Skip to contentSkip to stories

Updated

All AI news

Showing low-relevance items too. Hide low-relevance items

Oct 9

Oct 9Fri
  1. MarkTechPostNewsAI score44

    Google Research RRSI Guide: Mastering Self-Improving AI Agents

    AIMarkTechPost publishes a hands-on tutorial implementing RRSI (Regularized Recursive Self-Improvement), a method that lets an LLM agent revise its own harness around a frozen model. The full loop drafts edits with Claude Opus on Vertex AI and scores them in Docker benchmarks, but the edit-selection rules are plain Python that the tutorial runs in a simulated environment with a calibrated noise band.

  2. indigoXAI score28

    AI can build features, but defining requirements and design remains the gap

    AICurrent AI can quickly implement or replicate features, but clearly defining requirements and describing design is still missing, and the author expects this gap to persist. As requirements grow more abstract, humans may only specify goals and check results while agents handle implementation, leaving the software's logic layer as model-generated tokens.

  3. MiniMax Design (H3)OfficialAI score22

    MiniMax H3 Person Remover LoRA erases people from video

    AIA LoRA for MiniMax H3 removes a person from video by tracking them with SAM 3.1 and generating the replacement background in overlapping windows. Users supply the original video and a clean version of its first frame.

  4. X.PINXAI score46

    Manus parent Butterfly Effect raises over $500M led by Boyu Capital

    AIManus parent Butterfly Effect announced funding of over $500M led by Boyu Capital and IDG, with Tencent, HSG and ZhenFund returning, its first disclosed raise since resuming independent operations. The reported $4B post-money valuation was not confirmed. Manus also launched version 2.0 and the personal agent Cue on September 29, and is building a China-focused product team and partnerships with domestic model developers.

    Image from @thexpin's post
  5. X.PINXAI score46

    Seed preprint finds DeepSeek V4 long-context retrieval varies by position

    AIA Seed team preprint reports "phase sensitivity" in DeepSeek V4 and V4.1-Flash, where identical information becomes harder to retrieve depending on its position within compressed KV-cache blocks. The compression reduces memory and attention costs, but long-context retrieval accuracy varied by up to 40 percentage points across positions. The authors note that average benchmark scores can hide these recurring weak spots, though the findings concern retrieval specifically rather than all model behavior.

    Image from @thexpin's post
  6. LeiphoneNewsAI score42

    Doubao Work adds Canvas feature and Doubao 2.1 Lite model

    AIDoubao Work has added a Canvas feature for complex creative tasks, placing materials, design plans and outputs on one infinite canvas where users can keep editing text, colors and layout after images are generated. The update also integrates the lightweight Doubao 2.1 Lite model, aimed at everyday Q&A, document writing, spreadsheets and PPT creation, with optimized response speed and usage consumption.

  7. LeiphoneNewsAI score40

    TRAE Merges TraeWork and TraeCode Into a Full-Chain Development Platform

    AITRAE announced on October 9 that it has merged TraeWork and TraeCode into a single platform offering Agent mode and IDE mode with seamless switching between them. The upgraded product covers desktop, web, and mobile, letting users start tasks on a computer, check progress on mobile, and continue development back on desktop.

  8. GeekParkNewsAI score62

    Krinwave raises 400 million yuan to bring brain imaging ultrasound to AI

    AIGeekPark reports that Krinwave, a Shenzhen company, has completed a new 400 million yuan round with XVC and Sequoia China as new investors. The company says its low-frequency ultrasound captures brain structures through the skull, which conventional high-frequency devices cannot, and it aims to supply hardware and its kOS software platform to brain-computer interface and NeuroAI groups.

  9. GeekParkNewsAI score47

    Ten Days With Today AI, a Domestic Personal AI Assistant That Connects Chinese Apps

    AIToday AI, built by Teambition founder Qi Junyuan, launched its China version on September 24 and connects to Feishu, DingTalk, Tencent Docs and email. The author found it proactively sends morning and evening briefings and handles a single chat window across tasks, but struggled with misjudging task weight and gave confident yet wrong mod-installation instructions that cost an hour of testing.

  10. Simon WillisonBlogAI score14

    ttok 1.0 switches default tokenizer to GPT-5 family and GPT-6

    AISimon Willison released ttok 1.0, changing its default tokenizer from GPT-4 to a newer GPT-5/GPT-6 tokenizer after finding the old default was outdated. OpenAI has not confirmed that GPT-6 shares the GPT-5 tokenizer, but William Liu's experiment found all seven GPT models tested reported 44,794 tokens on a 31-fixture corpus.

  11. ModelScopeOfficialAI score63

    Google releases EmbeddingGemma 2, a lightweight multimodal embedding model for on-device search

    AIGoogle released EmbeddingGemma 2, a 740M-parameter multimodal embedding model under Apache 2.0 for private, on-device search and retrieval. It maps text, code, images, video, and audio into one shared space and reports a 9.92-point gain over EmbeddingGemma 1 on MTEB Code. The post lists about 191MB active RAM for quantized text-only weights and about 567MB for the full multimodal model on a Pixel 11 Pro.

    Video from @ModelScope2022's post
  12. Jeremy HowardXAI score8

    Jeremy Howard Says He Was Barred From Criticizing OpenAI

    AIJeremy Howard posted a sardonic remark that he was not allowed to say he is quite unhappy with much of what OpenAI does. The post quotes Tomek Korbak's similar statement and his relief at being permitted to say it.

  13. Neroitech Inventions (NITI)XAI score40

    Sui Agent Pass proposes bounded, enforceable limits on AI agent spending

    AIEvan Cheng, CEO of Mysten Labs, presented the Sui Agent Pass at Sui Basecamp in Singapore as a way to bound AI agent authority over money. Users would define allowed actions, assets, recipients, and permission duration, with the system itself enforcing those limits so that losses stop at a predefined cap even if an agent is compromised. The article argues that the real challenge is making such limits impossible to bypass when failures occur.

  14. PitchforkXAI score29

    Grimes announces AI-inspired album Psy Opera

    AIMusician Grimes has announced a new album titled Psy Opera, described as AI-inspired. The announcement was shared via a Pitchfork article linking to the first song from the record.

  15. QbitAINewsAI score64

    Tsinghua-linked VPP2 world action model tops RoboDojo simulation leaderboard

    AIStar Motion Era's VPP2, a world action model, ranked first on the RoboDojo simulation leaderboard with a 32.26% average success rate and 39.26 average score. The article attributes gains to staged training that separates video prediction from action learning, and reports a 58.5% zero-shot success rate on a real ALOHA dual-arm robot versus 40% for π0.5. The code is open source on GitHub.

  16. QbitAINewsAI score44

    Sharpa unveils D01 humanoid robot, W02 dexterous hand, and AE01 haptic glove at IROS

    AISharpa launched D01, a fully self-developed humanoid robot with electronic skin covering the whole body and tactile coverage of the upper body, sensing forces from 0.1 to 20N at 100Hz. It also unveiled the W02 dexterous hand, which has 21 active degrees of freedom, about 30% smaller than the W01, and the AE01 exoskeleton data glove with 22 encoders for teleoperation and data collection.

  17. Nace AIXAI score22

    NDI 1.0 document processing model launches for coding agents at 90% lower cost

    AINACE introduces NDI 1.0, a document processing model for coding agents that it says is 90% cheaper and ranks first on the Parse Index. The company says it offers native MCP, SDK, and CLI integration for Claude Code, Codex, Hermes, OpenClaw, and PI, and supports 50 languages. NACE also states the model was trained on over 15M financial files and offers $25 in free API credits to developers.

    Video from @NaceAI's post
  18. Ethan MollickXAI score34

    Gemini 2.5 models rated comparable to doctors in urgent care advice

    AIIn an urgent care study, physicians rated advice from the older Gemini 2.5 Pro and Gemini 2.5 Flash, which lacked access to patient medical records, as similar in quality to doctors' advice. No safety issues were identified. The author notes that models have improved significantly since.

    Image from @emollick's post
  19. 雷电芽衣XAI score24

    ELY Office Suite, an early Rust and GPUI native office suite, is released for macOS

    AIELY Office Suite is an open-source native office suite built with Rust and GPUI, now in an early test build for macOS, with Linux and Windows planned next week. It includes Word, PPT, Excel, PDF, Markdown, and Git client components, using LibreOffice's LOK for the office kernel, MuPDF for PDF, and MarkText for Markdown rendering. The author warns the build still has many bugs and should not be used in production.

    Video from @ZacharyZhang's post
  20. indigoXAI score30

    Guo Yu on ByteDance's neural-network roots and Vibe Coding's hidden cost

    AIIn a podcast, former ByteDance engineer Guo Yu says that in 2015 he first saw a company whose products were driven by neural networks that even engineers could not fully explain. He also argues that after AI took over coding, one person now carries the decision load of an entire former team, which has driven him to burnout.

  21. Arena.aiOfficialAI score38

    Mistral Large 4 ranks in Agent Arena top 15 at -6.6% net score

    AIMistral Large 4, a preview model from Mistral AI, ranks #43 overall in Agent Arena with a -6.6% net improvement score across more than 5,000 real-world agentic sessions. That is 11 rankings above its predecessor, Mistral Medium 3.5 (-12.60%), and places it in the top 15 labs, the only European lab there. Open weights are expected at the end of October, and at its current score the model would rank #13 among open models.

    Image from @arena's post
  22. X.PINXAI score49

    Apple's HomeHub ships only after LLMs finally improved Siri

    AIAccording to a source familiar with the project, Apple's homeOS was finished years ago, but the HomeHub was held back until large language models made Siri good enough to serve as its voice-driven interface. The hardware team reportedly refused to sign off on a device whose main interface was Siri, given its long-running poor performance. HomeHub is slated for an Oct 13 unveiling alongside a smart-home push with LG.

    Image from @thexpin's post

Oct 8

Oct 8Thu
  1. TechNode · AINewsAI score42

    XPENG names robotaxi service YOYO and opens invite-only public testing

    AIXPENG has named its robotaxi business XPENG YOYO and launched a ride-hailing mini program that lets invited members of the public test the service. The company says its first production robotaxi, based on the flagship GX model, rolled off the line in May 2026, and it has completed more than 2,000 internal test rides in Guangzhou. XPENG says YOYO uses four in-house Turing AI chips delivering 3,000 TOPS, a second-generation VLA model, and a vision-based approach without high-definition maps or LiDAR.

  2. Kling AIOfficialAI score14

    Kling AI hosts LA Tech Week panel on AI video production

    AIKling AI will host a panel at LA Tech Week 2026 on October 14, 2026, from 4:00 to 7:00 PM in Santa Monica, with speakers from Kling AI, Oxen AI, and Obsidian Studio. The discussion will cover how AI video is moving from model capabilities and platform integration into enterprise workflows and professional content production, following the launch of Kling 4.0. Attendees can RSVP via the event link and stay for networking afterward.

    Image from @Kling_ai's post
  3. QwenOfficialAI score22

    Free week of Qwen3.8-Max, Qwen3.8-Flash, and Wan3.0 on GMI Cloud

    AIQwen3.8-Max, Qwen3.8-Flash, and Wan3.0 are available free for a week on GMI Cloud, which is extending the offer by seven days and raising rate limits across all three models. GMI Cloud is also running a contest where three winners each receive $200 cash plus $200 in GMI credits for the most creative, most challenging, or most effort-driven projects built with Qwen or Wan.

  4. ClaudeOfficialAI score46

    Anthropic pauses Claude Startups Team and API credit offers amid demand

    AIAnthropic has paused the Claude Team plan and $1,000 API credit offers for its Claude Startups program after underestimating demand, with hundreds of thousands of applicants. Claimed offers will remain in accounts, but some approved applicants who had not yet claimed their offers will lose access as applications are re-reviewed. Startup Stack and Applied AI office hours remain available to accepted members.

  5. meng shaoXAI score43

    Unsloth integrates Microsoft's mxc sandbox for Windows AI agent isolation

    AIUnsloth has integrated Microsoft's open-source mxc sandboxing system into Windows as an OS-level sandbox for isolating AI agent code execution. Its High mode provides real operating-system isolation that confines tool calls to specified directories, while its Low mode adds language-level checks that block dangerous commands and shell escapes. Both modes also strip secret environment variables and enforce resource limits such as 8GB memory and 600-second CPU time.

    Image from @shao__meng's post
  6. meng shaoXAI score41

    Matt Pocock shares a framework for matching AI coding workflows to change size

    AIMatt Pocock recommends matching AI coding agent workflow weight to change size: one-shot small diffs, start medium-to-large changes with /grill-with-docs for requirements clarification, and escalate to /wayfinder for mapping and tickets only when planning becomes complex. He warns against starting with /wayfinder, since a simpler-than-expected solution can leave the generated map and tickets unnecessary.

    Image from @shao__meng's post
  7. meng shaoXAI score62

    Stanford CS146S Week 3 covers Agent Skills and CLI for coding agents

    AIStanford's CS146S course, taught by Mihail Eric, has published its Week 3 materials on Agent Skills and CLI. The lecture covers how SKILL.md files and scripts encode workflows, and it lists practical advice such as keeping each skill focused, mining one's own transcripts for skill ideas, and writing descriptions that name real trigger phrases.

    Image from @shao__meng's post