99 out of 111 rounds; in addition, each check consumes about 5–6k equivalent input tokens, and Mods run with user permissions and have no sandbox.

Quoted postDan McAteer · @daniel_mac8

There's not enough buzz around Claude Mods. Read this Claude Mods article if you want to: - Understand Claude Mods basics - Install and use the most useful mods today - Ponder where the agent-human interface will go For @latentspacepod.

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

AAlex Heath10:46Opinion

AI score58

Sundar Pichai responds to Google's free cash flow turning negative for the first time while still ramping up AI capital expenditure

Alex Heath asked Google CEO Sundar Pichai why the company recently let free cash flow turn negative for the first time in order to increase AI capital expenditure, and Pichai responded that it is not based on hope alone. He said making this decision requires confidence on multiple levels: Google sees early signals of diverse opportunities and strong demand from consumers to enterprises; at the same time, the company itself, as an AI user, also has demand signals and can independently assess progress at the capability frontier, and research breakthroughs led him to judge that Gemini 5.0 will be significantly stronger than Argon. He also mentioned that Google is signing long-term contracts with committed revenue, so it is planning its business portfolio and investing in a responsible and disciplined manner.

Quoted postAlex Heath · @alexeheath

Sundar Pichai on Gemini 4 and the AI compute crunch Even Google can’t get enough compute. I sit down with Google and Alphabet CEO @sundarpichai shortly after the wider release of Gemini 4 Argon to discuss its early reception, where he still wants the model to improve, and what it takes to keep Google at the frontier of AI. Pichai explains the recent leadership shakeup at Google DeepMind and why building frontier models has become an “all-consuming” job. We also get into Google’s answer to the personal agent race, how he predicts Gemini will do more work on your behalf, and whether agents will finally bring Search, Gmail, and Gemini together. I ask him about Google turning free cash flow negative, the risks in the AI infrastructure buildout, and how he plans for demand years before knowing what the models will be capable of. He explains the signals giving him confidence to keep spending and how he uses AI agents to follow model development and compute usage inside Google. Timestamps: 00:00 Gemini Argon launch reception 01:49 Model access and safety 06:20 The AI model race 08:35 Agentic breakthroughs and 2027 11:34 Recursive self-improvement 15:41 DeepMind leadership changes 19:52 Google's AI compute investments 25:20 Is AI spending sustainable? 33:18 How Sundar uses AI agents 35:35 Gemini's personal agent plans 38:37 How AI agents change the web 43:52 Google's space data centers 47:38 Waymo's expansion strategy 50:26 Wing's drone delivery plans 51:36 AI safety and regulation 55:58 Google's quantum computing bet YouTube: https://t.co/1Cc25LitmB Spotify: https://t.co/XnD4pEEsh8 Substack: https://t.co/0SrnNGKqlm Apple Podcasts: https://t.co/8UKvyXTlFF This episode's presenting sponsor is @hyperagentapp. Thanks to the show’s premiere sponsors: @Atlassian, @meetgranola, and @mercury.

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

TThomas Wolf10:46Opinion

AI score61

Thomas Wolf comments on Recursive's new model Flint: quantizing lm_head to 4 bits still comes close to bf16

Hugging Face co-founder Thomas Wolf said that his AGI standard remains AI winning a Nobel Prize, but the AI researchers at Recursive quantized lm_head—which is usually advised to be kept at high precision—to 4 bits, and the result differs from bf16 by only 0.7 percentage points, which impressed him. He also mentioned that the model was released as open weights on Hugging Face, and congratulated Recursive. The quoted original Recursive post introduced that Flint is its first model, positioned as a fast local coding model, built by the automated research system Eureka Machine, reaching 125+ tokens/sec on a MacBook Pro and 150+ on RTX Spark, and built from scratch in under 4 months.

Quoted postRichard Socher · @RichardSocher

Introducing Flint, @Recursive_SI’s first model: a fast local coding model built with our automated research system, the Eureka Machine. • Best-in-class fast local coding • Up to 2x faster than comparable models: 125+ tokens/sec on a MacBook Pro, 150+ on RTX Spark • Built from scratch in under 4 months Private, unmetered intelligence on your own machine. Open weights available now on @huggingface, run it locally on your favorite framework or try it for free on @opencode or @pidotdev.

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

EEthan Mollick10:45Opinion

AI score70

Ethan Mollick: AI attack-defense acceleration is creating a gap with human response systems

Ethan Mollick reposted a post summarizing multiple security incidents of the day, calling it a visible sign of the phenomenon that people in the AI circle have been warning about since the release of Mythos. He believes that attack and defense are accelerating at non-human speed and interacting with human systems that need to respond. The reposted post listed events including two critical PHP vulnerabilities that allow unauthenticated remote code execution, a GitLab vulnerability with a CVSS score of 10.0, LiteLLM 1.107.0 being hijacked to implant a credential stealer and a remote access trojan, Railway suspending all deployments, and TiDB Cloud being breached.

Quoted post⁵⁄₉ · @fiveoutofnine

So basically, today PHP got two critical bugs that let anyone run code on your server without even logging in, which is bad, and then the details leaked before the patch was even out, which is worse, and it works on WordPress and MediaWiki and Magento and Joomla and Laravel, and Censys says like 1.5 million servers are running a version with no fix, so the NSA was like "take your system offline," and the FBI was like "take your system offline," and CISA and the UK and CERT-EU were all like "take your system offline," and then the Prime Minister of Japan, the actual Prime Minister, posted at 5:55 in the morning telling 48.8 million people to take their system offline, so everyone took their system offline, like the Dutch cybersecurity agency took its own website offline, and the Dutch parliament went offline, and the Port of Rotterdam went offline, and Finland and France and Denmark started pulling government sites, and Wikipedia went down, and https://t.co/QlOXb3COq3 went down, and Asana and GOG and Something Awful went down, and the University of Vienna's webmail went down for some reason, and that's just PHP. Because GitLab also dropped a CVSS 10.0, which is the highest score they give out, for a bug that's been there since 2014, so that's every self-hosted GitLab for twelve years, and then LiteLLM 1.107.0 got hijacked and shipped a credential stealer and a remote access trojan in every single wheel and the Docker images, including latest, and you didn't even have to import it, just installing it was enough, and it's the second time this year, and then Railway paused every deployment on the entire platform because someone got into the environment variables, so now everyone has to rotate every secret they've ever stored there, and TiDB Cloud said someone was walking around inside with stolen credentials for basically a month and an extortion group says they took 10.6 terabytes, and Zeabur got breached for the second time in two months and is just shutting down, like the whole company, and Node.js patched remote code execution and a TLS bypass, and CISA says Plex is being actively exploited, and someone took over 300 Tor exit relays to steal SSH passwords, and SafePay ransomware has hit over 600 critical infrastructure orgs, and India's biggest bank filed a data loss incident, and Japan has been in an official state of emergency in cyberspace since October 9, which honestly explains the 5:55 AM post. Oh, and eMudhra, a certificate authority that's issued over 60 million certificates, whose entire job is proving websites are who they say they are, was supposed to be using private keys that are random numbers over a hundred digits long, and one of them was 4. Just 4. And the others were like 66 and 237. Anyway, it's Wednesday.

#English text is being prepared.#Safety / alignment