Every has extended its eight levels of AI adoption with Level 9 and Level 10, both centered on simulation. Level 9 pits strategies against each other to find the best one, while Level 10 runs that competition repeatedly to see which strategies win consistently. The author cautions that both levels remain experimental, burn far more tokens than lower levels, and do not guarantee better results today.
MiniMax's H3 family expanded beyond video this week, with H3 Image, H3 Image Highspeed, H3 Highspeed, and H3 Lite now available in Recraft Studio. The two image models generate and edit at 2K or 4K, while the two video models produce 4–15 second clips at 480p or 768p with sound. H3 Image starts at 6 credits per image and H3 Lite at 4 credits per second of video.
Figma has released 3D transforms in open beta, letting designers tilt, rotate, and arrange text, images, vectors, and frames in space, with a perspective effect on parent frames for shared spatial compositions. The non-destructive effects stay editable and can be combined with shaders, animated with Figma Motion, or explored via the Figma agent. Figma's Brand Studio used the tools, including an agent-built generative plugin, to produce the feature's launch video.
Google has launched an experimental AI tool called Tabletop that generates customized collaborative hubs for coordinating with family and friends. Users add links, files and ideas, then prompt @tabletop or pick a starter template to create interactive trip itineraries, dinner polls, expense splitters and tournament brackets, and share an invite link so others can join in real time like a Google Doc. The beta is available via a waitlist and is currently limited to people over 18 in the US.
Anthropic launched Claude for Manufacturing, a plugin with 17 skills covering sales, quality, operations, supply chain, design and setup, from drafting job quotes to diagnosing a stopped machine. The plugin pulls from a manufacturer's own records through connectors to Autodesk Fusion, Infor, Microsoft 365 and Oracle NetSuite, and each skill ends at a human review step before output reaches a customer or supplier. Anthropic cites a 2025 Manufacturing Leadership Council survey in which 79% of manufacturers said they enter data manually, and says it plans to add more manufacturing skills over time.
J.Crew launched Style Advisor, an AI stylist on JCrew.com that recommends complete outfits rather than single products, available now for women's and men's styles in the U.S. J.Crew's team built it with Claude Code using multiple Claude models: Claude Fable 5.1 to plan the stylist reasoning, Claude Opus 5.5 for the prototype, Claude Sonnet 5.5 as the stylist in each conversation, and Claude Haiku 5.5 for quick tasks. Every recommendation is drawn from looks created by J.Crew's own stylists for catalog shoots.
Dan Shipper used his closing keynote at Every's Thesis: 2027 conference to argue that once everyone has the same AI models, human judgment about what's good becomes the differentiator. He contrasted the "aggregate" view, which extrapolates trend lines from data and suggests AI will replace us, with the "intimate" view that zooms in on one real situation. The newsletter also covers turning AI corrections into tests and where hours saved by AI agents actually go.
Marktechpost launched GPU Leaderboard, a free cloud GPU price comparison site for AI inference that ranks 201 live offers from 24 providers by $/GPU-hr. Its cost calculator prices a full run including idle time: an 8×H100, 72-hour job at 90% utilization ranges from $967.68 at Latitude to $3,545.28 at CoreWeave. The companion Deploy It Chrome extension estimates VRAM at FP16, INT8 and INT4 and shows the cheapest fitting GPUs on Hugging Face model pages and GitHub repos.
German defense tech company Helsing unveiled Clarity, an agent-based platform that fuses sensor feeds, open-source intelligence and written reports into target recommendations for human review before data is packaged for a weapons system. Helsing says the platform runs on the customer's own infrastructure with models and interfaces developed in Europe, and that a human must authorize every handoff to a weapons system, with each decision documented under the responsible person's name. The article notes Helsing did not say whether Clarity is already running with any customer, and contrasts the promises with reported cases in Iran, Gaza and a 2026 near-miss involving a Chinese ship where AI-generated assessments were trusted uncritically.
Anthropic kicked off its IPO roadshow yesterday, and its roadshow video disclosed previously unreleased details about how the business works, including a section on compute strategy with co-founder Tom Brown. Brown said Anthropic trains and serves Claude across AWS Trainium, Google TPUs, NVIDIA GPUs and AMD GPUs, and that when the company began working with AMD GPUs this year, a system called Mythos got Claude running on AMD in three days with basically no human intervention, porting or writing almost 100% of the kernels. He also described compute multipliers as a second driver of compute advantage, saying each multiplier stacks on the ones before it and that Claude may eventually discover them entirely by itself.
Why it matters: Anthropic's IPO roadshow video details its multi-chip training and serving strategy across four chip families, including a Claude-assisted AMD port completed in three days.
a16z's Olivia Moore argues that AI agents now make up the majority of web traffic, write more than half of new online articles and generate most production code, marking the early stage of a distribution shift. She cites Shopify's report that AI-driven traffic to its stores grew eightfold year over year in Q1 '26 and orders from AI-powered searches grew nearly thirteenfold, plus Human Security data showing agentic browser traffic grew 8x between August and October. Moore groups potential beneficiaries into agentic models, products newly enabled by agents, incumbents optimizing for agents and agent infrastructure, and notes OpenAI launched Businesses in ChatGPT the day before.
Google DeepMind announced AlphaGenome 2, a next-generation DNA sequence model that predicts how DNA sequences and their variants affect specific human cell types. Trained on more than 107 million individual cell profiles across 68 tissues, the conditional model identified nearly twice as many expression-changing variants as the previous model, and on the TenK10K immune cell dataset it recovered 30% of causal variants versus 16% for the original AlphaGenome fine-tuned on single-cell data. It is available for non-commercial research via the AlphaGenome API and in private preview for commercial research.
Why it matters: AlphaGenome 2 shifts from bulk-tissue averages to single-cell conditioning, roughly doubling causal variant recall on immune cell data, which matters for linking GWAS signals to specific cell types.
Google DeepMind has unveiled AlphaGenome 2, a successor to last year's AlphaGenome that predicts how genetic variants affect individual cell types and states rather than broad tissue averages. Trained on more than 100 million individual cell profiles across 68 tissues, the model identifies nearly twice as many variants that can change gene expression as the previous state-of-the-art model, and can make predictions about new cell types and biological states without retraining. Google is offering it for non-commercial research via the AlphaGenome API from today, with a limited private preview for commercial use; DeepMind's Žiga Avsec said its finer predictions are more computationally expensive, so it will not replace AlphaGenome soon.
Google DeepMind completed a four-week project that mathematically verified key parts of Signal, according to Ars Technica. The work hints that AI could make formal verification far more practical, though the report provides no further technical details.
Adobe announced experimental AI tools at its annual Max conference, including Project Screen Test for directing AI characters through conversation, Project Sight Span for turning flat images into navigable 3D scenes, and Project Sound Bender for mixing sound effects on a square sound canvas. Project Rinse Repeat copies edits and applies them to other subjects in the same photo using native Photoshop tools. These "sneaks" are tech previews that may eventually reach Creative Cloud or Firefly, with no guarantee of launch.
ElevenLabs launched its Partner Program with enhanced commercial incentives, new certifications, joint go-to-market opportunities and dedicated technical support. The program covers 550+ active partners in 44 countries, and the partner ecosystem has generated $315 million in pipeline and over $52 million in partner-sourced and partner-influenced revenue this year. ElevenLabs aims for partners to earn $3 in services revenue for every $1 it generates, backed by forward-deployed engineers and a new certified partner directory.
At its New York Summit, ElevenLabs announced new models, agents and deployment options for its AI interaction layer, including the Interaction Engine that manages turn-taking and routes each part of a conversation to a cheaper or frontier LLM. ElevenAgents gains Memory, which builds a structured record of returning customers, and Routines, an automation layer for multi-step work before, during and after calls. ElevenLabs also previewed the ElevenAI workplace app with free early access opening today, launched a Partner Program now at more than 550 partners in 44 countries plus a McKinsey alliance, and introduced the ElevenLabs Marketplace; NVIDIA's Nemotron 3.5 Super VL is now available in ElevenAgents alongside 15 MW of new Vera Rubin compute capacity.
ElevenLabs outlines how voice cloning lets content creators record a short sample once and then generate narration for videos, podcasts, courses, and social clips without returning to the booth. The guide distinguishes Instant Voice Cloning, which produces a clone from a short sample in under a minute, from Professional Voice Cloning, which trains on a larger recording set and includes a verification step confirming the voice belongs to the creator. It also lists selection criteria including voice fidelity, delivery control, language coverage, consent safeguards, export options, and commercial licensing.
ElevenLabs launched the ElevenLabs Marketplace, a curated collection of industry-specific AI and voice applications built on its models by technology partners, with launch partners including Salesforce, IBM, NiCE, Twilio, and Zendesk. Enterprises can explore solutions across financial services, insurance, healthcare, telecommunications, retail, and government, and connect directly with the partners behind them. Starting next year, eligible partner solutions will be purchasable through existing ElevenLabs commitments.
Modal is one of three companies, alongside Ant Group and Tines, joining gVisor's maintainers as Google donates the project and its trademarks to CNCF under a maintainer-led governance model. Modal says it launches upwards of 40 million gVisor containers a day and has contributed upstream since 2023, including nvproxy GPU support, an RDMA device proxy called rdmaproxy, and tmpfs and memory-allocator optimizations. The company says it is working on multi-GPU memory snapshots and GPUDirect Storage support for gVisor.
ElevenLabs and NVIDIA have expanded their collaboration across models, infrastructure and sovereign deployments, bringing NVIDIA Nemotron 3.5 Super VL into the ElevenAgents platform. ElevenLabs plans to add 15MW of NVIDIA Vera Rubin capacity on top of its existing Blackwell Ultra systems, and is powering a voice agent for Poland's Centralna e-Rejestracja appointment system on-premises using NVIDIA Confidential Computing. On the ElevenAgents customer-support benchmark, Nemotron 3.5 Super VL reached leading end-to-end task-completion accuracy at 70% lower cost than other leading open models, according to ElevenLabs.
McKinsey and ElevenLabs announced an alliance to help organizations prove the value of conversational AI through initial use cases before scaling across the enterprise. The partnership combines McKinsey's industry and transformation expertise, including QuantumBlack's AI engineering, with ElevenLabs' frontier conversational AI technology and ElevenAgents platform. Dutch telecom KPN is already using the new AI-powered solutions, and the alliance plans to extend the approach to service-intensive industries such as insurance.
ElevenLabs says enterprises now account for more than 55% of its revenue, with its technology in daily operations at five of the 10 largest tech companies, five of the 10 largest insurers and four of the 10 largest telecoms. At its NYC Summit, clients including Ramp, Customers Bank, Banco Inter, Clover Health and FOX Sports described using ElevenAgents and ElevenCreative for support agents, onboarding, multilingual service and an AI persona of commentator Colin Cowherd in the FOX Sports App.
US data vendors including Mercor, Surge AI, Turing, AfterQuery and Pareto are selling reinforcement learning environments to Chinese AI labs, and the report identifies nine vendors with China business. It argues these sales let Chinese labs skip costly rollout experimentation and save memory-constrained compute, while noting the deals fall largely outside export controls and that the lag behind US labs limits the damage.
Mercor raised a $750M Series D at a $20B valuation after reaching $3B in annualized revenue, up from $1B in February and $2B in June. The company reports 5 million experts and frames its market as the 800-plus US occupations whose workflows models have yet to be evaluated against. Mercor also says it already spends 3x as much on inference as on employee salaries, and argues that if the ratio holds across the economy it points to double-digit GDP growth over the next five years.
Google Arts & Culture has published the first online retrospective of South African artist Dr. Esther Mahlangu, timed to her 91st birthday, with more than 45 curated stories and a rich archive of her art, life and legacy. Because archives held nothing from before 1980, the team used AI to visually recreate her undocumented memories, generating still images and then using Veo to turn them into videos. The exhibition was curated with institutions including the British Museum, Iziko Museums of South Africa and Serpentine.
More than 20,000 veterans and transitioning service members have claimed a free year of ChatGPT Plus, OpenAI said, and the company highlighted how they use it to prepare for civilian careers.
MBZUAI's Institute of Foundation Models released PAN 2, a 29B-parameter open world model that takes a scene plus a text action and simulates the outcome as video, with weights, training code and inference code under Apache 2.0. It pairs an LLM planner that writes structured JSON world states with a 29B MMDiT diffusion transformer conditioned by a frozen Qwen3.5-9B encoder, and IFM reports 34.0% full-plan completion on 100 multi-step tasks, ahead of MiniMax-H3 at 32.50%, while trailing MiniMax-H3 on single actions and human preference (Elo 1595 vs 1620). IFM's ablation shows removing the LLM planner drops full-plan completion to 0.0% and step completion from 75.1% to 8.7%.
Tether AI Research released BotPsy, a 0.8B-parameter agentic model built on Qwen3.5-0.8B for on-device personal-assistant tool use. It roughly doubles its base model's agentic scores, with BFCL v4 rising from 23.4 to 44.0 and τ-bench from 23.2 to 51.5, and beats every model up to 1.2B parameters tested across six tool-use benchmarks. It also cuts unsafe responses to 21% from 36% and uses about 3× fewer reasoning tokens on GPQA Diamond.
OpenAI Developer Experience engineers Dominik Kundel and Charlie Guo presented "New Ways of Working with Codex" at the AI Engineer Code Summit, covering the revamped Codex Cloud v2 announced at DevDay, voice mode as a control plane for managing coding threads, and personal agents called dots. Kundel said Codex Cloud now acts as a background executor, letting local Codex hand off tasks to the cloud, though he acknowledged the cloud workflow still needs to be easier.
inclusionAI released LemonDecider 4B v3, a decision model fine-tuned from Qwen3.5-4B that returns choice, yes/no or ordinal probabilities via a single joint forward pass instead of generating answer tokens. It was trained on about 1.20M decision/QA rows and scored 61.67 on the public Decision Index 0.4 suite, with median request latency of 49.9 ms.
NVIDIA highlights four Inception startups using AI and robotics to sort mixed waste at recycling plants. AIRECO's Jetson Orin NX pilot line tops 95% recognition accuracy and targets raising annual capacity from 18,000 to 48,000 tons while cutting operating costs by 85%, and EverestLabs has deployed about 150 vision-guided robots across North America and Australia.
Vapi launched Agent Loop in early access, with GA planned for early next year, as a self-improving harness that builds a voice agent from a plain prompt, scores every call against a business goal, drafts fixes for the misses, tests them in simulations and ships them to a slice of live traffic before full rollout. The loop runs on Vapi primitives including Simulations and Traffic Splitting, available in the dashboard now, plus Goals, Call Health and Experiments, coming soon. Every Agent Loop agent is a real Vapi assistant, so engineers keep full dashboard and API access, and human approval is required before any change ships.
vLLM has expanded Encoder/Prefill/Decode (EPD) disaggregation for multimodal serving, adding metadata-only requests, a dedicated encoder execution path, remote embedding transfer via NIXL and Mooncake, cross-encoder output reuse, and orchestration through vLLM Router, SMG, and llm-d. Core capabilities including MiMo and XPU support ship in vLLM v0.32.0, while metadata-only support for DeepSeek V4.1 Flash is available on main via #62802. In a DeepSeek V4.1 Flash test on GB200, EPD delivered 6.6–9.5% higher per-GPU throughput and 44–47% lower P90 TTFT than an eight-GPU PD baseline at concurrency 64–128, but had lower per-GPU throughput at concurrency 16.
OpenRouter measured the same GitHub task across the GitHub MCP server, the gh CLI and a gh skill on Claude Sonnet 5.5, GPT-6 Luna and Gemini 3.8 Flash. On Sonnet 5.5 all three answered correctly, but the CLI used 1,421 input tokens, the skill 3,315 and the MCP server 59,417 with its 46 default tools, whose definitions alone add 19,436 tokens per request. Scoping the server to two tools cut the task to 23,228 tokens, capping results at 4,000 characters brought it to 5,357, and Tool Search cut the up-front cost of 46 tools to under 600 tokens, though it halved the whole-task cost on Sonnet 5.5 while raising it on Claude Haiku 5.5.
Why it matters: OpenRouter's measured comparison shows how much context an MCP server, a CLI and a skill each consume on the same GitHub task, and where trimming tool results or deferring definitions changes the bill.
Hugging Face shows how pooling-aware training lifts retrieval quality for multi-vector embedding models at a fixed storage budget. With pool_factor=4, a normally trained model keeps 94.6% of its unpooled NanoBEIR nDCG@10 while cutting document vectors 4x, and pooling-aware training raises that retention to 97.8%. The post walks through HierarchicalTokenPooling, budget calibration and a reproducible Sentence Transformers example.
OpenRouter published a guide on handling 429 Too Many Requests errors from OpenAI, Anthropic and its own API, testing three fixes on a 60-prompt batch. The OpenAI Python SDK's default retries left 23 of 60 requests failed in 12 seconds, while a retry that waits for Retry-After or X-RateLimit-Reset completed all 60 in 89 to 92 seconds in Python and 106 to 218 seconds in TypeScript. A models fallback finished all 60 in 13 seconds for $0.000066, and with provider failover on, all 150 pinned-provider requests succeeded where up to 32 failed with fallbacks off.
OpenRouter ran 280 exact-answer tasks against six endpoints of Qwen3.8 27B on October 20, 2026 and did not detect an accuracy difference between fp16, fp8 and fp4, with all six scoring between 98.3% and 99.2%. Across 1,455 endpoints, 59.5% report no precision at all and fp8 is the most common labeled format at 23.6%. Lower precision was cheaper for most models but not all, and provider choice moved price more than precision did, with the priciest same-precision endpoint a median 1.72 times the cheapest.
Why it matters: OpenRouter measured fp8, fp4 and fp16 endpoints of one model across 240 exact-answer tasks and found no detectable accuracy gap, while price differences tracked the provider more than the precision.
JioStar chief architect Vijay Seshadri shares lessons on running AI discovery at 500M-user scale, covering evals, guardrails, and taking AI into production. The OpenAI-published piece focuses on the operational practices behind large-scale AI deployment.
inclusionAI released the ZoomGain-9B and ZoomGain-4B vision-language model checkpoints, alongside the ZoomGainBench evaluation benchmark and ZoomGain RL training data. ZoomGain mines global-to-zoom perception gaps to build fine-grained visual supervision, and ZoomGain-9B reached superior performance on multiple fine-grained and general perception benchmarks. ZoomGainBench contains 1,000 questions over 941 images, while the RL dataset holds 9,000 training records and 7,713 source images.