Skip to content
TodayNov 17, 2026Tue6items
  1. ChinaTalk76

    BIS issues 132-page rule to curb China's AI compute access

    The US Commerce Department's Bureau of Industry and Security released a 132-page export control rule on Monday targeting three channels for China's AI compute: license requirements for chip exports to Singapore, Malaysia, Thailand, Indonesia and Brazil; caps on compute rented to Chinese firms; and a renewed block on Chinese chip designers using TSMC. ChinaTalk estimates remote access supplies about two-fifths of China's AI compute.

    Why it matters: The rule's remote-access cap and TSMC restriction could meaningfully slow Chinese AI training, but an exemption for compute inside the US leaves a major loophole that enforcement will determine.

  2. WIRED · AI58

    Ex-Google founders' Tiberius Aerospace raises $85M for AI-designed weapons

    Tiberius Aerospace, founded by former Google executive Chad Steelberg and ex-Apple executive Andy Baynes, said it raised $85 million in Series A funding led by Lakestar and Rochefort Asset Management, with Ukrainian VC firm UA1 also investing. The company used an AI model to design Sceptre, a 155-mm artillery round, and has turned that system into a product called GRAIL.

  3. SiliconANGLE · AI42

    Noeri launches with $110M to build self-improving 'living' AI systems for enterprises

    Noeri Inc., founded by Stanford researchers Yejin Choi and Carlos Guestrin, launched with $110 million led by Menlo Ventures, with Madrona, Kindred, Wing, NEA, Snowflake and Cisco Ventures participating. Its software studies a company's data and workflows, then proposes explained changes to tasks for human review, and works with both open-weight and proprietary models. Choi said a small fine-tuned model beat GPT-5 on a Stanford research team's projects.

  4. SiliconANGLE · AI78

    Alibaba launches Qwen4-Max with training and attention optimizations

    Alibaba released Qwen4-Max, its newest flagship LLM, built on a Qwen4 architecture that made training more efficient. Its backbone has 2.4 trillion parameters but activates only 71 billion at a time, supported by 328 billion N-gram parameters acting as a cache. Alibaba says it used up to 35% fewer tokens than Qwen3.8-Max on an internal benchmark and 49% fewer on Terminal-Bench 4.0.

    Why it matters: Token savings of this size on coding and knowledge-work tasks bear directly on inference economics, and Alibaba's plan to open-source smaller Qwen4 models could widen adoption beyond its cloud.

Nov 16, 2026Mon
  1. MarkTechPost60

    Meta moves LLM-generated JiTTests into production to catch risky code before merge

    Meta has moved JiTTests, an LLM-based system that writes executable tests designed to fail on risky code changes, from research prototype into production infrastructure. Meta reports 74 privacy SEVs caught at the diff stage in the first half of 2026 and a 20% drop in privacy SEV occurrence, with engineers self-reporting a 60% true-positive rate. JiTTests is internal only, with no code, models or API released.

Nov 15, 2026Sun
  1. MarkTechPost62

    Cmpnd releases dspy-wsp, a DSPy optimizer that edits model weights directly

    Cmpnd has released dspy-wsp, an MIT-licensed prototype optimizer that plugs weight-space programming into DSPy: a frontier LLM reads a student model's weights in a sandbox and patches them, with no gradients or training data. In tests on 0.6B–9B decision models, 25 writer passes cost $5.42–$7.63 and took 78–167 minutes; d1-omni-600M rose from 50.9% to 69.1% with WeightSpace plus ReAnchor.

Nov 12, 2026Thu
  1. Meta Engineering · ML Applications42

    Meta details Intelligent Delivery system for allocating ad-ranking compute

    Meta Engineering described Intelligent Delivery, an ML-driven system that allocates compute across its ads serving stack toward a system-wide optimum. It has delivered about 10% incremental ranking performance on top of model and system advancements from 2017 through the first half of 2026. In that half, AI agents contributed to roughly 20% of its optimizations.

Nov 11, 2026Wed
  1. ChinaTalk67

    ChinaTalk: US data vendors sell RL environments to Chinese AI labs

    ChinaTalk reported that US data companies including Mercor, Surge AI, Turing and AfterQuery sell roughly $500 million of data to Chinese AI labs, some of it reinforcement-learning environments used in post-training. It identified at least nine vendors with China business, citing Mercor sales of software-engineering environments to Alibaba and Surge AI sales to Tencent, Zhipu AI and Moonshot AI.

  2. Meta Engineering · ML Applications58

    Meta details TLX, a Triton extension for explicit GPU warp-group control

    Meta Engineering introduced Triton Low-Level Extensions (TLX), which adds explicit warp-group and cluster-level orchestration to Triton while keeping its programming model. TLX targets NVIDIA Hopper and Blackwell and AMD MI300/MI350x. Meta reports 20-25% end-to-end MFU on GEM training across thousands of Blackwell GPUs, up to two-thirds lower compute-intensive latency in RecSys inference, and production use in LLM pretraining.

Nov 10, 2026Tue
  1. Amazon Science60

    Amazon releases Mitra-v2 tabular foundation model, open-sourced on Hugging Face

    Amazon Science introduced Mitra-v2, a second-generation tabular foundation model for classification and regression, pretrained entirely on synthetic tables with 10 times as many rows as Mitra-v1. It leads the single-model field on TabArena's 51 real-world datasets and ranks first on ScoringBench for regression. Fine-tuned models and inference code are released under Apache-2.0 on Hugging Face.

    Why it matters: Open, synthetic-only pretraining that tops public tabular benchmarks could reduce per-table feature engineering and tuning, a common cost in enterprise ML pipelines.

  2. Recode China AI58

    Ulanqab becomes an AI data center hub, drawing young workers to Inner Mongolia

    Ulanqab, an agricultural city in Inner Mongolia, has turned into a computing hub thanks to cheap power, cool climate and China's East Data, West Computing initiative. Kuaishou opened a self-built data center there in 2023, DeepSeek advertised data center roles paying up to 30,000 yuan a month, and ByteDance's Volcano Engine signed an agreement in August. Hiring standards are now rising.

Nov 2, 2026Mon
  1. The Register · AI/ML58

    India proposes DEPA framework to train AI models inside secure data environments

    India's Ministry of Electronics and Information Technology (MeitY) has floated a "Data Empowerment and Protection Architecture (DEPA) for AI Training" framework in which models travel to the data rather than data travelling to models. Training Data Providers such as hospitals, banks and government agencies would share anonymized, encrypted data via Data Exchange Platforms running hardware-based Secure Data Environments, with all activity logged for audit. MeitY says the scheme could unlock fragmented health data and extend formal credit to 400 million underserved individuals and MSMEs.

Oct 24, 2026Sat
  1. Ahead of AI (Sebastian Raschka)58

    Sebastian Raschka releases 8-part video course on building reasoning models from scratch

    Sebastian Raschka published eight lecture videos, about 11 hours total, accompanying his book Build a Reasoning Model (From Scratch). Starting from a pre-trained Qwen3 0.6B model, the course implements evaluation, inference-time scaling, reinforcement learning with verifiable rewards via a simplified GRPO, and distillation in Python and PyTorch, evaluating checkpoints on MATH-500.

Oct 23, 2026Fri
  1. Amazon Science58

    Amazon releases TorchNeuron, an open-source PyTorch backend for AWS Trainium

    Amazon announced TorchNeuron, an open-source backend that translates PyTorch code into operations runnable on AWS Trainium chips, with minimal or no code changes for code already running on other accelerators. It adds an optional "adaptive eager" mode that groups and compiles operations, plus multi-level caching; a shared compilation service cut time to first training step by about a third in a 64-process test.

Oct 21, 2026Wed
  1. The Register · AI/ML58

    Linux Foundation survey: AI cuts coding time 52% but floods open source with low-quality pull requests

    A Linux Foundation survey found generative AI cut code-authoring time by 52 percent on average, but respondents estimated a quarter of incoming pull requests were low-quality "AI slop," rising to 31 percent at large projects. Only 35 percent of projects have a formal policy permitting GenAI use, while 32 percent have no policy and no plans for one.

Oct 13, 2026Tue
  1. Amazon Science75

    AWS priming cuts hybrid model pretraining cost by over 99%

    Amazon Web Services researchers introduced priming, a method that reuses pretrained Transformer weights to build hybrid state space models, cutting pretraining cost by more than 99% and using under 0.5% of the source model's token budget. A 32-billion-parameter Gated KalmaNet hybrid matched its source Transformer on reasoning while decoding up to twice as fast at 128,000-token context. Code and models were released under Apache 2.0.

    Why it matters: Cheap, controlled architecture experiments could accelerate the shift away from pure Transformers for long-context agents, where KV-cache memory costs dominate serving economics.

Jul 18, 2026Sat
  1. Ahead of AI (Sebastian Raschka)58

    Raschka explains how LLMs control reasoning effort settings

    Sebastian Raschka's newsletter article explains how reasoning-effort settings in models like OpenAI's GPT-5.6 are trained and implemented. It describes effort-conditioned supervised fine-tuning and reinforcement learning with token-length penalties, and compares disclosed recipes from six open-weight models including DeepSeek V4, Kimi K2.5, GLM-5, Qwen3 and Nemotron 3 Ultra. Higher effort generally means longer responses and higher cost.