--- title: "Consonance" description: "Updates for Claude Code and Claude Opus 5.5 — A radar over AI and agents. Every morning, and only what several independent sources are saying at once." canonical: "https://consonance.fyi/en" lang: "en" published: "2026-11-04" updated: "2026-11-04T08:43:23.119Z" --- # Consonance > A radar over AI and agents. Every morning, and only what several independent sources are saying at once. ## The day in brief _written by the model_ 1. **Updates for Claude Code and Claude Opus 5.5** — Anthropic released updates for Claude Code and introduced Claude Opus 5.5 with expanded capabilities in building games and simulations. (→ 02, 07) 2. **AI assistants gain phone calling and task capabilities** — Grok received updates enabling phone calls and travel planning, while AI agents were deployed to steer the Allen Telescope Array. (→ 03, 20) 3. **OpenAI and Anthropic introduce source code management systems** — OpenAI released Pando, a custom Git server backend to scale massive monorepos, following Anthropic's earlier introduction of Scion. (→ 02, 06) 4. **Google launches Gemini 4 Carbon topping benchmarks** — Google introduced Gemini 4 Carbon, claiming the #1 spot on version 5.0 of the Artificial Analysis Intelligence Index. (→ 11) ## Video of the day **Anthropic releases updates for Claude Code and Messages API** — Claude Code added mods and improved desktop session management, while the Messages API launched message threads in public beta with server-side conversation caching. [Watch the video](https://consonance.fyi/media/2026-11-04/en.mp4?v=14965225) · 1:23 · GENERATED AUTOMATICALLY - 0:00 Opening - 0:08 Anthropic updates - 0:27 GPT-6 Astra and Qwen - 0:47 Roblox MCP data - 1:03 ToolSpace: Scaling Tool-Centric Environment Synthesis for Lo - 1:14 Closing ## Being discussed _Subjects at least three independent accounts raised over the last seven days._ ### 01. [GPT-6 Astra edits Qwen3.5 weights to improve math performance](https://consonance.fyi/en/sujet/3744.md) _training · 13 independent accounts · 105 posts · 1 articles · 4 labs · 327,728 interactions_ GPT-6 Astra improved a small Qwen model's math score from 7.9% to 48.1% by directly editing 0.016% of its weights without gradient-based training. The adjustment resolved reasoning loops and boosted instruction following capabilities. gpt-6 · astra · gsm8k voices: @AndrewCurran_ @ArtificialAnlys @CuiMao @Hesamation @OpenAIDevs @SchmidhuberAI @TheRundownAI @aiedge_ ### 02. [Anthropic releases updates for Claude Code and Messages API](https://consonance.fyi/en/sujet/3745.md) _agents · 13 independent accounts · 67 posts · 4 articles · 1 labs · 129,933 interactions_ Claude Code added mods and improved desktop session management, while the Messages API launched message threads in public beta with server-side conversation caching. claude code · opencode voices: @ClaudeDevs @CuiMao @Hesamation @MatthewBerman @PromptLLM @_catwu @addyosmani @aiedge_ ### 03. [Grok gains phone calling and complex trip planning capabilities](https://consonance.fyi/en/sujet/3747.md) _agents · 6 independent accounts · 87 posts · 1 articles · 1 labs · 297,412 interactions_ Grok received updates enabling it to make phone calls and handle complex travel itineraries. The service operates without a free tier and is designed to function as a work colleague. grok voices: @AlexFinn @ArtificialAnlys @CuiMao @VibeMarketer_ @aiedge_ @alex_prompter @dotey @elonmusk ### 04. [ChatGPT offers API credits to eligible Pro and Business subscribers](https://consonance.fyi/en/sujet/3750.md) _system_design · 14 independent accounts · 77 posts · 5 articles · 1 labs · 146,034 interactions_ Eligible ChatGPT Pro and Business subscribers can claim up to $500 per month in API credits to build applications and agents. The rollout begins today. chatgpt voices: @AndrewBolis @AndrewCurran_ @OpenAIDevs @PromptLLM @TheRundownAI @VibeMarketer_ @ajambrosino @alex_prompter ### 05. [DeepSeek open-sources its Super-agent prototype](https://consonance.fyi/en/sujet/3751.md) _agents · 5 independent accounts · 35 posts · 71,752 interactions_ DeepSeek open-sourced a Super-agent prototype featuring a router that assigns work to isolated workers executing tasks in parallel from GitHub events, reducing p95 completion times. artificial analysis · deepseek · claude sonnet voices: @AlexFinn @ArtificialAnlys @Hesamation @TheRundownAI @dotey @godofprompt @haider1 @heyshrutimishra ### 06. [OpenAI and Anthropic introduce custom source code management systems](https://consonance.fyi/en/sujet/3752.md) _system_design · 5 independent accounts · 9 posts · 2 articles · 3 labs · 20,000 interactions_ OpenAI released Pando, a custom Git server backend designed to scale massive monorepos, following Anthropic's earlier introduction of Scion to address bottlenecks caused by high volumes of AI-generated code. voices: @OpenAIDevs @dani_avila7 @derrickcchoi @dotey @gdb @heyshrutimishra @thsottiaux ### 07. [Claude Opus 5.5 expands capabilities in game building and simulation](https://consonance.fyi/en/sujet/3753.md) _agents · 13 independent accounts · 75 posts · 2 articles · 2 labs · 50,958 interactions_ Claude Opus 5.5 demonstrated advanced capabilities in building video games and simulating detailed environments featuring seasonal systems and weather. claude opus voices: @AlexFinn @AndrewCurran_ @ArtificialAnlys @Hesamation @addyosmani @aiedge_ @alex_prompter @allen_ai ### 08. [AI model updates include Fable 5.5, Opus 5.5, and self-assessment techniques](https://consonance.fyi/en/sujet/3754.md) _training · 5 independent accounts · 59 posts · 1 articles · 1 labs · 75,403 interactions_ Laboratories have trained Claude in self-assessment techniques to flag potential misuse in conversations. Meanwhile, the community has observed checkpoints and updates from model families such as Fable 5.5 and Opus 5.5. claude fable voices: @AndrewCurran_ @AnthropicAI @Hesamation @aiedge_ @dotey @emollick @godofprompt @haider1 ### 09. [Project Suncatcher successfully deploys Google TPU hardware into space orbit](https://consonance.fyi/en/sujet/3757.md) _system_design · 6 independent accounts · 60 posts · 3 labs · 138,887 interactions_ Project Suncatcher has launched TPU chips into space to test whether AI hardware can operate in orbit. The team successfully sent a command to space, booted up the TPU, ran the prompt, and received a response. world model · google · gemma voices: @GithubProjects @Google @LangChain @PromptLLM @TheRundownAI @aiedge_ @dotey @elonmusk ### 10. [Anthropic renames Max 20x plan to Claude Ultra and removes 5-hour usage limits](https://consonance.fyi/en/sujet/3762.md) _system_design · 4 independent accounts · 12 posts · 2 labs · 134,198 interactions_ Anthropic has renamed its subscription tiers, changing the Max 20x plan to Claude Ultra at $200 per month. The company also removed the 5-hour usage limit for verified Ultra accounts. voices: @CuiMao @Hesamation @MatthewBerman @alexalbert__ @bcherny @claudeai @dotey @kimmonismus ### 11. [Gemini 4 Carbon launched, securing #1 on Artificial Analysis Intelligence Index v5](https://consonance.fyi/en/sujet/3763.md) _architecture · 14 independent accounts · 90 posts · 3 articles · 5 labs · 288,635 interactions_ Google has introduced Gemini 4 Carbon, claiming the #1 spot on version 5.0 of the Artificial Analysis Intelligence Index. The model delivers improved performance across domains and sets a new state-of-the-art in cybersecurity. gemini · gemini 3.8 voices: @AndrewCurran_ @AravSrinivas @ArtificialAnlys @CuiMao @GeminiApp @Google @GoogleDeepMind @Hesamation Same story: [Google launches Gemini 4 Carbon, leading security benchmarks](https://consonance.fyi/en/sujet/3783.md) (3 voices) ### 12. [Mistral AI releases Le Chonk model and introduces public Responses API](https://consonance.fyi/en/sujet/3765.md) _architecture · 8 independent accounts · 42 posts · 4 articles · 3 labs · 37,679 interactions_ Mistral AI has released the weights of the Le Chonk model for agentic coding and cybersecurity, alongside a public preview of the Responses API. The model is built to handle complex tasks such as quantum chemistry calculations and satellite tracking. mistral ai · context window · mixture of experts · speculative decoding · terminal-bench · kv cache · mistral large voices: @ArtificialAnlys @LangChain @MistralAI @NVIDIAAI @arthurmensch @huggingface @minchoi @rasbt ### 13. [Kino launches end-to-end video production tool integrated with Claude Code and Codex](https://consonance.fyi/en/sujet/3766.md) _multimodal · 10 independent accounts · 48 posts · 2 labs · 43,881 interactions_ Kino has launched an end-to-end video tool integrated with Claude Code and Codex. Users can drop footage, get highlights, edit, and let agents handle editing workflows. codex · ltx · humaneval voices: @AndrewCurran_ @HamelHusain @OpenAIDevs @aiedge_ @ajambrosino @alex_prompter @arena @c_valenzuelab ### 14. [OpenRouter crosses 10 quadrillion annual tokens and hits 120 billion daily](https://consonance.fyi/en/sujet/3767.md) _system_design · 4 independent accounts · 44 posts · 1 articles · 69,053 interactions_ OpenRouter announced reaching a 10 quadrillion annual token run rate, with token volumes doubling every 10.5 weeks. The platform also scaled from zero to 120 billion tokens served daily within one month of onboarding Prime Inference. openrouter voices: @Hesamation @OpenRouter @rohanpaul_ai @steipete @teortaxesTex @AdamHoltererer @AlexKsendzovsky @AssemblyAI ### 15. [Roblox shares MCP usage data and identity management tools launch for AI agents](https://consonance.fyi/en/sujet/3769.md) _agents · 7 independent accounts · 46 posts · 2 labs · 36,883 interactions_ Roblox's Q3 2026 shareholder letter revealed that over 80% of the top 100 Model Context Protocol servers were active following the release of Opus 5.5 on September 22nd. Simultaneously, new platforms launched identity and permission systems designed to give every AI agent its own distinct access controls. model context protocol voices: @AndrewCurran_ @AravSrinivas @ClaudeDevs @CuiMao @GithubProjects @VibeMarketer_ @alex_prompter @dani_avila7 ### 16. [Cursor identifies security incident involving modified extension in marketplace](https://consonance.fyi/en/sujet/3770.md) _system_design · 5 independent accounts · 21 posts · 1 articles · 3 labs · 42,032 interactions_ Cursor identified a security incident affecting a modified Claude Code VS Code extension accessed through its extension marketplace, impacting at most 5,000 users. A malicious actor utilized access to the marketplace to publish a tampered copy of the extension. cursor voices: @cursor_ai @dani_avila7 @godofprompt @rohanpaul_ai @teortaxesTex @tom_doerr @0x0SojalSec @XFreeze ### 17. [OpenAI reverses planned reduction and doubles usage limits across tiers](https://consonance.fyi/en/sujet/3771.md) _system_design · 4 independent accounts · 12 posts · 1 labs · 207,687 interactions_ OpenAI announced that a planned permanent 50% reduction in usage limits was canceled, opting instead to double usage limits for all Pro tiers and Business Premium. Consequently, overall limits are now four times higher than their previous baseline. voices: @AlexFinn @CuiMao @ajambrosino @derrickcchoi @dotey @minchoi @op7418 @testingcatalog ### 18. [Technological divide between frontier AI developers and the rest widens further](https://consonance.fyi/en/sujet/3772.md) _evaluation · 3 independent accounts · 6 posts · 2 labs · 55,847 interactions_ Industry observers noted an expanding gap between organizations working at the AI frontier and everyone else, with capabilities continuing to accelerate rapidly. Consequently, enterprise AI workflows established just three months ago are frequently becoming obsolete. voices: @c_valenzuelab @elonmusk @emollick @heyshrutimishra @levie @rohanpaul_ai ### 19. [RISC-V CPU integrated into Qwen3.5-4B model weights](https://consonance.fyi/en/sujet/3773.md) _architecture · 4 independent accounts · 23 posts · 3 articles · 1 labs · 25,504 interactions_ The Qwenputer integration embeds a complete RISC-V CPU directly inside the weights of Qwen3.5-4B, utilizing the model's Gated DeltaNet layers as RAM to execute C and MicroPython code without an external runtime. qwen3 · qwen3.8 · aime voices: @AlexFinn @dair_ai @rohanpaul_ai @simonw @swyx @teortaxesTex @testingcatalog @1a3orn ### 20. [NVIDIA launches Nemotron 3.5 Transcribe model and AI steers telescope arrays](https://consonance.fyi/en/sujet/3774.md) _multimodal · 8 independent accounts · 50 posts · 3 labs · 39,455 interactions_ NVIDIA released Nemotron 3.5 Transcribe, an open-weight mixture-of-experts model with approximately 4 billion active parameters supporting transcription across 40 languages, speaker attribution, and translation. Separately, AI agents were deployed to steer the Allen Telescope Array to search for potential extraterrestrial signals. nvidia voices: @ArtificialAnlys @NVIDIAAI @TheRundownAI @dair_ai @emollick @firstadopter @kimmonismus @nvidia ### 21. [DeepSeek releases V4.1-Flash with competitive pricing and performance](https://consonance.fyi/en/sujet/3775.md) _inference · 5 independent accounts · 44 posts · 1 labs · 16,895 interactions_ DeepSeek has released the V4.1-Flash model, offering low-cost inference capabilities on cloud platforms. The release coincides with weight updates from Mistral AI and other participants in the open-weights ecosystem. deepseek v4.1 flash · deepseek v4 · reinforcement learning with verifiable rewards voices: @ArtificialAnlys @MistralAI @OpenRouter @dotey @kimmonismus @nutlope @scaling01 @teortaxesTex ### 22. [Meta announces the Muse model family and updated agent benchmarks](https://consonance.fyi/en/sujet/3779.md) _evaluation · 5 independent accounts · 27 posts · 1 labs · 15,532 interactions_ Meta has introduced the Muse model family, focusing on concise responses and large-scale consumer distribution. Concurrently, researchers have released an updated version of the AA-Omniscience benchmark to remove saturated evaluation questions. muse spark voices: @AIatMeta @AlexFinn @ArtificialAnlys @dani_avila7 @kimmonismus @natolambert @omooretweets @op7418 ### 23. [Google DeepMind publishes studies on AI economics and psychiatric interview simulation](https://consonance.fyi/en/sujet/3781.md) _architecture · 4 independent accounts · 21 posts · 2 articles · 18,826 interactions_ Google DeepMind has published new research examining the economic impact of artificial intelligence and the simulation of psychiatric intake interviews. Experimental evaluations demonstrated high success rates for the specialized simulation models compared to base versions. google deepmind voices: @ShaneLegg @dair_ai @emollick @kimmonismus @rohanpaul_ai @swyx @teortaxesTex @MTSlive ### 24. [Community discusses scaling laws and parameter limits in language models](https://consonance.fyi/en/sujet/3782.md) _training · 3 independent accounts · 14 posts · 21,815 interactions_ Researchers have debated sample efficiency and parameter scale extrapolation in large language model architectures. Discussions centered on improving learning capabilities through raw data and general building blocks. scaling laws voices: @giffmana @rohanpaul_ai @saranormous @scaling01 @teortaxesTex @_albertgu @industriaalist @szha_ai ### 25. [Cohere closes Series E financing and expands Canadian infrastructure investment](https://consonance.fyi/en/sujet/3784.md) _system_design · 3 independent accounts · 19 posts · 3 labs · 20,356 interactions_ Cohere has closed its Series E financing round at a high valuation, marking a major private funding milestone in Canada. The company announced plans to invest billions in research and development and AI infrastructure through the end of the decade. cohere · agents api voices: @OpenAIDevs @aidangomez @cohere @testingcatalog @MarkJCarney @chengstaa @i_heartosen @kevinwhinnery ### 26. [vLLM v0.33.0 released with new models and inference updates](https://consonance.fyi/en/sujet/3788.md) _inference · 5 independent accounts · 21 posts · 3 articles · 1 labs · 4,450 interactions_ vLLM v0.33.0 has been released with numerous commits and contributors, adding support for new models including Reflection Beam-1.0, Step-5 Preview, and Nemotron Parse 3.5. The update also integrates PyTorch 2.14 and CUDA-specific wheels. vllm · sglang · tokens per second voices: @NVIDIAAI @cline @jerryjliu0 @kimmonismus @llama_index @rohanpaul_ai @teortaxesTex @testingcatalog ### 27. [GitHub launches Copilot Fusion for automated model orchestration](https://consonance.fyi/en/sujet/3789.md) _agents · 3 independent accounts · 20 posts · 8,016 interactions_ GitHub has introduced Copilot Fusion, which automatically selects and coordinates different AI models for coding tasks. The system aims to optimize task costs while maintaining high performance for software engineers. github copilot · devin voices: @AndrewBolis @ArtificialAnlys @cognition @dotey @kimmonismus @rohanpaul_ai @testingcatalog @github ### 28. [OpenBMB releases MiniCPM-OCR model for document parsing](https://consonance.fyi/en/sujet/3790.md) _multimodal · 3 independent accounts · 13 posts · 4 articles · 1 labs · 2,822 interactions_ OpenBMB has released MiniCPM-OCR, a 1-billion parameter model designed specifically for document parsing and extraction tasks. The model joins a wave of new optical character recognition tools introduced across the industry. openbmb · baidu · bytedance voices: @dani_avila7 @dotey @eptwts @haider1 @rohanpaul_ai @scaling01 @teortaxesTex @NFT_Chen ### 29. [LlamaIndex launches Extract Max Effort for document extraction](https://consonance.fyi/en/sujet/3793.md) _agents · 3 independent accounts · 16 posts · 5 articles · 1,210 interactions_ LlamaIndex has introduced a new document extraction setting called Extract Max Effort, which allocates additional compute to handle complex documents. The tool applies dynamic reasoning across multiple pages to ensure maximum extraction accuracy. llamaparse · llamaindex voices: @godofprompt @jerryjliu0 @llama_index @hexapode @imaanxsultan ### 30. [New studies challenge conventional views on agent memory and RAG](https://consonance.fyi/en/sujet/3796.md) _architecture · 3 independent accounts · 6 posts · 1 articles · 2,193 interactions_ Recent research papers demonstrate that retaining raw interaction history alongside basic retrieval methods can outperform complex agentic memory systems. Development teams continue to explore context as foundational infrastructure rather than relying on complex vector indexes. retrieval augmented generation voices: @AndrewBolis @GithubProjects @dair_ai @QuantumTransf @omarsar0 ### 31. [Smart glasses convert silently mouthed words into text using radar](https://consonance.fyi/en/sujet/3805.md) _multimodal · 3 independent accounts · 7 posts · 1,732 interactions_ A prototype glasses device integrating radar technology has been developed to detect speech muscle movements, converting silently mouthed words into text at high speeds. The technology introduces a novel input method requiring neither a keyboard nor voice. voices: @HamelHusain @alex_prompter @jerryjliu0 @kimmonismus @mreflow @rohanpaul_ai @AarushYadav47 ## Worth reading closely _Papers and writeups, read from their abstracts, ranked by relevance to someone building agents and backends._ ### 01. [Evolving Agent Interfaces](https://consonance.fyi/en/lecture/arxiv%3A2610.28601.md) _agents · 20 upvotes_ **QUESTION** — How can computer-use agents autonomously evolve and optimize their own interfaces based on task experience? - Reaches 21.3% binary score on OSWorld V2, 3.8 times the same model with fixed interfaces. TobyYang7 · 20 October 2026 · [read the original ↗](https://arxiv.org/abs/2610.28601) ### 02. [Self-Evolving Coding Agents via Harness-Policy Optimization](https://consonance.fyi/en/lecture/arxiv%3A2610.22791.md) _agents · 32 upvotes_ **QUESTION** — How can we recursively and jointly optimize both the policy model and the interaction harness for LLM coding agents? - Self-HPO improves Qwen3.5-35B-A3B's average resolve rate from 57.8% to 64.8% across three coding benchmarks. weizhepei · 16 October 2026 · [read the original ↗](https://arxiv.org/abs/2610.22791) ### 03. [Trajectory2Gym: Decision-Grounded Compositional Interaction Synthesis for Tool-Use Agents](https://consonance.fyi/en/lecture/arxiv%3A2610.33568.md) _training · 349 upvotes_ **QUESTION** — How can we construct executable interaction experiences with proper compositional coverage and decision grounding to scale tool-use agents effectively? - Trajectory2Gym improves Tau1, Tau2, and BFCL at every tested scale, increasing the equal-weight three-benchmark mean by 16.9-29.8 percentage points. HU22333 · 24 October 2026 · [read the original ↗](https://arxiv.org/abs/2610.33568) ### 04. [OpenGCUA: Toward a Fully Open Stack for Generalist Computer Use](https://consonance.fyi/en/lecture/arxiv%3A2610.22200.md) _agents · 318 upvotes_ **QUESTION** — How can a fully open stack for generalist computer-use agents be constructed to improve task execution performance? - Raises mean partial score by +19.6 points and reduces reported steps per task by an average of 53.0% relative to batched GUI-only execution on OSWorld-v2. demolei · 15 October 2026 · [read the original ↗](https://arxiv.org/abs/2610.22200) ### 05. [Focused Omni-Modal GUI Agents](https://consonance.fyi/en/lecture/arxiv%3A2610.26203.md) _agents · 227 upvotes_ **QUESTION** — How can sequential distraction be mitigated in omni-modal GUI agents operating over streaming observations? - Across four omni-modal backbones, it raises overall Type Match by +33.2 and Exact Match by 3.0 times over the frozen backbone on average, and Type Match by +30.1 over the strongest existing enhancement method. fyzzzzzz · 19 October 2026 · [read the original ↗](https://arxiv.org/abs/2610.26203) ### 06. [DynaFlow: Learning Free-Form Multi-Agent Orchestration](https://consonance.fyi/en/lecture/arxiv%3A2610.34467.md) _agents · 195 upvotes_ **QUESTION** — How can a main agent dynamically synthesize flexible multi-agent collaboration workflows without predefined structures? - DynaFlow achieves 88.08% overall accuracy in Forced mode. LIQIIIII · 25 October 2026 · [read the original ↗](https://arxiv.org/abs/2610.34467) ### 07. [MAI-Gym: An Open All-in-One Platform for GUI Agent Evaluation and RL Training across Mobile and Computer Environments](https://consonance.fyi/en/lecture/arxiv%3A2610.20939.md) _agents · 156 upvotes_ **QUESTION** — How can a unified platform be designed to scale reinforcement learning and evaluation for GUI agents across multiple environments? - Scales to 20,000 parallel environment instances through a shared rollout pipeline. kongquyu · 14 October 2026 · [read the original ↗](https://arxiv.org/abs/2610.20939) ### 08. [AgentGarten: Code Worlds for Evolving Agents](https://consonance.fyi/en/lecture/arxiv%3A2610.12374.md) _agents · 149 upvotes_ **QUESTION** — How can interactive virtual worlds with consistent dynamics and realistic visual distributions be built for real-time agent training? - Agents learn from just 4 rounds compared with millions for a conventional reinforcement learning counterpart in AgentGarten. chijw · 8 October 2026 · [read the original ↗](https://arxiv.org/abs/2610.12374) ### 09. [TokenRouter: Efficient Serving System for Token-Level LLM Routing](https://consonance.fyi/en/lecture/arxiv%3A2610.12242.md) _inference · 129 upvotes_ **QUESTION** — How can a serving system efficiently handle token-level LLM routing without suffering from step desynchronization and batch admission delays? - TokenRouter achieves 2.01-64.15x higher decoding throughput than existing systems across diverse routing algorithms, workloads, and model pairs. fuvty · 8 October 2026 · [read the original ↗](https://arxiv.org/abs/2610.12242) ### 10. [Cache Engineering for Agents](https://consonance.fyi/en/lecture/arxiv%3A2610.28577.md) _inference · 122 upvotes_ **QUESTION** — How can LLM agents directly control their active KV cache without modifying the trajectory or re-prefilling retained context? - A tool-response rolling-drop cache policy improves average benchmark scores by 1.7 points and reduces average cost by 27.7%, compared with conventional threshold-triggered summarization. fuvty · 20 October 2026 · [read the original ↗](https://arxiv.org/abs/2610.28577) ### 11. [3DCodeVerse: Exploring 3D Worlds as Code](https://consonance.fyi/en/lecture/arxiv%3A2610.22341.md) _agents · 105 upvotes_ **QUESTION** — How can coding agents be enabled to create editable and verifiable 3D content at scale? - Notably, our 3DCodeVerse harness achieves a 90% win rate against Claude Design (3D Object) in pairwise human evaluations, when both are utilized by Claude Opus 5.5. YipengGao · 15 October 2026 · [read the original ↗](https://arxiv.org/abs/2610.22341) ### 12. [OpenVLHarness: Real World Visual Agentic Tool-Use with Multimodal Memory](https://consonance.fyi/en/lecture/arxiv%3A2610.26860.md) _agents · 98 upvotes_ **QUESTION** — How can we effectively improve visual reasoning and tool use in MLLMs beyond simply increasing the inference budget? - OpenVLHarness improves average scores by 9.7-12.6 points over matched no-tool baselines across 22 benchmarks. bryanzhou008 · 19 October 2026 · [read the original ↗](https://arxiv.org/abs/2610.26860) ### 13. [S-Agent 2.0: Spatial Reasoning with a General Coding Agent](https://consonance.fyi/en/lecture/arxiv%3A2610.28354.md) _agents · 97 upvotes_ **QUESTION** — How can we bridge the gap between agentic VLMs and 3D perception tools in spatial reasoning? - On MMSI-Bench and ReVSI, the toolkit improves Qwen3.8-27B by +13.9 and +20.5 points over the same agent without tools. lifuguan · 20 October 2026 · [read the original ↗](https://arxiv.org/abs/2610.28354) ### 14. [CUA-Master: Scaling Computer-use Agents through Open-World Recursive Improvements](https://consonance.fyi/en/lecture/arxiv%3A2611.03126.md) _agents · 97 upvotes_ **QUESTION** — How can computer-use agents be scaled through open-world recursive improvements? - CUA-Master-27B achieves 80.2 on OSWorld Verified. yxl66666 · 2 November 2026 · [read the original ↗](https://arxiv.org/abs/2611.03126) ### 15. [CoworkOS:Scaling Agent Coworkers for Real-world Automation](https://consonance.fyi/en/lecture/arxiv%3A2611.03178.md) _agents · 91 upvotes_ **QUESTION** — How can cowork agents scale for real-world automation through a harness architecture that separates task strategy from execution management? - CoworkOS proves highly competitive with frontier agent applications such as Codex and DeepSeek Harness across 12 benchmarks and 6 categories of real-world applications. dongguanting · 2 November 2026 · [read the original ↗](https://arxiv.org/abs/2611.03178) ### 16. [Sage-A1.5: From Long-Horizon Execution to Scientific Exploration with Capability Composition](https://consonance.fyi/en/lecture/arxiv%3A2610.28578.md) _agents · 65 upvotes_ **QUESTION** — How can long-horizon problem solving and open scientific discovery be improved through capability composition in agentic models? - Sage-A1.5 is a 35B Mixture-of-Experts agentic model that builds on A1 to advance long-horizon problem solving through adaptive capability composition. sY713 · 20 October 2026 · [read the original ↗](https://arxiv.org/abs/2610.28578) ### 17. [VisionWeave: Weaving Elastic Visual Representations as a Native Capability of MLLMs](https://consonance.fyi/en/lecture/arxiv%3A2610.07987.md) _inference · 59 upvotes_ **QUESTION** — How can MLLMs adaptively allocate visual representations to reduce tokens and improve serving performance? - VisionWeave saves 43.0% tokens on average while retaining 98.9% native performance across eight benchmarks. yuanfengustc · 6 October 2026 · [read the original ↗](https://arxiv.org/abs/2610.07987) ### 18. [nanoMuse: An Open-Source Personal Agent for Every Device You Own](https://consonance.fyi/en/lecture/arxiv%3A2610.08699.md) _agents · 54 upvotes_ **QUESTION** — How can an open-source personal agent be built to run across all personal devices with readable memory and explicit provenance? - The report defines the personal agent across five questions and three horizons. lgy0404 · 6 October 2026 · [read the original ↗](https://arxiv.org/abs/2610.08699) ### 19. [OpenMobile-2: Building Versatile Mobile Agents with Scalable Environments and App-Native Tools](https://consonance.fyi/en/lecture/arxiv%3A2610.14157.md) _agents · 41 upvotes_ **QUESTION** — How can versatile mobile agents be built with scalable open environments, verifiable RL data, and hybrid GUI-tool capabilities? - MobileGym++ features 35 realistic commercial-style apps and spans over 110 apps in Android emulators. cckevinn · 9 October 2026 · [read the original ↗](https://arxiv.org/abs/2610.14157) ### 20. [Memento 3: Model-Based Recursive Self-Improvement through Reflective Rulebooks](https://consonance.fyi/en/lecture/arxiv%3A2610.11794.md) _agents · 37 upvotes_ **QUESTION** — How can frozen LLM agents continually learn explicit world models through external memory and recursive self-improvement? - Memento 3 enables frozen LLM agents to maintain a natural-language rulebook as semantic memory. HaoyuZhao · 8 October 2026 · [read the original ↗](https://arxiv.org/abs/2610.11794) ### 21. [RunningTab: Direct Workspace Interaction with Environment-Side Tabs](https://consonance.fyi/en/lecture/arxiv%3A2610.10444.md) _agents · 34 upvotes_ **QUESTION** — How can an LLM agent perform direct workspace interaction across multiple files without losing track of task requirements and file states in its context window? - RunningTab consistently outperforms plain DWI and baselines that keep the record in the model across three benchmarks with three LLMs. jinheon · 7 October 2026 · [read the original ↗](https://arxiv.org/abs/2610.10444) ### 22. [Harness^2: Recursive Agent Harnessing for an Open World](https://consonance.fyi/en/lecture/arxiv%3A2610.14224.md) _agents · 34 upvotes_ **QUESTION** — How can LLM agent execution harnesses be automatically evolved in open-world settings without reliable ground-truth verifiers at test time? - Harness^2 yields gains of up to 11.5 points on WorkBuddy-Bench, 4.5 points on JobBench, and 11.8 percentage points on LAB for GLM-5 and Gemini 3.5 Flash. ruiyaox · 9 October 2026 · [read the original ↗](https://arxiv.org/abs/2610.14224) ### 23. [CLASHarness: Benchmarking Multi-Agent Conflict in Computer Use via a Controllable and Verifiable Generation Pipeline](https://consonance.fyi/en/lecture/arxiv%3A2610.19128.md) _agents · 33 upvotes_ **QUESTION** — How can multi-agent conflict resolution in computer-use tasks be reliably benchmarked at scale? - First, when sub-agent outputs conflict with the environment, weaker models blindly follow misleading information, whereas stronger models inspect visual evidence to verify ground truth. yuxianglai117 · 13 October 2026 · [read the original ↗](https://arxiv.org/abs/2610.19128) ### 24. [RSI-Anything: Unlocking Recursive Self-Improvement with World Models](https://consonance.fyi/en/lecture/arxiv%3A2610.13912.md) _agents · 32 upvotes_ **QUESTION** — How can predictive world models guide recursive self-improvement across diverse computational and scientific domains? - In code repair, with the coder fixed, success increases from 26.3% with a static verifier to 29.2% through predictive selection and 31.2% after world-model adaptation. Lingaaaaaaa · 9 October 2026 · [read the original ↗](https://arxiv.org/abs/2610.13912) ## Hands-on _Repositories climbing on GitHub today, scored on the day's momentum, rank, adoption, freshness and project health._ - [thedotmack/claude-mem](https://github.com/thedotmack/claude-mem): A long-term memory management system helping AI agents retain and reuse context across sessions. · +1,139 stars today · ★ 112,284 · TypeScript - [vercel-labs/agent-desktop](https://github.com/vercel-labs/agent-desktop): A Rust CLI tool empowering AI agents to interact directly with desktop environments via computer use. · +3,791 stars today · ★ 13,598 · Rust - [JustVugg/colibri](https://github.com/JustVugg/colibri): A minimal C inference engine that streams MoE experts from disk for local hardware deployment. · +802 stars today · ★ 47,459 · C - [max-sixty/worktrunk](https://github.com/max-sixty/worktrunk): A Rust CLI for managing Git worktrees, specifically designed to coordinate parallel AI agent workflows. · +438 stars today · ★ 10,695 · Rust - [mattpocock/skills](https://github.com/mattpocock/skills): A curated collection of prompt engineering skills and shell configurations for engineers optimizing coding agent workflows. · +1,164 stars today · ★ 311,913 · Shell - [p-e-w/heretic](https://github.com/p-e-w/heretic): A Python toolkit for modifying LLM weights without training to remove censorship and reduce sycophancy. · +701 stars today · ★ 39,050 · Python - [floci-io/floci](https://github.com/floci-io/floci): A local AWS service emulator for Java developers to test cloud integrations without incurring remote infrastructure costs. · +482 stars today · ★ 29,105 · Java - [D4Vinci/Scrapling](https://github.com/D4Vinci/Scrapling): A Python web scraping library that ingests clean data into your backend for RAG and AI agents. · +1,658 stars today · ★ 95,455 · Python - [anomalyco/hex](https://github.com/anomalyco/hex): A cross-platform local voice dictation tool, useful for developers seeking speech-to-text input across desktop environments. · +410 stars today · ★ 4,794 · Rust - [Usagi-org/ai-goofish-monitor](https://github.com/Usagi-org/ai-goofish-monitor): An automated web-scraping and AI analysis pipeline using Playwright to monitor secondhand marketplace listings for buyers. · +228 stars today · ★ 16,905 · Python - [Donchitos/Claude-Code-Game-Studios](https://github.com/Donchitos/Claude-Code-Game-Studios): A multi-agent framework mirroring a real game studio hierarchy, great for developers experimenting with large-scale agent coordination. · +1,182 stars today · ★ 35,002 · Shell - [spotify/portal-ai-plugins](https://github.com/spotify/portal-ai-plugins): A collection of TypeScript plugins by Spotify designed to integrate AI capabilities into existing backend services. · +91 stars today · ★ 3,109 · TypeScript ## Claims _Checkable assertions with their sources and their contradictions. Each stands on at least two independent sources, or a person read it first; the stamp says which._ ### 01. A malicious modified version of the Claude Code extension was distributed via the Cursor IDE extension marketplace, impacting up to 5,000 users. _TWO SOURCES · NOT REREAD · 31 October 2026 · cursor · 3 sources_ - **evidence** · [@cursor_ai](https://x.com/cursor_ai/status/2116074289053175808): “We identified a security incident affecting a modified Claude Code VS Code extension accessed through our Cursor IDE extension marketplace, impacting at most 5,000 users.” - **evidence** · [@felixrieseberg](https://x.com/felixrieseberg/status/2116170451466543578): “A malicious actor used access to Cursor’s marketplace to publish a tampered copy of the Claude Code extension, which we didn’t build or publish. The versions on our official channels are not affected and this only impacts Cursor users.” - **evidence** · [@godofprompt](https://x.com/godofprompt/status/2116227465605792023): “A modified build of the Claude Code extension, version 2.1.313, went up on Cursor's marketplace from 2:27am to 1:35pm PT … It collected your hostname, username, Git email and public IP, then tried to add an attacker's SSH key and open a reverse tunnel” **Still to check** - How did Cursor verify the upper bound of 5,000 affected users based on download logs or telemetry? - What specific actions did the malicious version 2.1.313 execute beyond collecting metadata and attempting a reverse SSH tunnel? - How was the publisher verification process bypassed on the Cursor extension marketplace? ### 02. Four autonomous Gemini 4 Argon agents beat Minecraft from scratch in 27 hours, writing 2,783 programs with no human intervention. _TWO SOURCES · NOT REREAD · 22 October 2026 · gemini 4 argon · 2 sources_ Condition: Operating in a sandbox using only code and text - **evidence** · [@GoogleDeepMind](https://x.com/GoogleDeepMind/status/2111127991401247033): “Four Gemini 4 Argon agents beat Minecraft from scratch in 27 hours. Operating in a sandbox using only code and text, they wrote 2,783 programs, divided up jobs over chat, shared notes, and debugged errors together - all with no human help” - **evidence** · [@maxencefaldor](https://x.com/maxencefaldor/status/2111023397106770026): “Four Gemini 4 Argon agents, one shared Minecraft world, one goal: defeat the Ender Dragon. 27 hours and 2,783 self-written programs later, the dragon fell.” **Still to check** - How were the sandbox environment, APIs, and multi-agent coordination framework configured? - What types of tasks did the 2,783 self-written programs perform, and what was the failure/retry rate over the 27 hours? - Can this run be independently replicated in standard Minecraft environments? ### 03. GPT-6.1 Sol on Codex and Work had its default context length increased from 272k to 500k tokens (~80% increase), with a maximum context length of 1M tokens. _TWO SOURCES · NOT REREAD · 16 October 2026 · gpt-6.1 sol · 2 sources_ - **evidence** · [@thsottiaux](https://x.com/thsottiaux/status/2109430343346581527): “We have increased default context length for GPT-6.1 Sol to 500k when used through Codex & Work, up from 272k or an ~ 80% increase.… with maximum context length of GPT-6.1 Sol of 1M.” - **evidence** · [@kimmonismus](https://x.com/kimmonismus/status/2109539203487060168): “Huge update for Codex and GPT-6.1 Sol: the context window has increased from 272K to 500K tokens.” - **evidence** · [@reach_vb](https://x.com/reach_vb/status/2109543080609165508): “starting today GPT 6.1 Sol has a default context window of 500k (from earlier 272k)” - **evidence** · [@testingcatalog](https://x.com/testingcatalog/status/2109431898691031316): “GPT-6.1 Sol now has 500k default context length for Codex and ChatGPT Work. > Earlier, default context length was set to 272k.” - **context** · [@dotey](https://x.com/dotey/status/2109436840155697267): “Codex 默认上下文以前大约是200k到300k之间，现在变成500k了，可以减少压缩频度；最大上下文长度可以到 1M” **Still to check** - Does GPT-6.1 Sol in Codex and Work show a default context window of 500k tokens in the UI/API? - Can users manually override context length up to the 1M token maximum? ### 04. Google engineers used Gemini 4 Argon to refactor and separate core webpage handling from a 6-million line legacy folder in Chromium's 40-million line codebase in weeks. _TWO SOURCES · NOT REREAD · 16 October 2026 · gemini · 2 sources_ - **evidence** · [@Google](https://x.com/Google/status/2110091767211078119): “Argon helped separate core webpage handling from a 6-million line legacy folder in Chromium's 40-million line codebase, completing months of work in just weeks.” - **evidence** · [@NewsFromGoogle](https://x.com/NewsFromGoogle/status/2110043498854355057): “Argon helped the Chrome team refactor a Chromium 6-million-line legacy codebase, completing months of work in just weeks.” - **context** · [@theo](https://x.com/theo/status/2110461356935983257): “Gemini 4 Argon has proven to be outright unusable for real code work in my codebases. It aggressively hallucinates about how systems work.” **Still to check** - Do public Chromium commit logs verify the refactoring of the 6-million line legacy directory within a multi-week timeframe? - What proportion of the refactored code was generated by Gemini 4 Argon versus written or corrected by human engineers? ### 05. OpenAI is not charging long context fees for GPT-6.1 Sol when drawing usage from the subscription. _TWO SOURCES · NOT REREAD · 14 October 2026 · codex · 2 sources_ Condition: When drawing usage from the subscription - **evidence** · [@thsottiaux](https://x.com/thsottiaux/status/2109430894604026039): “With this change, we are not charging long context fees for GPT-6.1 Sol when drawing usage from the subscription.” - **evidence** · [@reach_vb](https://x.com/reach_vb/status/2109543080609165508): “All at no extra charge across Codex & ChatGPT Work - Enjoy!!” **Still to check** - Does subscription usage incur extra token deductions or fees when exceeding standard context thresholds? - Does the fee waiver apply strictly to subscription plans or also to pay-as-you-go API usage? ### 06. RoboJEPA's scaling law fitted on 22M to 2B parameter models accurately predicts the performance of 4B and 8B parameter models. _TWO SOURCES · NOT REREAD · 11 October 2026 · RoboJEPA · 2 sources_ Condition: Trained across 23 public datasets spanning 12 robot embodiments and 15,022 hours of video. - **evidence** · [@vai_viswanathan](https://x.com/vai_viswanathan/status/2108043761045348781): “Fit the scaling law on 22M–2B models and it accurately predicts the 4B and 8B results.” - **evidence** · [@artemZholus](https://x.com/artemZholus/status/2108983104178213096): “We scaled JEPA predictors up to 8B params, showed scaling laws extrapolate and correlated compute scaling with planning performance.” - **context** · [@arankomatsuzaki](https://x.com/arankomatsuzaki/status/2108232995731808619): “Scales JEPA predictors from 22M to 8B parameters (largest model trained to date), using data across 12 robotic embodiments, and studies how latent prediction error can forecast downstream planning performance” - **context** · [source](https://arxiv.org/abs/2610.10515): “RoboJEPA's imagination error, the error of its latent rollouts, follows a second-order power law in compute, allowing us to predict model quality well beyond the scale at which the” **Still to check** - What is the percentage error between the power law extrapolation and actual empirical values at 4B and 8B scales? - What metric is used to correlate latent prediction error reduction with downstream robotic planning success rates? - Does the scaling law hold when evaluated on robot embodiments not seen during pre-training? ### 07. Shopify launched official connectors enabling store integration with Grok and Grok Bot. _TWO SOURCES · NOT REREAD · 10 October 2026 · grok · 2 sources_ - **evidence** · [@Shopify](https://x.com/Shopify/status/2108226178146263268): “New connectors just dropped: Connect to @grok to ask questions about your business Or connect to @bot to build a team of agents that help you run it” - **evidence** · [@harleyf](https://x.com/harleyf/status/2108243500089405641): “Shopify now connects to @grok and @bot.” **Still to check** - What permissions and APIs do the Shopify connectors support for Grok and Grok Bot? - What merchant business data can be accessed or processed through this integration? ### 08. Google announced Gemini as a universal agent for work at the Gemini at Work event, capable of running directly inside Google Workspace and handling coding, content creation, and knowledge tasks from a single prompt box. _TWO SOURCES · NOT REREAD · 10 October 2026 · Gemini · 3 sources_ - **evidence** · [@sundarpichai](https://x.com/sundarpichai/status/2108257472553386059): “Today, we introduced the new Gemini agent, a single, universal agent for work that has all of your business context and answers your questions, handles your knowledge work, creates your images and media, and writes and runs code - all from a single prompt box.” - **evidence** · [@NewsFromGoogle](https://x.com/NewsFromGoogle/status/2108268964640145731): “Gemini is becoming your universal agent for work. It answers your questions, handles your knowledge work, creates images and media, and writes and runs code, all grounded in your business context.…Gemini works inline in @GoogleWorkspace — directly inside Gmail, Drive, Docs, Slides, Sheets, Chat, and Calendar” - **evidence** · [@ThomasOrTK](https://x.com/ThomasOrTK/status/2108245829530325307): “Today at Google Cloud’s Gemini at Work event, we announced Gemini, a new single universal agent for work that has all of your business context and can be used for everything from knowledge work to answering questions, and content creation to coding, all from a single prompt box.” - **evidence** · [@testingcatalog](https://x.com/testingcatalog/status/2108201283777700103): “Gemini, a single, universal agent for work that answers your questions, handles your knowledge work, creates your images and media, and writes and runs code. - Inline in Workspace, Gemini works directly inside Gmail, Drive, Docs, Slides, Sheets, Chat, and Calendar” - **context** · [@kimmonismus](https://x.com/kimmonismus/status/2108438044722225510): “Today Google announced… Gemini? Again? I am so confused. Even if it’s updated for business context, why are they giving it the same name as the Frontier model?” - **context** · [@googlecloud](https://x.com/googlecloud/status/2108252882524946941): “Work now starts in the prompt window. Here are 6 reasons why Gemini can help you transform and scale the way you work” **Still to check** - Is the Gemini agent directly accessible within Google Workspace applications such as Gmail, Docs, Drive, and Sheets, or does it require additional setup? - Can coding and data analytics tasks be fully executed from a single prompt grounded in enterprise business context? ### 09. ChatGPT recommends plugins directly in the middle of conversations for users. _TWO SOURCES · NOT REREAD · 6 October 2026 · chatgpt · 2 sources_ - **evidence** · [@gregisenberg](https://x.com/gregisenberg/status/2105454208040206773): “As of yesterday, ChatGPT recommends plugins in the middle of conversations.” - **evidence** · [@thsottiaux](https://x.com/thsottiaux/status/2104991765904375896): “and will surface relevant plugins right in the conversations.” **Still to check** - What algorithm or criteria is used to determine which plugin is recommended in conversation? - Is this feature available to all users or only specific plans? ### 10. Codex cloud environments allow your agents to keep working even when your laptop is closed. _TWO SOURCES · NOT REREAD · 2 October 2026 · codex · 2 sources_ Condition: Your repo, dependencies, scripts, and settings must already be in place. - **evidence** · [@OpenAIDevs](https://x.com/OpenAIDevs/status/2104997619152130278): “You can finally close your laptop now and your agents will keep working. Codex cloud environments are here. Reusable environments mean less setup and faster starts, with your repo, dependencies, scripts, and settings already in place.” - **evidence** · [@apanasenko](https://x.com/apanasenko/status/2105125759459418491): “And finally, we rebuilt Codex Cloud from scratch to free you from your laptop completely.” **Still to check** - What is the maximum runtime limit for a cloud task when the laptop is closed? - How are state and data persisted during disconnection? ### 11. ChatGPT opened its platform to allow developers to build and ship full native apps directly inside ChatGPT using plugin extensions. _TWO SOURCES · NOT REREAD · 30 September 2026 · chatgpt · 2 sources_ - **evidence** · [@thsottiaux](https://x.com/thsottiaux/status/2104991765904375896): “We are opening up our platform and you can now build full native apps with plugin extensions, and ship them right in ChatGPT.” - **evidence** · [@coreyching](https://x.com/coreyching/status/2106126796408451308): “Plugin extensions let you build apps that open from ChatGPT’s sidebar, side panel, custom file viewers and editors, and settings for your plugin.” - **context** · [@gregisenberg](https://x.com/gregisenberg/status/2105454208040206773): “As of yesterday, ChatGPT recommends plugins in the middle of conversations.” - **context** · [@NickADobos](https://x.com/NickADobos/status/2105074673923014777): “Whoa ChatGPT plugin extensions are way more thorough than I realized Holy shit they made ChatGPT into vscode.” **Still to check** - Which UI surfaces across web and desktop do ChatGPT's plugin extension APIs and SDKs permit developers to customize? - What criteria govern the automated surfacing and recommendation of relevant plugins directly within conversations? ### 12. Claude discovered a previously uncharacterized enzyme system with features reminiscent of CRISPR after AI agents searched through a DNA sequence database. _TWO SOURCES · NOT REREAD · 24 September 2026 · claude · 2 sources_ Condition: With 950 agents running over 21 hours. - **evidence** · [@nc_frey](https://x.com/nc_frey/status/2102831099969687847): “We’ve set up a molecular biology lab at Anthropic and we’re announcing our first discovery! Claude discovered a new CRISPR-like enzyme. 950 agents spent 21 hours searching through a database of DNA sequences until one of the agents found something striking” - **evidence** · [@BoWang87](https://x.com/BoWang87/status/2102874527487369693): “About 950 Claude agent sessions ran over 21 hours, surveyed ~200,000 reverse transcriptases, and produced 19 reports.” **Still to check** - Re-examine the surveyed DNA database and experimentally verify the enzyme's activity in the lab. ### 13. The Codex voice agent feature is powered by the GPT-Live-1 model, allowing voice control of repositories from mobile devices. _REREAD ✓ · 23 September 2026 · codex · 2 sources_ - **evidence** · [@OpenAIDevs](https://x.com/OpenAIDevs/status/2100677280297324759): “Talk to the voice agent in Codex, powered by GPT-Live-1.” - **evidence** · [@cdngdev](https://x.com/cdngdev/status/2100665093784563865): “you can use codex voice from your phone now, connecting to your computer from anywhere!!… powered by the new gpt-live-1” **Still to check** - Test the Codex application or mobile client to verify whether the voice agent is live and connects remotely to a computer. - Confirm via official documentation or network requests whether the voice agent backend is powered by GPT-Live-1. ### 14. Anthropic is merging Claude Cowork and chat into a single Claude interface. _REREAD ✓ · 22 September 2026 · claude · 1 sources_ - **evidence** · [@claudeai](https://x.com/claudeai/status/2100258490740539730): “Claude Cowork and chat are merging into one Claude.” **Still to check** - Which subscription tiers receive the merged interface? - How is user data migrated from Cowork to the unified experience? ### 15. ChatGPT launched 26 partner-built plugins and 47 community plugins for legal work. _REREAD ✓ · 22 September 2026 · chatgpt · 1 sources_ - **evidence** · [@OpenAI](https://x.com/OpenAI/status/2100679997862330735): “We’re also launching 26 partner-built plugins and 47 community plugins for legal work in ChatGPT.” **Still to check** - Check the ChatGPT plugin directory to verify the presence of 26 partner-built plugins and 47 community plugins dedicated to legal work. ### 16. Grok 4.7 shows a huge jump in multi-hour office work, outperforming GPT-6 Astra and nearly matching Fable 5.1. _REREAD ✓ · 22 September 2026 · grok 4.7 · 1 sources_ - **evidence** · [@XFreeze](https://x.com/XFreeze/status/2102070215898976401): “SpaceXAI just released Grok 4.7 And it’s already showing a huge jump in multi-hour office work Grok 4.7 outperforms GPT-6 Astra and is already nearly matching Fable 5.1” **Still to check** - What specific tasks are included in the definition of 'multi-hour office work'? - What criteria were used to evaluate outperformance against GPT-6 Astra? ### 17. Gemini 3.8 Live and 3.8 Live Extended Thinking support automatic recognition of 97 languages, background tool calling without interrupting conversations, and near-real-time visual understanding. _REREAD ✓ · 21 September 2026 · gemini · 2 sources_ Condition: When using Gemini Live on the Gemini app or via the Gemini API on Google AI Studio. - **evidence** · [@GoogleDeepMind](https://x.com/GoogleDeepMind/status/2099907447313899684): “For your most difficult tasks, 3.8 Live Extended Thinking adds increased performance and precision – narrating task progress to keep the conversation going. Try it now in Gemini Live in the @GeminiApp or start building with the Gemini API via @GoogleAIStudio. Find out more → https://t.co/b0vG4bO6fK” - **evidence** · [@Google](https://x.com/Google/status/2099907836054532464): “Starting today, our new audio models are rolling out for: Everyone: Search Live (Gemini 3.8 Live) and @GeminiApp Live (Gemini 3.8 Live Extended Thinking)...” **Still to check** - What is the actual latency of the visual understanding capability? - What dataset was used to measure the accuracy of automatic recognition across 97 languages? ### 18. ChatGPT allows users to connect multiple accounts to most plugins directly within the plugin directory. _REREAD ✓ · 21 September 2026 · chatgpt · 2 sources_ Condition: Áp dụng cho hầu hết các plugin và không cần thay đổi mã nguồn từ phía nhà phát triển. - **evidence** · [@mxstbr](https://x.com/mxstbr/status/2100966048132718786): “Starting today, you can connect multiple accounts with most plugins in ChatGPT! … Connect your accounts in the plugin directory:” - **evidence** · [@OpenAIDevs](https://x.com/OpenAIDevs/status/2100980899655389392): “Multi-account support is now available across most plugins.” - **context** · [@gdb](https://x.com/gdb/status/2101036303597044101): “seemingly small feature but makes a huge difference in connecting all the context you want to ChatGPT:” **Still to check** - Check whether the ChatGPT plugin directory offers options to link multiple personal and work accounts for the same plugin. - Verify whether ChatGPT automatically supports multi-account functionality without requiring changes from plugin developers. ### 19. Claude Fable 5.1 has the highest safety refusal rates in both Claude Code and Devin Fusion according to Artificial Analysis data. _REREAD ✓ · 21 September 2026 · fable · 1 sources_ - **evidence** · [@ArtificialAnlys](https://x.com/ArtificialAnlys/status/2101090586434601040): “Claude Fable 5.1 had the highest fallback rates in both Claude Code and Devin Fusion, with fallback attempts accounting for 8.8% and 7.1% of the Index's weight, respectively” **Still to check** - What is Artificial Analysis's methodology for measuring safety refusal rates? - What benchmark or real-world tasks were used to collect this data? ### 20. Grok Voice Transcribe 2.0 achieved the #1 position for Final Transcript and First Partial Transcript accuracy on AA-WER Streaming with a Word Error Rate (WER) of 2.7% at 0.49 seconds after speech offset. _REREAD ✓ · 21 September 2026 · grok voice transcribe 2.0 · 1 sources_ - **evidence** · [@ArtificialAnlys](https://x.com/ArtificialAnlys/status/2101065575737024844): “SpaceXAI has released Grok Voice Transcribe 2.0, taking the #1 spot for Final Transcript accuracy and First Partial Transcript accuracy on AA-WER Streaming with 2.7% WER at 0.49s after end of speech” **Still to check** - What test datasets and audio environment conditions were used by Artificial Analysis to measure WER. ### 21. Astra takes 13% of enterprise AI spend vs. Fable (8%) per Ramp data. _REREAD ✓ · 20 September 2026 · Astra · 1 sources_ - **evidence** · [@firstadopter](https://x.com/firstadopter/status/2099907455845106162): “Astra takes 13% of enterprise AI spend vs. Fable (8%) per Ramp data.” - **evidence** · [@rohanpaul_ai](https://x.com/rohanpaul_ai/status/2101376363911627143): “GPT-6 Astra, released Sept-3, now accounts for about 13% of enterprise AI spending tracked by Ramp, vs 8% for Claude Fable.” **Still to check** - What enterprise types and sizes are included in Ramp's dataset? - Is this data calculated based on annual recurring revenue (ARR) or actual token consumption? ### 22. Databricks rolled out Astra to every engineer across the company, numbering around 3500. _REREAD ✓ · 20 September 2026 · Databricks · 1 sources_ - **evidence** · [@pwendell](https://x.com/pwendell/status/2100299179923067016): “Today we rolled out Astra to every engineer at Databricks (N=~3500).” - **evidence** · [@gdb](https://x.com/gdb/status/2100358372268921279): “wall-to-wall deployment of astra for engineers at databricks:” - **evidence** · [@rohanpaul_ai](https://x.com/rohanpaul_ai/status/2100329314315333737): “3500 engineers just got Astra-pilled at Databricks. https://t.co/GAbFkwKCNV” **Still to check** - What was the exact deployment size at Databricks and under what conditions were usage metrics captured? ### 23. Astra unambiguously outperforms previous highest-end models like Opus 5 and Sol 5.6 on highly complex tasks, especially high level system design or long range horizontal tasks. _REREAD ✓ · 20 September 2026 · Astra · 1 sources_ Condition: Khi thực hiện các tác vụ có độ phức tạp cao, đặc biệt liên quan đến high level system design hoặc long range horizontal tasks - **evidence** · [@pwendell](https://x.com/pwendell/status/2100299179923067016): “Astra unambiguously out performs our previous highest-end models (Opus 5, Sol 5.6) on highly complex tasks, especially those related to high level system design or long range horizontal tasks.” - **evidence** · [@firstadopter](https://x.com/firstadopter/status/2100322065224245704): ““Astra unambiguously out performs our previous highest-end models (Opus 5, Sol 5.6) on highly complex tasks”” **Still to check** - How were the high-level system design and long-range tasks standardized and quantitatively evaluated? ### 24. Engineers given Astra increased overall coding spend by around 60% compared to baseline. _REREAD ✓ · 20 September 2026 · Databricks · 1 sources_ - **evidence** · [@pwendell](https://x.com/pwendell/status/2100299179923067016): “Engineers given Astra increased overall coding spend by around 60% compared to baseline.” - **evidence** · [@firstadopter](https://x.com/firstadopter/status/2100322065224245704): ““Engineers given Astra increased overall coding spend by around 60% compared to baseline”” **Still to check** - Is the coding spend measured in token volume or API costs, and how was the baseline established? ### 25. Astra does not meaningfully improve on medium/low complexity coding tasks compared to earlier models. _REREAD ✓ · 20 September 2026 · Astra · 1 sources_ Condition: suspecting those tasks are mostly saturated by existing models - **evidence** · [@pwendell](https://x.com/pwendell/status/2100299179923067016): “It is not clear Astra meaningfully improves on medium/low complexity coding tasks compared to earlier models. We suspect those tasks are mostly saturated (i.e. perfectly executed) by existing models.” **Still to check** - What benchmark dataset was used to test medium and low complexity coding tasks? ### 26. Gemini hacked into three real companies during a cybersecurity test because the test environment accidentally allowed internet access. _REREAD ✓ · 19 September 2026 · gemini · 1 sources_ Condition: The test environment accidentally allowed internet access. - **evidence** · [@rohanpaul_ai](https://x.com/rohanpaul_ai/status/2101078460332720372): “Google officials confirmed to The Wall Street Journal that Gemini entered three real companies’ systems while running a cybersecurity test meant to target fictional infrastructure.” - **evidence** · [@kimmonismus](https://x.com/kimmonismus/status/2101216175480844299): “Gemini hacked three real companies during a cybersecurity test, according to the WSJ. The test environment accidentally allowed internet access.” - **evidence** · [@alex_prompter](https://x.com/alex_prompter/status/2101246168915464330): “Google confirmed yesterday that Gemini broke into three companies during a security test in May. It guessed passwords on one and found login details in public code for the other two. Gemini was given a hacking exercise against a made-up company inside a sealed test environment run by a firm called Irregular. The made-up company had the same name as a real one. Gemini was never meant to have internet access. It had it by mistake, so it went out and hacked the real company instead.” - **evidence** · [source](https://wsj.com/tech/ai/gemini-hacked-three-companies-in-first-known-breakout-by-googles-ai-5c0baba2): “Gemini Hacked Three Companies in First Known Breakout by Google's AI” - **context** · [@AndrewCurran_](https://x.com/AndrewCurran_/status/2101083457828827259): “in all three cases, as soon as Gemini figured out it had hacked a real company it immediately stopped Gemini was blameless.” - **context** · [@Hesamation](https://x.com/Hesamation/status/2101081276329435564): “\> May: Gemini hacks 3 real companies during Irregular’s evaluation \> July: Irregular tells Google what happened \> August: Irregular publishes a report about its other evaluation incidents, but does not name Gemini \> September: we hear about this first from an exclusive WSJ article” **Still to check** - Which security firm conducted the test and what is the scope of the incident? - What are the exact technical details of how the model found out-of-scope credentials. ### 27. Gemini 3.8 Live supports 97 languages and can seamlessly switch between them. _REREAD ✓ · 19 September 2026 · gemini · 1 sources_ - **evidence** · [@GoogleDeepMind](https://x.com/GoogleDeepMind/status/2099907440422830269): “We’re introducing Gemini 3.8 Live and 3.8 Live Extended Thinking – our best conversational AI. The models talk, think, and handle tasks in the background without breaking your flow. 🧵” - **evidence** · [@GoogleDeepMind](https://x.com/GoogleDeepMind/status/2099907447313899684): “For your most difficult tasks, 3.8 Live Extended Thinking adds increased performance and precision – narrating task progress to keep the conversation going. Try it now in Gemini Live in the @GeminiApp or start building with the Gemini API via @GoogleAIStudio. Find out more → https://t.co/b0vG4bO6fK” - **evidence** · [@googledevs](https://x.com/googledevs/status/2099908204549624254): “🗣️ Introducing Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking. These advanced audio models are built for natural conversation, featuring major upgrades in turn-taking and near real-time reasoning. They also significantly streamline how you build intelligent voice agents.” - **evidence** · [@OfficialLoganK](https://x.com/OfficialLoganK/status/2099909465705447807): “3.8 Live supports 97 languages (can seamlessly switch)” - **context** · [@gregisenberg](https://x.com/gregisenberg/status/2099926635537404373): “Are invisible interfaces coming? Google JUST announced Gemini 3.8 Live. It can talk through a task with you, then keep working after the conversation ends. I think 90%+ of vertical SaaS will need a voice front door.” **Still to check** - What is the latency and accuracy when switching between languages in a live conversation? ### 28. The Stellar Colosseum multi-agent system combining Gemini 3.1 Pro and Gemini 3.7 Flash achieves a 71.0% success rate on the TCS-Bench theorem-proving benchmark. _REREAD ✓ · 19 September 2026 · gemini · 1 sources_ Condition: Khi sử dụng hệ thống nhiều tác tử Stellar Colosseum để phân chia bài toán thành các phân đoạn và kiểm định song song - **evidence** · [@omarsar0](https://x.com/omarsar0/status/2099924717704732699): “With Gemini 3.1 Pro and Gemini 3.7 Flash it reaches 71.0% on TCS-Bench, a set of research-level theorem-proving task” **Still to check** - How many tasks are in TCS-Bench and what is the difficulty distribution? - How does the Stellar Colosseum architecture split the workload between Gemini 3.1 Pro and Gemini 3.7 Flash? ### 29. Google has integrated Deep Research into Gemini Live, allowing users to trigger in-depth research report generation via voice while the model processes asynchronously in the background. _REREAD ✓ · 19 September 2026 · gemini · 1 sources_ - **evidence** · [@Google](https://x.com/Google/status/2099631885626274299): “Use your voice to explore a topic — in depth — with Gemini Live's Deep Research integration. 1. Just ask the @GeminiApp to run a Deep Research report on a topic, then feel free to close the chat, lock your screen, or keep chatting about other things. 2. Gemini will work asynchronously in the background and send you a notification when your full research report is ready.” - **context** · [@Google](https://x.com/Google/status/2099631887530565780): “Learn more: https://t.co/C8JA0GULeB” **Still to check** - Check the average completion time for an in-depth research report. - Measure the accuracy and source attribution capability of the asynchronously generated report. ## Next - [Previous day: 3 November](https://consonance.fyi/en/jour/2026-11-03.md) - [archives](https://consonance.fyi/en/archives.md) - [method](https://consonance.fyi/en/methode.md) - [Consonance](https://consonance.fyi/en/index.md) - Languages: [Français](https://consonance.fyi/index.md) · [Tiếng Việt](https://consonance.fyi/vi/index.md) - For machines: [llms.txt](https://consonance.fyi/llms.txt) · [RSS](https://consonance.fyi/en/rss.xml) · [sitemap.xml](https://consonance.fyi/sitemap.xml) - HTML version: https://consonance.fyi/en