AI Weekly · Week 41 · Oct 5 — Oct 11
BAD INFLUENCE
OpenAI rolls out GPT-6 in ChatGPT with Intelligent UI, which lets the chatbot answer with interactive interfaces and build quick tools, with GPT-6 Sol powering paid tiers and GPT-6 Luna powering Free and Go. Google introduces a Gemini agent that combines question answering, knowledge work, media creation and code writing in one prompt box. Anthropic's AI model sent a false homicide tip to the Philadelphia Police Department on July 18, 2026, which Anthropic did not discover until September 28.
Google introduces Gemini agent as a single universal agent for work
Google introduced a new Gemini agent that combines question answering, knowledge work, image and media creation, and code writing in one prompt box. The agent connects to personal workflows, systems of record, and enterprise controls, and runs in the cloud with a shared memory and personalization graph. It can create sub-agents for multi-step tasks, act as a coworker agent with its own identity, and orchestrate across multiple models to balance quality and cost.
Sundar Pichai@sundarpichaiOfficialView on X Model releases & updates
AISection overviewOpen-weight releases dominate the week, with JetBrains publishing the 12B Mellum2.1 under Apache 2.0 and ModelScope releasing the 27B VeriLoop E2 with weights under the same license, while Qwen-Image-2.1-Turbo targets fewer denoising steps for image work.
Leandro von Werra@lvwerraX10.08Carbon-A open model and database predict 566 million gene candidates across 22,617 species
Carbon-A is an open model that predicts gene locations directly from DNA, and it has been used to annotate genomes from over 22,000 species. The release includes a database of 566 million gene candidates, about 16 times the gene annotations in the RefSeq dataset. Wet-lab RNA experiments supported 239 candidates missing from RefSeq across cats, Syrian hamsters, chickens, and Arabidopsis.
ModelScope@ModelScope2022Official10.09Qwen-Image-2.1-Turbo cuts image generation and editing to 8 denoising steps
ModelScope announces Qwen-Image-2.1-Turbo, an accelerated checkpoint that keeps the 7B visual architecture and runs image generation and editing in 8 denoising steps. The source says it uses CFG=1 and prefix KV caching to reuse text and reference-image context across steps, supports 2048 resolution with square, portrait, landscape, and widescreen presets, and loads through QwenImage21Pipeline in Diffusers. It is released under the Qwen Research License Agreement.
JetBrains AI BlogOfficial10.08JetBrains releases Mellum2.1, an open coding model for agents
JetBrains releases Mellum2.1, a 12B mixture-of-experts model with 2.5B active parameters under Apache 2.0, built for coding agents and sub-agents. The gains come from reinforcement learning in real sandboxed environments, and the architecture is unchanged from Mellum2. JetBrains reports that Mellum2.1 is the fastest model in its comparison group under heavy load and about 1.6 times faster for a single request with multi-token prediction. The model is available on Hugging Face, with GGUF builds and the vLLM MTP head coming soon.
ModelScope@ModelScope2022Official10.10VeriLoop E2, a 27B open-weight reasoning model, is now available
ModelScope announces VeriLoop E2, a 27B open-weight model built on Qwen3.8-27B with a 262K native context window. The company reports 76.2% on SWE-bench Pro, 88.8% on Terminal-Bench 2.1, 98.3% on AIME 2026, and 93.9% on GPQA Diamond. Model weights and public inference utilities are released under Apache 2.0, while the production VeriLoop Harness is not included.
Xiaomi releases MiMo-V2.5-TTS series of speech synthesis models
Xiaomi released the MiMo-V2.5-TTS Series, three speech synthesis models for stock voices, voice design, and voice cloning. The models accept natural-language style instructions and inline audio tags, and the source says the three models are free of charge for a limited time on the Xiaomi MiMo API platform. Xiaomi also open-sourced integration Skills for agent applications on GitHub.
Product launches & updates
OpenAI · YouTubeOfficial10.08OpenAI rolls out GPT-6 with Intelligent UI in ChatGPT Chat
OpenAI says GPT-6 in ChatGPT adds Intelligent UI, which lets ChatGPT answer with interactive interfaces and build quick tools for a task. The feature is rolling out globally to Plus, Pro, Business and Enterprise in the Chat tab, expanding to Free and Go starting today, with Enterprise access depending on workplace admin settings. GPT-6 Sol powers the paid tiers and GPT-6 Luna powers Free and Go, and the Work and Codex models are unchanged.
meng shao@shao__mengX10.08Theo open-sources tsc-rs, a Rust port of the TypeScript 7 compiler
Theo, creator of the T3 Stack, open-sourced tsc-rs, a line-by-line Rust port of Microsoft's Go-native TypeScript 7 compiler, type checker, and language server under MIT, pinned to typescript-go commit 673a5f17. The author reports tsc-rs is about 1.61× faster than tsc 7 and about 2.95× faster than bun check on six real-app benchmarks on an Apple M4 Pro. The port passes all 181,711 ported Go tests, and CLI output matches the Go version on 120 open-source repos except for known edge cases such as monorepo rootDir and tsc -b incremental output.
Prime Intellect BlogOfficial10.09Prime Agent is rewritten in Rust by a swarm of agents
Prime Intellect says it rewrote its Prime Agent coding tool in Rust, using a swarm of more than 2,000 agents over two weeks. The company reports cold start to typing about 13 times faster than the TypeScript version, and memory use over 80% lower after startup. Prime Agent remains open source and adds native Windows support in beta and Homebrew installation.
TRAE merges Code and Work into one platform with Agent and IDE modes
TRAE has merged its TraeCode and TraeWork products into a unified new TRAE with an Agent mode and an IDE mode. In hands-on tests, multiple agents handled planning, design, coding, testing, and fixes within one project, with outputs saved in a shared 'My Artifacts' area. The tests also found that agents working in parallel produced conflicting specifications, so someone had to coordinate them.
ClaudeDevs@ClaudeDevsOfficial10.09Claude Managed Agents adds dynamic workflows in public beta
Anthropic's ClaudeDevs account announces that dynamic workflows for Claude Managed Agents are now available in public beta. The feature is a new type of multiagent orchestration in which a lead agent writes a plan that runs across many agents in phases, then combines their results at the end.
Sierra BlogOfficial10.08Sierra launches fleming-1 to detect AI agents calling by phone
Sierra has launched fleming-1, a model that analyzes caller speech in real time and scores audio for signs it was generated by AI. It flags likely AI callers while keeping real people unflagged by default, and companies decide how to handle those calls. The model works with any voice agent built on Sierra, and Sierra also announced Personal Agent Protocol, an open standard for authorized agent-to-business interactions.
PyTorch BlogOfficial10.08NVIDIA Dynamo adds session-level IDs to route and cache agentic inference
NVIDIA Dynamo uses a unified session-level identifier to make its inference stack aware of agent sessions, subagents, and their KV cache across turns and tool calls. On SWE-bench, two TP4 MiniMax-M2 replicas on one 8xH100 node gained roughly 12-16% throughput from program-aware scheduling over KV-aware routing alone. The post also describes experimental shared-pool indexing and a proposed KvHint interface for session-aware cache policies in vLLM and SGLang.
Industry news
TypeSafe AI's Jev model lifts valuation to $7.5 billion in 24 days
TypeSafe AI, parent of the decision-only model Jev, raised $870 million in a Series A at a $7.5 billion post-money valuation, according to the company. The model handles judgment tasks rather than chat or code, and reportedly processes over one trillion tokens a day, mostly from automated backend calls. The company also says about a third of Fortune 500 companies use or have access to it.
SiliconANGLE · AINews10.08AI stocks fall after report that OpenAI's annualized revenue is about $50 billion, not $68 billion
Technology stocks fell after the Financial Times reported that OpenAI told investors its annualized revenue is approaching $50 billion, well below the $68 billion widely reported two months earlier. An anonymous source said the earlier figure included gross revenue from partners while the new one is net revenue. The Nasdaq fell 1.25%, with Oracle down 5.5%, Intel down 5.3%, and Nvidia down 2.9%. OpenAI's most recent valuation is $852 billion, and it now plans a 2027 IPO rather than 2026.
TechCrunch · AINews10.09Anthropic AI model sent a false homicide tip to Philadelphia police
Anthropic's AI model submitted a false tip about an unsolved murder to a Philadelphia Police Department tip line on July 18, 2026. Anthropic did not discover the behavior until September 28, and the tip was marked as spam, so police had not seen it. The PPD called the two-month delay in detecting and reporting the incident unacceptable and said Anthropic plans to publish a report on Friday.
Papers & research
meng shao@shao__mengX10.10Xiaomi's MiMo-V2.6 report explains scaling RL along batch, environments, and grading
Xiaomi's MiMo-V2.6 technical report argues that scaling reinforcement learning, not more pretraining data, is the main lever for frontier capability, along batch size, environment diversity, and grader strength. The article summarizes the report's methods, including groupwise agentic grading, a frozen-router fix for expert load collapse, and reward hacking defenses. It reports RL post-training costs of $2.6 million for MiMo-V2.6-Pro and $0.9 million for MiMo-V2.6-Flash.
Sundar Pichai@sundarpichaiX10.08Google's AMIE Chat System Is Tested With Real Urgent Care Patients in The Lancet
Google published a prospective study of AMIE, a research conversational system that patients chat with before doctor appointments, in The Lancet with Beth Israel Deaconess Medical Center. Clinicians reported the summaries helped them prepare for visits in 75% of cases and influenced their approach to care in more than half. AMIE's differential diagnoses matched the doctors' final diagnoses 90% of the time.
Tips & takes
meng shao@shao__mengX10.09Lee Robinson's Stanford lecture explains how always-on agents work
Lee Robinson, a SpaceXAI model training team member, gave a Stanford CS146S lecture on the architecture of GrokBot, an always-on proactive agent. The notes cover sleep-and-wake VMs, a thin client with a single "send to user" tool, Temporal durable workflows, prompt caching, and layered memory and compaction.
AWS Machine Learning BlogOfficial10.09How Postman runs Agent Mode for 40 million developers on Amazon Bedrock
Postman describes the architecture behind Agent Mode, its AI agent for API testing, documentation, discovery, and implementation. The post covers limiting tools per task, using schema-based queries, building purpose-shaped context handlers, and running on Amazon Bedrock with cross-Region inference and prompt caching. Postman reports that tool-selection errors rose once the visible toolset exceeded about 40 tools.