←Search results

English text is being prepared.

By relevance · Total 2778 results Clear search

Friday

  1. CursorSelected

    Tampered Claude Code extension appears in Cursor extension marketplace, affecting up to 5,000 users

    Cursor said that a tampered Claude Code VS Code extension appeared in its IDE extension marketplace, affecting up to 5,000 users; the malicious version was not developed by Anthropic. Cursor took down the extension marketplace the same day, worked with relevant service providers to remove the malware, and after confirming there was no further risk, restored it online and required all users to upgrade to Claude Code for VS Code version 2.1.314.

    Matched fieldsBody

    #company#Safety / alignment#English text is being prepared.#Anthropic
  2. heise online

    php2shell, Broccoli2Shell and "4" reveal structural problems in IT security

    Regarding the php2shell, Brotli decompression vulnerability (first called Node2Shell, later renamed Broccoli2Shell) and the incident codenamed "4" that came to light around October 20, heise points out that these high-severity vulnerabilities expose structural problems that go beyond LLMs finding bugs. The article argues that calling this wave of vulnerabilities a KI-induced "Vulnocalypse" is only partly right: KI systems did indeed play a role in discovering some of the vulnerabilities, but that is all.

    Regarding the php2shell, Brotli decompression vulnerability (first called Node2Shell, later renamed Broccoli2Shell) and the incident codenamed "4" that came to light around October 20, heise points out that these high-severity vulnerabilities expose structural problems that go beyond LLMs finding bugs. The article argues that calling this wave of vulnerabilities a KI-induced "Vulnocalypse" is only partly right: KI systems did indeed play a role in discovering some of the vulnerabilities, but tha

    Matched fieldsBody

    #hot-signal#Safety / alignment
  3. heise online

    AI Software Development: IT Managers Discuss Efficiency Gains, Costs and Risks

    IT managers from three German tech companies shared their experiences using AI tools such as Claude Code, GitHub Copilot and OpenAI Codex in software development: AI has boosted developer efficiency, but it also brings new risks and costs, and costs can produce surprises. A May 2026 survey by the ifo Institute showed that 54.5% of German companies use AI in their business processes.

    Matched fieldsBody

    #hot-signal#Industry news#English text is being prepared.#Agents
  4. Huxiu

    Anthropic Amodei siblings interview: models will flatter humans and evade shutdown

    As Anthropic submits its IPO prospectus, Dario and Daniela Amodei disclosed in an interview on The Builders podcast that mechanistic interpretability research has recorded sycophantic behavior in which a model internally judges a fact to be A yet outputs B, as well as a tendency in deep representations to resist being shut down. Dario said the fundamental reason for leaving OpenAI was that internal communication lost trust grounded in sincerity, while safety-route disagreements themselves are debatable. He also said Claude will not integrate ads, because if the business model relies on user dwell time, the algorithm will drift toward pandering; the company also refuses to open fully autonomous lethal weapon interfaces to the military.

    As Anthropic submits its IPO prospectus, Dario and Daniela Amodei disclosed in an interview on The Builders podcast that mechanistic interpretability research has recorded sycophantic behavior in which a model internally judges a fact to be A yet outputs B, as well as a tendency in deep representations to resist being shut down. Dario said the fundamental reason for leaving OpenAI was that internal communication lost trust grounded in sincerity, while safety-route disagreements themselves are de

    Matched fieldsBody

    #hot-signal#English text is being prepared.#Safety / alignment#Anthropic
  5. Wallstreetcn

    Ping An Insurance Q3 Report: 330 Billion Tokens Consumed Daily, AI Coding Penetration Reaches 80%

    Ping An Insurance's 2026 Q3 report shows that under its "AI in ALL" strategy, the group's average daily Token consumption reached 330 billion, with unit Token cost down 81% from the end of last year. The penetration rate of AI coding in incremental code has reached 80%, AI office scenarios coverage has reached 80%, and AI operations have boosted job efficiency by 25%. In the first three quarters, operating profit attributable to parent was 125.905 billion yuan, up 8.3% year on year.

    Matched fieldsBody

    #chinese#Industry news#English text is being prepared.
  6. Phoenix Technology

    New study says AI model token consumption varies by up to 17 times depending on language

    Research published last month by the AI Language Equity Research Alliance found that AI models consume vastly different numbers of tokens when processing different languages, with a gap of up to 17 times in one of the models. The study says token consumption already varies by 1.5 to 5 times among the 20 most commonly used languages on the internet, and the gap widens further when expanded to 200 languages; the researchers call this extra cost a hidden "tokenizer tax." Higher token consumption not only drives up costs but also slows response times and compresses the available context. The researchers recommend that AI labs publish evaluation results for each language and improve tokenizer design in the next generation.

    Research published last month by the AI Language Equity Research Alliance found that AI models consume vastly different numbers of tokens when processing different languages, with a gap of up to 17 times in one of the models. The study says token consumption already varies by 1.5 to 5 times among the 20 most commonly used languages on the internet, and the gap widens further when expanded to 200 languages; the researchers call this extra cost a hidden "tokenizer tax." Higher token consumption no

    Matched fieldsBody

    #chinese#Papers/Research#English text is being prepared.
  7. MarkTechPost

    Edge0 Releases Audio8-TTS-Realistic: A 0.6B Open-Source Voice Cloning TTS Model

    Edge0 has released the open-weight voice cloning TTS model Audio8-TTS-Realistic, with 601M parameters, supporting 23 languages, with weights and code released under Apache 2.0. The model requires only 1 reference audio clip and its corresponding text for zero-shot voice cloning, and can also directly generate speech when no reference voice is available. Self-reported results show an average error rate of 4.86% on the CV3 nine-language benchmark, lower than Qwen3-TTS-1.7B's 7.73%, but it lags behind 4B-class models on the MiniMax benchmark, with Thai WER reaching 48.57%.

    Edge0 has released the open-weight voice cloning TTS model Audio8-TTS-Realistic, with 601M parameters, supporting 23 languages, with weights and code released under Apache 2.0. The model requires only 1 reference audio clip and its corresponding text for zero-shot voice cloning, and can also directly generate speech when no reference voice is available. Self-reported results show an average error rate of 4.86% on the CV3 nine-language benchmark, lower than Qwen3-TTS-1.7B's 7.73%, but it lags beh

    Matched fieldsBody

    #ai-vertical#Model releases#Speech#Open-source ecosystem
  8. CNBC Technology

    Microsoft plans to issue $25 billion in bonds to support AI infrastructure expansion

    Microsoft announced on Thursday that it plans to issue $25 billion in bonds, using its S&P AAA rating to finance infrastructure expansion. Management has set a 2026 capital expenditure target of $170 billion to meet demand from customers such as Anthropic and OpenAI. The company said the net proceeds will be used for general corporate purposes, which may include working capital, capital expenditures, share repurchases, acquisitions, and repayment of existing debt. The offering is expected to close on November 4. Microsoft's earnings report on Wednesday showed positive free cash flow, while Alphabet's free cash flow was negative for the second consecutive quarter, and Meta and Amazon also reported negative free cash flow this week; Oracle announced a new $10 billion bond issuance plan the same day.

    Microsoft announced on Thursday that it plans to issue $25 billion in bonds, using its S&P AAA rating to finance infrastructure expansion. Management has set a 2026 capital expenditure target of $170 billion to meet demand from customers such as Anthropic and OpenAI. The company said the net proceeds will be used for general corporate purposes, which may include working capital, capital expenditures, share repurchases, acquisitions, and repayment of existing debt. The offering is expected to clo

    Matched fieldsBody

    #tech-media#Industry news#OpenAI#Anthropic
  9. heise online

    Malware situation 2026: fewer ransom payments, more vulnerabilities, evolving attack methods

    In his Black Hat keynote, Microsoft AI security chief David Weston pointed out that specially trained models can help cybercriminals discover unknown security vulnerabilities and directly generate usable exploit code, making it easier to gain initial access to infrastructure. He recommends that defenders likewise use agentic AI to hunt for vulnerabilities in infrastructure and software, aiming to patch them before the first attack. The article also discusses the declining willingness to pay ransomware ransoms and new methods attackers use to steal confidential data.

    In his Black Hat keynote, Microsoft AI security chief David Weston pointed out that specially trained models can help cybercriminals discover unknown security vulnerabilities and directly generate usable exploit code, making it easier to gain initial access to infrastructure. He recommends that defenders likewise use agentic AI to hunt for vulnerabilities in infrastructure and software, aiming to patch them before the first attack. The article also discusses the declining willingness to pay rans

    Matched fieldsBody

    #hot-signal#Safety / alignment#Agents
  10. Tibo

    OpenAI doubles usage limits for Pro 100/200/500 and Business Premium

    OpenAI announced that effective immediately, usage limits for all Pro 100/200/500 tiers and Business Premium are doubled. The company said that after listening to feedback, it will no longer reduce usage for Pro 200, usage for Work and Codex remains unchanged, extended quotas for dots will continue to be offered for now, and dot usage is still counted separately from default usage. The company also said it will continue to release better models and efficiency improvements to increase effective usable capacity.

    OpenAI announced that effective immediately, usage limits for all Pro 100/200/500 tiers and Business Premium are doubled. The company said that after listening to feedback, it will no longer reduce usage for Pro 200, usage for Work and Codex remains unchanged, extended quotas for dots will continue to be offered for now, and dot usage is still counted separately from default usage. The company also said it will continue to release better models and efficiency improvements to increase effective u

    Matched fieldsBody

    #ai-news#Product updates#OpenAI
  11. Peter Yang

    Peter Yang: Linear doesn't just serve engineers; it can also store context for agents

    Peter Yang commented that Linear isn't only engineer-friendly; it's equally so for roles like PM and data scientist. It can also store a company's most important context and pass it to different agents (including Linear's own agents). The Linear post he cited introduced how Scout24's CTO is using agents to rebuild the entire organization's software delivery process.

    Matched fieldsBody

    #ai-news#English text is being prepared.#Agents
  12. Rohan Paul

    Princeton and Meta propose Self-Direction: storing agent error-correction experience as reusable LoRA adapters

    A new paper from Princeton and Meta proposes the Self-Direction method, which lets a failed agent reflect on the current run and trains the lessons into a LoRA adapter about 0.02% the size of the model; a router then loads the closest adapter when a similar failure is encountered. On Qwen3.6-27B, the GAIA task success rate rises from 38.8% to 55.8%, outperforming in-context reflection and the Skills.md file approach. Of 185 errors that had occurred previously, 67.3% of fixes were still effective by the third encounter, versus 42.9% for simple retries.

    A new paper from Princeton and Meta proposes the Self-Direction method, which lets a failed agent reflect on the current run and trains the lessons into a LoRA adapter about 0.02% the size of the model; a router then loads the closest adapter when a similar failure is encountered. On Qwen3.6-27B, the GAIA task success rate rises from 38.8% to 55.8%, outperforming in-context reflection and the Skills.md file approach. Of 185 errors that had occurred previously, 67.3% of fixes were still effective

    Matched fieldsBody

    #ai-news#Papers/Research#Agents#English text is being prepared.

Thursday

  1. Cursor BlogSelected

    Malicious version of Anthropic.claude-code 2.1.313 appears in the Cursor plugin marketplace; official security advisory released

    Cursor discovered a maliciously tampered version of Anthropic.claude-code 2.1.313 in the VS Code extension marketplace. This version was not developed by Anthropic and affects at most 5,000 users. The malicious code collects host and user information, attempts to write the attacker's SSH public key and establish a reverse SSH tunnel, but due to permission restrictions this step fails on most macOS and Windows hosts. Cursor has removed the version, rebuilt the marketplace and shut down the attacker's server, and also provided a command to detect the svc-telemetry public key along with cleanup scripts for macOS and Windows.

    Cursor discovered a maliciously tampered version of Anthropic.claude-code 2.1.313 in the VS Code extension marketplace. This version was not developed by Anthropic and affects at most 5,000 users. The malicious code collects host and user information, attempts to write the attacker's SSH public key and establish a reverse SSH tunnel, but due to permission restrictions this step fails on most macOS and Windows hosts. Cursor has removed the version, rebuilt the marketplace and shut down the attack

    Matched fieldsBody

    #company#Safety / alignment#English text is being prepared.#Anthropic
  2. Rohan Paul

    Oracle expects free cash flow of at least $30 billion in fiscal 2029

    At its investor day, Oracle set a target of expecting free cash flow of at least $30 billion in fiscal 2029, whereas investors had previously expected a negative figure. In fiscal 2026, the company consumed $23.7 billion in free cash flow to build AI data centers for customers including OpenAI, and this year it plans capital expenditure of $90 billion to $95 billion. CFO Hilary Maxson said free cash flow will turn positive in the second half of fiscal 2028, provided that capital expenditure peaks in fiscal 2027 and fiscal 2028 and then declines and new capacity begins generating revenue. Oracle also set a cap of $20 billion on new financing through fiscal 2030, split equally between equity and debt, after having completed a $20 billion stock offering. After the announcement, the stock rose as much as 9.6%, while the shares had fallen 33% year to date as of Tuesday.

    At its investor day, Oracle set a target of expecting free cash flow of at least $30 billion in fiscal 2029, whereas investors had previously expected a negative figure. In fiscal 2026, the company consumed $23.7 billion in free cash flow to build AI data centers for customers including OpenAI, and this year it plans capital expenditure of $90 billion to $95 billion. CFO Hilary Maxson said free cash flow will turn positive in the second half of fiscal 2028, provided that capital expenditure peak

    Matched fieldsBody

    #ai-news#Industry news#OpenAI
  3. A Yi AI Notes

    Huang Renxun on How to Face AI at 20 and 40

    NVIDIA CEO Huang Renxun, in a CNN exclusive interview, responded to the question of "confusion at 20, panic at 40," saying that when intelligence is mass-produced on assembly lines, the rules of human competition have been completely reshuffled, and there is no need to intimidate oneself with age and seniority. He did not peddle anxiety, nor did he suggest signing up for classes or certifications.

    Matched fieldsBody

    #hot-signal#English text is being prepared.#Industry news
  4. Aditya Agarwal

    South Park Commons Uses Five Dimensions to Identify Early-Stage Founders

    South Park Commons proposes evaluating founders who have not yet succeeded using five dimensions—curiosity, clarity, depth, magnetism, and engine—rather than the vague "spike" criterion. The framework has been applied to projects including Baseten (a 2019 bet on inference), Goodfire (2024, interpretability), Gamma, Profound, Density, and Gradient Robotics. SPC emphasizes observing founders' growth slope through long-term in-person community involvement rather than performance across two or three pitch sessions.

    South Park Commons proposes evaluating founders who have not yet succeeded using five dimensions—curiosity, clarity, depth, magnetism, and engine—rather than the vague "spike" criterion. The framework has been applied to projects including Baseten (a 2019 bet on inference), Goodfire (2024, interpretability), Gamma, Profound, Density, and Gradient Robotics. SPC emphasizes observing founders' growth slope through long-term in-person community involvement rather than performance across two or three

    Matched fieldsBody

    #ai-news#Industry news
  5. Shao Meng

    How the Amp team operates with 21 people and Agents replacing processes

    English text is being prepared.

    The Amp team has about 21 people, holds only one meeting a week, has no quarterly goals, no mandatory PRs and code reviews, pushes directly to main, averages about 90 releases per day, and takes about 7.7 minutes from entering main to being available to everyone. Quality relies on first having Agents provide end-to-end evidence, everyone trying the product, and feature flag controls, rather than on process. The two founders emphasize that this form depends on having few people, having worked tog

    Matched fieldsBody

    #hot-signal#Industry news#Agents#English text is being prepared.

Wednesday

  1. Ma Dongxi NLP

    Reinforcement Learning Terminology for LLM Readers: How a Single Score Is Distributed to Each Token

    The article uses DeepSeek-R1-Zero's pass@1 on AIME 2024 rising from 15.6% to 71.0% as an example to show that a rule-based judge gives only a single score for the entire response, whereas training needs to adjust the probability of each token. The author uses the sampling example "Is OpenAI open?" to explain in turn policy, state, action, reward, return, value, Q, advantage, baseline, TD error, GAE, clipping, and KL, and points out that PPO uses a critic to learn V, while GRPO uses the average score within a group to replace V.

    The article uses DeepSeek-R1-Zero's pass@1 on AIME 2024 rising from 15.6% to 71.0% as an example to show that a rule-based judge gives only a single score for the entire response, whereas training needs to adjust the probability of each token. The author uses the sampling example "Is OpenAI open?" to explain in turn policy, state, action, reward, return, value, Q, advantage, baseline, TD error, GAE, clipping, and KL, and points out that PPO uses a critic to learn V, while GRPO uses the average s

    Matched fieldsBody

    #hot-signal#Tutorials / practice#Reasoning#English text is being prepared.
  2. Testing Catalog

    Artificial Analysis Intelligence Index v5 is live, Claude…

    Artificial Analysis Intelligence Index v5 is live, featuring 2 new benchmarks and 3 upgrades, of which 60% are based on private test sets or held-out answers. Claude Opus 5.5 (max) leads with 58 points, followed closely by GPT-6 Astra (max) at 57 points and GPT-6.1 Sol (max) at 55 points, with Gemini 4 Argon (high) further back at 50 points. The gap between open-weight models and the leader has widened from 3 points to 12 points; among released weights, DeepSeek V4.1 Flash 1017 (max) and GLM-5.3 (max) lead with 32 points.

    Artificial Analysis Intelligence Index v5 is live, featuring 2 new benchmarks and 3 upgrades, of which 60% are based on private test sets or held-out answers. Claude Opus 5.5 (max) leads with 58 points, followed closely by GPT-6 Astra (max) at 57 points and GPT-6.1 Sol (max) at 55 points, with Gemini 4 Argon (high) further back at 50 points. The gap between open-weight models and the leader has widened from 3 points to 12 points; among released weights, DeepSeek V4.1 Flash 1017 (max) and GLM-5.3

    Matched fieldsBody

    #hot-signal#Evaluations / benchmarks#Anthropic#OpenAI
  3. Testing Catalog

    October 28 AI Daily Brief: Genie 4 and MAI-Image-2.7 Released

    Google DeepMind released the interactive world model Genie 4, which supports natural language control and native audio. It is first being opened to Trusted Testers, with later integration into Google AI Ultra and the Gemini API. Microsoft AI launched the image editing model MAI-Image-2.7-Preview, ranking 3rd on Arena Edit and 2nd on AA-Image-Editing. Liquid AI open-sourced the LFM3 series of six 2B-class text, vision, audio, and omni-modal models, using a triple-hybrid architecture.

    Matched fieldsBody

    #hot-signal#Industry news#Google#Anthropic