OrangeBot.AI Digest — 2026-11-06

Hidden songs on your walk, chosen by people, not algorithms

A live board of security flaws

Turn your run into a ready-to-post Instagram Story or Reel

Which "Open" companies are actually open source

Turn selected text into quick actions on your Mac

Agent orchestration in the terminal, designed for Rex

Safae database backups that AI can restore but not delete

Release infrastructure for developers

Open-source AI coding agent

Test infrastructure for hardware teams

Turn any spreadsheet into a data story in seconds

Changelog + feedback board for closing the loop on your git

Control your Mac’s volume and brightness from the trackpad

Your autonomous research lab

The database for the autonomous AI era.

Agent decisions depend on empirical relationships that deployment observations reveal and task descriptions alone may leave unresolved. We introduce Decoupled Predictive Decision Making (DPDM), an architecture that formulates heterogeneous agent decisions as conditional prediction with a shared pretrained learner. Language models provide semantic representations of current and historical records, while a general-purpose conditional predictor predicts task-defined targets from the represented evidence. Pretraining supplies reusable inference capability; deployment evidence instantiates problem-specific predictions without updating either model's neural parameters. Tool judgments, candidate preferences, and reasoning assessments share an evidence-and-query interface, with predictions mapped to choices by a decision rule. This architecture establishes complementary sources of agent capability: semantic knowledge acquired through language pretraining and empirical inference acquired through structured-data pretraining. A single frozen LimiX-2 predictor serves all five task categories. With Qwen3.5-27B as the semantic encoder, DPDM raises mean accuracy across 50 benchmarks from 78.80% with few-shot prompting to 84.92% under their respective evaluated history budgets, and reaches 82.33% with identical histories and queries. On ToolPRMBench's BFCL and ToolSandbox tasks, absolute accuracy gains are 34.26% and 28.87%, respectively, at the methods' respective evaluated history budgets. Accuracy improves as history grows. With identical imbalanced histories, mean balanced accuracy improves by an absolute 14.51% across five reasoning-step assessment tasks and four encoders. DPDM makes learning from experience an independently improvable foundation-model capability, creating a shared route for advances in conditional prediction to improve decisions across agent tasks.

Models for physical AI tasks, such as robotics and autonomous driving, need not reason solely in language. They can use visual plans--points, bounding boxes, and trajectories--to represent spatial goals and anticipate motion. However, autoregressive vision-language models generate these plans one coordinate token at a time, making richer predictions increasingly costly. We introduce VFlash, a method for fast, accurate visual planning through Flash Primitive Decoding. VFlash exploits the known structure of visual primitives to predict their coordinates jointly, using a lightweight drafter that generates an entire primitive in one forward pass. The layout adapts to the primitive's type, dimensionality, and length, supporting compact boxes and long, variable-length trajectories. Crucially, spatial accuracy need not require reproducing an autoregressive model's exact token sequence. VFlash therefore trains on annotated primitives and directly accepts structurally valid predictions without target-token agreement. Across the evaluated 2D and 3D robot trajectory tasks, VFlash consistently reduces prediction errors relative to autoregressive decoding. Preliminary single-rollout measurements indicate approximately 25x faster generation for 3D point motion forecasting. A world action model conditioned on VFlash plans achieves 70.2% average success across four MolmoSpaces task suites, compared with 62.6% for the baseline.

Harness evolution from a common base exhibits Locality, Complementarity, and Nonstationarity: variants repair different failures, and their utility changes as the policy learns. Rather than seeking a universally superior harness, we introduce Harness-S to selectively rescue failures of a base execution system through sibling harnesses and transfer the resulting capabilities into a single policy that remains effective under the original harness. Directional Harness Evolution develops specialized sibling harnesses along distinct repair directions, forming a reusable Sibling Harness Pool that remains fixed during policy training. Complementary Harness Rescue Training invokes this pool only when all initial attempts under the base harness fail. The same policy generates fresh rollouts under selected auxiliary harnesses, and these trajectories replace the failed group for reinforcement learning. An Adaptive Harness Router selects the Top-k variants using Harness Behavior Profiles and voting across multiple LLMs. Policy-aware profile re-anchoring refreshes this evidence as the policy learns. Across AppWorld, SWE-Gym, and DBBench, the trained policy substantially outperforms policy-only training when returned to the original base harness. Behavioral analyses and held-out cases support the transfer of harness-elicited behaviors into the policy itself. Reusing the router for Rank-1 inference provides additional gains at the same executor rollout count.

Benchmark compression reduces the often prohibitive time and cost of model evaluation by evaluating models on a small benchmark subset that closely matches full-benchmark results. To quantify estimation error, we derive an accurate analytical approximation to the error distribution when estimating full-benchmark scores from a subset, showing that most subsets yield small errors, while a few produce much larger errors, resulting in a long-tailed distribution. This means random sampling alone can achieve surprisingly strong performance with high probability, but remains unstable due to rare large errors. We further show that this instability is inherent to modern foundation models and persists even as more models are evaluated. Based on these findings, we introduce BestRand, a stable and low-cost benchmark compression method that preserves the efficiency of random sampling while substantially reducing its high-error tail. We validate BestRand on 70+ text, multimodal, and agent benchmarks, and further extend it to VLA evaluation. Compared with competitive methods, BestRand reduces high-error risk by 50.2% and estimation error by 16.4% on average, while requiring only 51.3% of the construction cost. It further reduces OOD estimation error by 21.8% on average across score-range, model-family, and temporal shifts. Together, our theory provides practical guidance for understanding and predicting the accuracy and reliability of benchmark compression, while BestRand enables stable and accurate compression at low cost. The code and data are publicly available onhttps://github.com/George-Ling3/BestRand.

Reliable robotic manipulation requires both semantic understanding and precise control in three-dimensional space. However, vision-language-action (VLA) foundations largely inherit visual representations from 2D vision-language pre-training, while spatial supervision is often introduced only during task-specific adaptation. Developing spatial understanding as a transferable foundation-model capability therefore calls for learning from broader robot experience. We present SLA, a Spatial-Language-Action Model that scales implicit spatial alignment to continual pre-training and maintains it during downstream post-training. Initialized from a pretrained pi05 policy, SLA jointly optimizes action prediction and alignment between intermediate visual embeddings and spatial representations from a frozen 3D perception model. We curate and unify heterogeneous robot datasets to support spatial learning across diverse embodiments, tasks, and scenes. To accommodate this diversity, we introduce feature-dependent gates that adaptively aggregate visual embeddings from multiple VLA layers for spatial supervision. The teacher further exploits partially available depth and camera calibration alongside RGB observations, enabling effective use of heterogeneous geometric annotations. Probabilistic state dropout encourages reliance on visual cues and mitigates overfitting to demonstration trajectories. All auxiliary components are used only during training, allowing direct reuse of pretrained VLA weights and preserving the original inference architecture.

We present WeLM-Image, a unified image foundation model that integrates multimodal understanding into text-to-image generation, instruction-guided editing, and multi-reference synthesis. Designed for single-GPU deployment in WeChat's large-scale consumer applications, the model processes text, reference images, and target latents as one interleaved sequence. A Mixture-of-Transformers (MoT) backbone combines shared attention with independent understanding and generation parameters to limit optimization interference. Equipping each path with Mixture-of-Experts (MoE) modules provides sparse capacity, with 31.7B stored parameters and approximately 5.4B generation parameters activated per target latent token. A generation-oriented data engine combines a knowledge-grounded visual vocabulary with large-scale visual clustering to balance semantic coverage and editing tasks, while training-time evaluation guides iterative data refinement. Bilingual annotations and verified editing supervision support progressive training up to native 1.5K resolution. We optimize generation and editing separately with rubric- and checklist-based rewards, then consolidate the specialized policies through noise-aware on-policy distillation. In the reported overall comparisons, the 1536-resolution model ranks first among the evaluated open-source models on PRISM-Bench, Qwen-Image-Bench, and UnicBench. MeanFlow-based few-step distillation yields WeLM-Image-Turbo, which generates a 1024times1024 image in 4.98\,s on a single NVIDIA H20 GPU using eight guidance-free steps.

Strong single-turn editing performance does not guarantee reliable multi-turn interaction: as edited outputs are repeatedly reused, small unintended changes can accumulate into severe visual drift. This long-term drift is often mitigated through training on multi-turn data or cross-turn rollouts, potentially altering pretrained priors, or through specialized inference-time mechanisms that are difficult to scale to foundation models. We introduce TailSteer, a final-stage training framework that improves multi-turn robustness while retaining the generative and editing capabilities acquired through earlier training. TailSteer follows a simple principle: keep early denoising dynamics close to those of the base model and concentrate preservation corrections near the trajectory tail, where fine-grained deviations can be suppressed. These corrections are learned from the editor's own on-policy denoising rollouts, using unchanged source content as a preservation reference without requiring curated ground-truth editing pairs. Experiments across the FLUX, SenseNova, and Qwen families show that TailSteer substantially reduces cumulative drift over 15-20 editing turns while largely preserving base-model performance on extensive text-to-image and single-image editing benchmarks. These results demonstrate that long-horizon editing robustness can be unlocked in foundation models after TailSteer, rather than by relearning their existing capabilities.

Long-term memory enables large language model (LLM) agents to provide continuous personalized assistance by retaining user information across interactions. Existing systems support this process through fact extraction, organization, updating, and retrieval, but largely maintain memories as independent facts. This fact-centric paradigm may overlook how new facts update, constrain, or conflict with earlier ones, while leaving their applicability, role, and priority for the current request implicit. In this paper, we introduce EvoMemState, a state-centric framework that augments persistent facts with two complementary states: Relational States that capture how facts remain valid and relate to one another as memory evolves, and a Reasoning State that determines how these facts should be applied to the current request. Specifically, EvoMemState consists of three key components: (i) Ontology-Guided Memory State Induction extracts Persistent Facts and constructs Relational States that maintain their validity, scope, and evolving relations. (ii) RL-Based State Evolution internalizes memory-use experience into a policy through reinforcement learning to generate a request-specific Reasoning State. (iii) State-Aware Reasoning then combines the relevant facts with their Relational States and the generated Reasoning State to guide the personal agent. Extensive experiments on long-term dialogue and memory-use benchmarks demonstrate consistent improvements over existing memory systems in retrieval completeness and memory-grounded response accuracy. Our code and implementation details are available at https://github.com/XMUDeepLIT/EvoMemState.git.

Autoregressive approaches with large language models, such as long chain-of-thought reasoning, have become the default for new AI applications because they reach higher accuracy on a wide range of tasks. In production, however, they give inconsistent answers, require parsing of outputs, and add cost and latency. Decision models address this: given a question and a set of options supplied by the user, they return a probability distribution over the options without autoregressive decoding. We present Drex, a family of four open decision models with 8B (DLM), 9B (v1.5), 35B with 3B active (v2.0), and 27B (Max) parameters. Each request consists of a state (a document, record, or agent context, optionally with images) and questions about it, each asking for a choice among options, a yes/no judgment, or a score on an ordered scale. Training combines supervised fine-tuning with reinforcement learning from adjusted feedback (RLAF), plus vision tuning for Max. Ablations show that RLAF improves accuracy and that consistency losses improve robustness to option order. On Decision Index 0.4.1, Max sets a new state of the art among the decision models we compare (67.52 on the public suite vs. 60.70 for Jev 1.14.0). On Decision Bench v4, it is statistically tied with d3-turbo and GPT-6 Sol and trails Claude Opus 5.5 by 1.6 points, and it has the highest accuracy among decision models on DecideBench. Max answers a typical request in 183 ms, against about 2.0 s for GPT-6 Sol and 3.5 s for Claude Opus 5.5, and its outputs are stable under option reordering. All four models are open source: https://github.com/nace-ai/drex-decision-models

Multimodal agents use web and image tools to answer questions beyond their internal knowledge. Some search tasks use images mainly to identify an initial entity, then rely on text retrieval. These tasks offer limited insight into how agents connect information across images, objects, and search steps. We therefore focus on visual context management. This involves organizing and using visual information and linking it to other findings during search. Visually intensive tasks require agents to combine object appearance, local details, or spatial relations with external information. We introduce VisSearch-Bench, with 204 human-written tasks across 7 main domains and 25 subdomains. It covers Set-wise Visual Search, Sequential Visual Search, Hybrid Wide-Deep Search, Competitive Visual Disambiguation, and Visual-Structure-Guided Search. Each task provides a reference answer, traceable evidence, and scoring rules. Accuracy measures complete task success, while Process Score measures the correctness and completeness of individual results and subquestion answers. Across 16 multimodal models, the best system fully solves only 26.47% of tasks, even with search tools. These results highlight the challenge of linking visual evidence with search results to produce complete answers.

Egocentric human video offers a scalable source of manipulation priors for robot learning, yet translating raw data scale into generalizable robot policies demands rigorous data curation and training recipes. We introduce Open-HumanScale, the first fully open recipe for scaling egocentric pretraining to 50,000 hours, anchored by Open-HumanScale-50K---a curated egocentric video corpus densely annotated with hand motion and natural language. Leveraging this corpus, we systematically investigate pure human pretraining via closed-loop simulation evaluations and uncover an empirical log-linear scaling law linking pretraining data hours to downstream robot success rates. Grounded in these findings, we develop HumanWAM-Vanilla, a world--action model pretrained exclusively on 50k hours of egocentric human video. Upon robot post-training, HumanWAM-Vanilla reaches 92.08% success on RoboTwin 2.0 Clean2Clean and 68.88% on Clean2Random, alongside a 14.17% mean success rate and a 19.00 mean progress score on RoboDojo. To push the capability frontier further, we explore joint human--robot pretraining with HumanWAM-Goal, a visual-goal-conditioned world--action model. Hybrid data scaling unlocks broad zero-shot generalization across both simulation and physical benchmarks, propelling post-trained RoboTwin 2.0 performance to new state-of-the-art levels: 92.70% on Clean2Clean and a substantial leap to 84.60% on Clean2Random. On physical robots, HumanWAM-Goal further exhibits emergent embodied in-context learning, dynamically adapting its manipulation behavior to unseen visual goals without parameter updates. To support reproducible scaling research in embodied AI, we open-source the Open-HumanScale-50K corpus, data curation and pretraining pipelines, alongside the pretrained and post-trained HumanWAM checkpoints, inference code, and simulation evaluation suites.

Prefill-to-Autoregressive Cross-model Transfer (PACT), a form of latent communication between language models, combines a larger model's prompt representations with a smaller model's efficient generation, offering a promising way to balance capability and inference cost. However, existing systems provide limited support for independently evaluating large-small model collaboration, and improved task performance alone does not establish whether the receiver benefits from the sender's information. We introduce PLatour, an open-source PACT framework that connects frozen models through lightweight, zero-initialized linear bridges trained directly with task rewards and supports hybrid-attention architectures. PLatour significantly outperforms open-source baselines on generation tasks and surpasses Mostik's reported results on shared mathematics and coding benchmarks using a substantially smaller sender. However, controlled experiments show that the gains on some tasks can be reproduced by uninformative sender signals, self-communication, or receiver-only training. We find that trainable bridges can act as parameter-efficient adaptation mechanisms without exploiting useful sender information. We therefore propose PACT-Check to quantify how much of a trained PACT system's gain genuinely comes from communication. Finally, we discuss the opportunities and challenges of PACT systems. Code is available at https://github.com/zhenyi4/platour.

Recent advances in Gaussian splatting have enabled efficient and photorealistic head avatar reconstruction, yet monocular video-based or image-based methods still suffer from under-constrained geometry and texture, leading to artifacts such as blurring, stretching, and temporal instability. We present EDGAR (Efficient Diffusion-Driven Gaussian Avatar Refinement), a conditional diffusion framework designed to enhance the quality of Gaussian avatars by correcting artifacts while preserving identity fidelity. To train EDGAR, we construct EDGAR-SET, a hybrid triplet dataset composed of real-captured and synthetic-rendered samples, which simulates realistic Gaussian avatar failure cases by pairing degraded renderings with ground-truth and reference images. EDGAR refines degraded frames through a fusion block that injects identity features extracted by the self-supervised Sapiens model into a one-step diffusion backbone, enabling effective artifact removal without sacrificing efficiency. Our EDGAR model incorporates a Mixture-of-Experts (MoE) architecture to address diverse degradation types and severity levels. The refined multi-view renderings are subsequently used as pseudo-targets to optimize a single underlying Gaussian avatar, so that the final avatar can be rendered without applying EDGAR to every output frame. Experiments on three different Gaussian avatar reconstruction methods, each evaluated under three different settings, show improvements in both perceptual and reconstruction measures over the raw outputs and an SDEdit-style refinement baseline. In addition, cross-view evaluation demonstrates improved calibrated consistency, while tests under large yaw and open-mouth expressions verify robustness under extreme settings. These results collectively indicate a useful operating regime for recoverable appearance artifacts. The code is available at https://kwanyun.github.io/edgar/

Recent advances in Vision-Language-Action (VLA) models have enabled fine manipulation capabilities, yet performing tasks that require working memory remains a key challenge. Most existing VLA architectures operate under a Markovian assumption, mapping only the current observation to an action, which leads to failure in non-Markovian environments. Inspired by the human 'inner monologue', we propose Hierarchical Explicit Language Memory (HeLM), a hierarchical VLA framework that maintains task-relevant history in explicit language through factorized event detection and event-triggered memory updates. We evaluate HeLM across 10 non-Markovian real-world tasks with 56 variants, including five newly formulated inter-episode tasks that require persistent context across task boundaries. HeLM improves intra-episode success by 12.8 percentage points over the strongest baseline and achieves 78% subtask success on compositional generalization tasks, while using up to 4.5x less memory and running 2.5x faster than the keyframe baseline at long horizons.

Chunk-interleaved full-duplex speech language models listen and speak in fixed time chunks, and each chunk must be processed before the next one arrives. Their workload is asymmetric across chunks: listen chunks leave compute idle while awaiting audio, and speak and tool-call chunks require serial generation, which delays responses and tool calls. Using the idle compute for the next chunk requires decoding it before its audio arrives, under an assumed input. Taking silence as the assumed input and comparing the output with the output under real audio on four models, we observe two properties. First, the next audio chunk does not change the output in 95.3% of chunks. Second, 88% of changes begin at the first output token. Building on these properties, we propose Speculative Listening, a training-free, lossless framework for input-time self-speculative decoding. During idle time, the model drafts the next chunk under silence and, for changes at the first token, a second branch from the runner-up first token. It keeps all drafts in a cross-chunk hypothesis pool and verifies them when the audio arrives. On our eight-task benchmark TickTock-Bench, Speculative Listening speeds up speak chunks by 2.6-2.8x and tool-call chunks by up to 23.4x. On average, it reduces the decoding delay before a response starts by 73%, to 74-241 ms, close to the 200 ms gap between human turns, and the maximum lag behind real time by 94.4%.

New York Times : Sources: US intelligence concluded that Megaspeed executive Alice Huang was recruited by China's top spy agency MSS to collect intelligence on Jensen Huang —  U.S. intelligence concluded that Alice Huang, a Chinese business executive who developed a close relationship with Nvidia's leader …

Ryan Barwick / The Information : Filing: Google spent roughly $548M to buy around 3M shares of Qualcomm, which is pushing into serving the AI data center market —  Google's investment arm Google Ventures has acquired nearly 3 million shares of the chip maker Qualcomm, as a result of Qualcomm's purchase of chip software startup Modular …

Bloomberg : Sources: OpenAI has closed on new funding of roughly $11B at a $1.25T valuation; the round could increase to as much as $20B in total —  OpenAI has closed on new funding of roughly $11 billion, the initial part of a round that values the company at $1.25 trillion, according to people familiar with the matter.

Financial Times : Sources: Anthropic is continuing with plans to list before the Thanksgiving holiday this month, expecting to raise $80B-$100B at a valuation of $1.25T to $1.75T —  Claude-maker intends to list before Thanksgiving despite brutal tech rout  —  Anthropic is pressing on with its high-stakes IPO …

PSI : AI physics startup PSI says a system running a non-public Anthropic model proved the Aaronson-Ambainis conjecture, a quantum computing question open since 2008 —  The Aaronson-Ambainis conjecture is now a theorem.  —  Open since 2008, the conjecture of Scott Aaronson and Andris Ambainis concerns …

Bloomberg : Elon Musk says US government contracts will account for under 10% of SpaceX's revenue this quarter, down from 16.6% last quarter and around 20% in 2025 —  SpaceX's reliance on US government-funded contracts will continue to shrink, Elon Musk said, signaling a shift in focus away …

Brody Ford / Bloomberg : Sources: Microsoft has been turning off rented computing capacity from Oracle after declining to renew a 2023 deal, because the hardware used older Nvidia chips —  Microsoft Corp. is winding down an arrangement to rent AI computing from Oracle Corp., part of a larger reassessment of billions of dollars in cloud partnerships.

Rebecca Cohen / NBC News : The US DOJ indicts three foreign nationals for allegedly using AI to impersonate CEOs and attorneys at major international law firms in a $129M+ fraud scheme —  Federal prosecutors say the suspects posed as corporate executives and high-powered attorneys, using AI to convince employees to transfer millions of dollars.

Lawrence Abrams / BleepingComputer : Researcher: *sshole, a decade-old OpenSSH flaw, lets attackers use a password of 256+ characters to log into passwordless accounts on some musl Linux systems —  A newly disclosed flaw involving OpenSSH and musl, dubbed “*sshole,” allows attackers to log in to certain Alpine Linux …

Maggie Eastland / Bloomberg : Democrats, with Elizabeth Warren a top contender to chair Senate Banking, are poised to step up oversight of Nvidia over chip exports and circular financing —  Nvidia Corp. is in for tougher scrutiny from a new Democratic majority in Congress that's set to demand answers …

Todd Spangler / Variety : Fox closes its $22B Roku acquisition well ahead of the expected 2027 date; Anthony Wood joins Fox's board and will co-chair a new Tubi, Fox One, and Roku unit —  Surprise!  Fox Corp. has closed its acquisition of streaming platform Roku, a deal valued at $22 billion, well ahead of the expected 2027 date.

Qianer Liu / The Information : Sources: TSMC is considering expanding into manufacturing 3D DRAM at the request of Nvidia, its biggest customer, amid memory chip shortages due to the AI boom —  Taiwan Semiconductor Manufacturing Co. is considering expanding into the advanced memory chip manufacturing sector …

Emmy Martin / New York Times : Employees at Microsoft, long known for its stability, are grappling with its decision to offer buyouts to 60,000 people as its spending priorities shift to AI —  Within an hour of Microsoft's announcement of buyouts last week, employees rushed to an internal site to see whether they had been offered one.

Eric Geller / Cybersecurity Dive : Experts say the new Democratic congressional majority will likely seek to rebuild CISA, prioritize AI security, revive local cyber grants, probe DOGE, and more —  From CISA to AI and beyond, the new Democratic majority has many opportunities to scrutinize government programs and push new policies.

Alexander Martin / The Record : Preliminary findings: the EU says TikTok violated the DSA by not preventing its recommendation algorithm from trapping users in harmful content “rabbit holes” —  The European Commission accused TikTok on Friday of failing to prevent its recommendation algorithm from drawing users into …

Linux 和维基百科代表了一种被称为同侪生产（Peer Production）的模式——这一名字由法学教授 Yochai Benkler 提出。古典经济学认为，如果想要人们开展合作以及有效分配合作所产生的成果，那么最有效方式是通过私人营利性企业实现，即由这些企业在开放市场中销售商品或服务。企业雇用员工并支付报酬，而员工则通过合同承担义务，按照管理者分配的任务开展工作。员工劳动所创造的产品归公司所有，公司负责为产品定价、管理销售交易，支付员工薪酬以及分配经营过程中产生的利润。从理论上说，这些方法能解决几个关键问题：如何协调各方的利益动机、如何组织和安排工作任务，以及如何确保最终产出的成果能分配给有意愿且有能力支付费用的人。然而同侪生产为古典经济学传所依赖的行为模式提供了一种替代方案。在同侪生产中，无论是合同义务、经济激励，还是官僚式权威，都无法确保参与者一定会贡献自己的劳动，也无法保证这些劳动能有效组织。由此产生的成果属于公共资源（Public Commons），受到非排他性知识产权的保护。这意味着同侪生产所创造的资源无法直接在开放市场上进行定价。在理想形态中，同侪生产项目对自身工作的范围、领导者权威的合法性，以及值得投入精力的任务选择，都保持着完全的集体自主控制权。但在实践中，同侪生产往往只是以不同程度的形式得到运用，实施主体也包括一些并不完全符合上述所有定义标准的机构。

特斯拉 CEO 马斯克（Elon Musk）宣布将于 2027 年年中停产电动皮卡 Cyber​​truck。生产电动皮卡的工厂很可能转为生产电动出租车 Cybercabs。马斯克于 2019 年末宣布了外形奇特的 Cyber​​truck，预计售价低于 4 万美元，宣称它会和其主力电动汽车 Model 3 和 Model Y 一样随处可见。然而 Cyber​​truck 研发比特斯拉的其它车型更坎坷，为量产还修改了底盘结构，而 25 万年预售目标也根本无法实现。如今马斯克的粉丝将只剩下几个月的时间购买这款奇特的电动皮卡。

NASA 宣布了新试验机 X-70 Fractal，它展示了一幅生成图像——即并非实图，而是 AI 生成的。NASA 称，X-70 是载人飞行器，设计巡航速度为 0.75 马赫，最大飞行高度约 13700 米，航程 2778 公里。X-70 起飞重量为 4536 公斤，设计可容纳 900 公斤的机内外试验载荷。NASA 局长 Jared Isaacman 表示，这架试验机有望在两年内首飞。关于其具体设计细节和研发时间表目前知之甚少。NASA 公布的生成图像显示 X-70 与超音速侦察机 SR-71 非常相似。

根据 Carbon Brief 的分析，中国第三季度 7-9 月的 CO2 排放量同比下降 3.1%，为 2022 年以来最大降幅。今年至今中国碳排放量累计下降 0.5%。数据显示，中国第三季度石油消费量同比下降 4%，煤炭使用量同比下降 3%。第三季度 CO2 排放量下降的关键因素之一是燃煤发电量减少 3.3%。背后的原因是电力需求增长放缓，同时风能和太阳能未被利用的发电量也有所减少。太阳能、风能和核能的发电量同比增加了 110 TWh，超过了同期电力总需求增幅的两倍。

私募股权公司 Silver Lake 和沙特主权财富基金 PIF 在今年八月完成以 550 亿美元收购 EA 之后正在磋商收购游戏引擎提供商 Unity Software。Silver Lake 持有 8% 的 Unity 股权。在 Unity 发布最新财报前，收购财团论过约 250 亿美元的估值；财报发布后 Unity 市值升至约 215 亿美元。如果交易达成，Unity 将和 EA 一样私有化。Unity 的游戏引擎和开发者工具在游戏行业被广泛使用。Unity 于 2004 年在丹麦成立。自上市以来，公司业务已从游戏引擎扩大到广告技术和实时 3D 软件。

有 31 年历史的 Salon 于 2026 年 11 月 5 日宣布从即日起停止出版。Salon 创办于 1995 年 11 月 5 日，当时的互联网与今天截然不同。一小群记者开始尝试创造一些与众不同的东西：他们希望创办一种具有鲜明风格、对世界充满好奇，并且坚信互联网能以全新的方式将新闻报道与读者连接起来的出版物。他们开创了一种全新的新闻报道形式，这种形式后来成为定义了一个全新的出版时代。

摩托罗拉宣布了第一款支持安全加固发行版 GrapheneOS 的旗舰手机 Signature 27，但售价高达 1600 美元。Signature 27 配备了高通目前最强大的移动芯片骁龙 8 Elite Extreme Gen 6，内存只有 12GB 没有 16GB 版本（目前内存供应紧张而且价格昂贵），存储空间为 512G​​B (UFS 4.1）；显示屏为 6.8英寸 Extreme AMOLED Duo，120Hz刷新率，12,000 尼特峰值亮度；电池 7000 mAh，90W 有线充电，15W (Qi) 无线充电，等等。

微软画图应用曝出了一个越界读取导致的远程代码执行漏洞，受影响版本从 11.0.0.0 到 11.2605.91.0 之前。画图应用被预装在每一台 Windows PC 上，最近一段时间微软为画图应用引入了越来越多的新功能，其中包括 AI 生成功能，它同时也给画图引入了新 bug。微软用户目前可能需要通过 Microsoft Store 官方应用商店下载修正后的版本，Windows Update 尚未推送新版本。

付费无广告搜索引擎 Kagi 宣布了最新更新，向所有人开放搜索，但免费用户搜索次数有限额。现在任何人无需注册账号可搜索 50 次，注册账号之后再增加 100 次。Kagi 还抨击了 Google 最近的一项决定：如果美国学校选择禁用 AI 那么它的搜索引擎也将会对学校关闭。Kagi 认为：教育不应沦为广告公司的人质，应当停止将企业利润置于教育学生这一核心使命之上。它呼吁美国学校可选择包括 Kagi 在内的其它搜索引擎。

2017 年发售的《尼尔：自动人形（Nier: Automata）》释出了最新更新，加入了对简繁体中文以及韩文的支持，但游戏的打包显然存在 bug，整个游戏文件的容量从 50GB 暴涨至 366GB。《尼尔：自动人形》共有八个 DLC，每增加一个 DLC，游戏的容量就会增加 40GB，也就是每个 DLC 都包含了主游戏文件。现在想要玩《尼尔》的中韩玩家最好先等待游戏开发商释出修正，除非你的 SSD 有几百 GB 的可用空间。

印度计算机应急响应小组（CERT-In）星期三发布安全通报，被称为“狡猾猎鹰”（Crafty Falcon）的黑客组织，已入侵非银行贷款机构、金融科技公司及加密货币交易所。攻击者利用公开可用的 AI 工具，实现高速自动化入侵。印度计算机应急响应小组形容这项威胁“仍在发生，且持续活跃”。该机构点名四款用于入侵的工具，分别是 ARTEX、PentAGI、Cairn 和 CyberStrikeAI。ARTEX 此前曾被用于攻击韩国银行。印度计算机应急响应小组没有说明这些攻击事件由谁发动，也未透露该组织的运作地点。该机构也没有称这些攻击由国家支持或主导，且其点名的攻击工具均可在网络上免费获取。

今天的电网接入了数以百万计的数字控制设备，为了维持电力供需平衡，精准预测和实时控制必不可少。电力系统因此转变成为一种信息物理融合的基础设施（Cyber–Physical Infrastructure）。物理设备与数字控制系统之间形成了紧密的相互依存关系，在带来前所未有的效率与灵活性的同时也引入了新的系统性风险。根据发表在《Nature Energy》期刊上的一项研究，德国研究人员指出全球电力系统存在的一种由信息物理同步效应所引发的、周期为一分钟的相干振荡现象。除了故障和恶意攻击外，即使是良性的控制程序，当大量设备以协同一致的方式运行时，也可能产生系统层面的影响。在大规模信息物理能源系统中，非预期的同步行为可能引发反复出现且具有潜在放大效应的振荡现象。

越南一家法院对 14 名参与盗版足球赛事直播网站 Bun Cha TV 的团队判处最高一年徒刑。4 名内容、编程和运营团队负责人被判处 12 个月徒刑，另外 5 人判处 10 个月监禁，2 人判处 8 个月监禁，3 名解说员被判处 10 个月缓刑。所有 14 名被告都承认了指控，配合了调查。Bun Cha TV 累计直播逾 15 万场体育赛事，通过博彩及赌博广告非法获利逾 200 亿越南盾（约 512 万人民币）。内容团队负责人获利逾 4.3 亿越南盾，解说员非法所得则在 3100 万到 4.8 亿越南盾不等。美国 MPAA 的代表称赞了这一判决。

安全公司 Layer3 Intel 报告，他们在分配给白宫的自治系统 AS2650 中监测到住宅代理服务的活动。住宅代理服务会将用户的网络地址租给其他人使用。Layer3 在 10 月 5 日-11 月 3 日期间观察到了 236 次住宅代理连接活动。住宅代理的源头被认为来自 Windows 11 设备安装的俄罗斯代理程序 Infatica。Layer3 Intel 指出，在30 天内有 26 天出现了源自白宫的代理出口节点流量，涉及 6 个 IP，大部分活动发生在傍晚，即下午 5 点至午夜之间。

基于 Fedora 的游戏发行版 Bazzite 释出了 v45。主要变化包括；基于 Fedora 45，更新 Open Gaming Collective(OGC)内核，支持罗技 TrueWheel 设备等等。过去几年 Linux 对游戏的支持有了显著改善，但对很多用户而言仍然需要自行配置。Bazzite 项目旨在让 Linux 桌面用户无需额外配置就能畅玩游戏，设计类似 Valve 的 SteamOS，但更适合通用桌面使用情景：SteamOS 主要支持 AMD GPU，Bazzite 包含了私有的 NVIDIA 驱动。