OrangeBot.AI Digest — 2026-07-31

Your Mac's vitals in the menu bar. Free and open source.

Natural-language prospecting: find, enrich and sync leads.

The customer brain for your company

Why is this running? Trace process, port, container or file

The system of record for your company's AI spend.

OpenTelemetry for production AI applications

Edit tracks & stems, get consistent music with new engine

Turn your team's real workflows into AI training data

Google's AI brain for the next generation of robots

Know who’s real on every video call

New ways to reach customers in the moments that matter

Unified video generation for motion design and branding

Let every AI remember the same you.

See what your Claude Code sessions actually cost

Prompt analytics and citation mapping for AI search

GUI agents have the potential to become a general purpose executor over existing digital devices. To advance them toward real-world use, we envision agents that operate reliably on real devices, execute workflows across platforms, combine GUI interaction with CLI execution, complete long-horizon tasks, proactively initiate useful services, and autonomously improve their capabilities with minimal human effort. Guided by this vision, we present Qwen-UI-Agent, a real-world centric foundation GUI agent spanning mobile, computer-use, web, and DeepSearch environments. Qwen-UI-Agent combines diverse sandbox environments with a large-scale real-device mobile runtime. Its unified action space interleaves GUI operations with CLI execution and generates batched actions in a single model turn. An AutoResearch-style data flywheel uses agents to construct tasks and environments, diagnose failures, and plan subsequent iterations. Online RL supports training on trajectories exceeding 100 turns, with over 10,000 concurrent environments accelerating rollout. A lightweight harness layer supports proactive service initiation and stateful workflows across mobile and computer. Across a broad suite of evaluations, Qwen-UI-Agent sets state-of-the-art performance on mobile-use benchmarks while delivering competitive performance on computer- and browser-use tasks against frontier models, including Opus 4.8, Gemini 3.1 Pro, and GPT-5.6 Sol. On mobile use, it achieves 82.1% on MobileWorld, 92.2% on MobileWorld-Real, and 97.5% on AndroidDaily. On computer use, it achieves 79.5% on OSWorld-Verified and a 40.0% partial-progress score on OSWorld-v2. On browser use and GUI grounding, it achieves 73.6% on WebArena and 81.5% on ScreenSpot-Pro, respectively.

Chemistry literature synthesis often requires assembling specific findings scattered across many publications, yet existing literature-search systems primarily return ranked document lists. As a result, scientists and AI agents need to locate relevant information, verify their provenance, and assemble cross-paper answers manually. We present AskChem, a claim-centered infrastructure for cross-paper chemistry search. AskChem changes the unit of retrieval from the paper to the provenance-carrying claim: each paper is converted into atomic, typed claims, each grounded by a source DOI and a verbatim quote or an explicit evidence locator. Over this shared claim store, AskChem exposes complementary structures for search and synthesis: a stabilized faceted taxonomy for hierarchical retrieval and browsing, an evidence graph linking claims through relations, and an exploratory living taxonomy that situates indexed papers under scientific principles. AskChem currently indexes 2.4M claims from 147K papers and provides a web interface, as well as REST, SDK, and MCP access for AI agents. On AskChem-Bench, grounding a GPT-5.5 reader in AskChem yields 100% resolvable DOIs, compared with 88.3% without retrieval, and the highest citation density among five tested systems. AskChem is live at https://askchem.org.

Recent advances in AI agents have increasingly internalized native capabilities into their underlying foundation models, giving rise to multimodal foundation models and large reasoning models. However, agent memory is still primarily implemented through external modules, leaving the native memory capability largely unexplored. In this paper, we take a first step toward this direction by introducing memory foundation models, which empower foundation models with native memory capabilities. We formalize native memory from two perspectives: a persistent and dynamically evolving memory state within the backbone, and native memory procedures that autonomously store and utilize information through model computation. We show that native memory offers advantages in architecture, end-to-end optimization, and efficiency. Based on this formulation, we propose Metis, the first prototype of memory foundation models. Metis introduces a new architecture that equips a foundation model with a native memory state, allowing historical information to be compressed into the model and accessed through memory attention. We construct large-scale memory-specific training data and introduce multiple optimization objectives to acquire these native memory procedures through mid-training. The online memory maintenance of Metis is gradient-free, and the memory update requires only a forward pass. At inference time, all learned model weights remain frozen, while the native memory states are autonomously transformed through standard forward computation. Through extensive experiments, we show that Metis exhibits native memory capabilities and further provide a detailed analysis of its strengths, limitations, and behaviors. To facilitate future research on memory foundation models, we release our project and model checkpoints.

We introduce PhiZero, a physical world model built around physical language, a compact discrete representation of world-state transitions. Existing physical world models typically predict future videos directly in pixel space, leaving the underlying world dynamics implicit within high-dimensional visual predictors. Motivated by humans' ability to abstract predictive structure from visual experience and organize it in natural language for explicit reasoning, we learn physical language from in-the-wild videos through self-supervision and use it to explicitly reason about how the physical world evolves. Accordingly, PhiZero adopts a reason-then-render paradigm: it first infers future world evolution as a physical-language sequence and then renders the inferred transitions into videos. Extensive experiments across generation and understanding benchmarks validate the ability of PhiZero to model physically coherent world evolution. We further show its potential for realistic and interactive world modeling, fine-grained action-conditioned simulation, and zero-shot motion transfer.

Recursive self-improvement (RSI) requires AI systems that improve the process of building AI (i.e., AI4AI); machine learning engineering (MLE) offers a concrete, executable testbed for studying this capability. We introduce OpenMLE, an open full-stack system for RSI research in MLE, spanning verifiable task environments with execution feedback (OpenMLE-Gym), operator learning (OpenMLE-RL), and long-horizon search (OpenMLE-Evo). On this stack we post-train Frontis-MA1 (35B) as a meta-evolution agent for MLE, aligning post-training and inference around four atomic program-evolution operators (Draft, Improve, Debug, Crossover): the same operators are trained via execution-grounded SFT and RL on data deduplicated against all evaluation benchmarks, then composed into long-horizon search, coupling learning and evolution in a single loop. On MLE-Bench Lite under a 12-hour per-task budget on one RTX 4090 capped at 12 GB VRAM, Frontis-MA1 (35B) improves Medal Average from 39.39% to 60.61% over its base model with OpenMLE-Evo, and reaches 71.21% with OpenMLE-Evo-Max (benchmark-independent experience priors and asynchronous search), exceeding GPT-5.5 + Codex and approaching GPT-5.6 Sol and the 2.8T Kimi K3. On held-out NatureBench Lite, both components transfer: with the framework fixed, swapping in the trained model raises Match-SOTA from 50% to 70%; with the model fixed, swapping in OpenMLE-Evo raises it from 20% to 50%. We release the model weights and the full OpenMLE stack to enable reproducible research on executable AI4AI toward RSI. Code: https://github.com/FrontisAI/OpenRSI

Text-to-video models have achieved remarkable visual quality, yet they still struggle to generate physically consistent dynamics because the temporal evolution of a scene must be inferred implicitly from a highly compressed text prompt. Existing chain-of-thought approaches introduce intermediate plans or visual states, but these representations are typically non-executable or temporally sparse, limiting their ability to instantiate and control the complete spatiotemporal process. To address this limitation, we introduce VideoCoCo, an agentic dual-engine framework in which executable Blender code serves as a process-level chain of thought. Given a text prompt, a coding agent synthesizes a Blender program that explicitly specifies the scene and its temporal evolution. The executable simulation engine runs the program to produce a deterministic spatiotemporal draft, which is subsequently transformed into a photorealistic video by a generative video engine through draft-conditioned editing. This decomposition separates process-level reasoning from high-fidelity visual realization. To adapt the video editor to simulated drafts, we construct VideoCoCo-3K, a curated dataset of draft-instruction-target triplets. VideoCoCo improves the OmniWeaving baseline from 0.475 to 0.558 on PhyGenBench and from 52.18 to 77.88 on VBench-2.0, achieving the best average score on both benchmarks. These results demonstrate that executable code provides an effective, controllable, and inspectable intermediate representation for physically consistent video generation.

Decoder-only language models entangle long-term memory and reasoning in a single parameter set, making it difficult to scale memory capacity independently. Memory Decoder introduces a parametric long-term memory module but only studies it at a relatively small scale. In this work, we present Memory Decoder at Scale, scaling memory models up to 6.9B parameters and pretraining them on 300B tokens. At this data scale, the combined cost of indexing and search makes a standard Faiss pipeline infeasible. We address this bottleneck with a distributed pipeline for Faiss indexing and retrieval, together with sparse, batch-wise loading of kNN distributions. Across model scales, we find that allocating more parameters to memory yields a better parameter-performance tradeoff than scaling the base model alone. On 17 benchmarks, pairing a 6.9B general memory with Pythia-410M raises its average score from 29.86 to 37.34, surpassing Pythia-12B (37.24) with 39% fewer total parameters. For Qwen3 Base models ranging from 0.6B to 14B, 1.7B domain memories improve the average score across the three domains by more than 9 points at every scale. Overall, our results demonstrate that independently scaling pretrained memory offers a more parameter efficient path to improving language model performance.

The fundamental goal of agentic visual reasoning is to improve the success rate of multimodal large language models (MLLMs) on complex tasks, rather than merely equipping them with a sophisticated yet inefficient reasoning paradigm. In this work, we rethink agentic visual reasoning through two key dimensions of tool use: Mode Adaptiveness (MA) and Tool Effect (TE). Mode Adaptiveness characterizes whether an MLLM can recognize when tools are truly necessary and invoke them accordingly, thereby avoiding unnecessary computational overhead while improving performance on challenging problems that require tool assistance. Tool Effect characterizes the actual impact of tool use: tools should extend the model's capabilities on problems unsolvable through text-only reasoning, while avoiding additional errors on problems that the model can already solve without tools. We conduct a comprehensive analysis to quantify these two properties and empirically reveal that existing agentic visual reasoning models exhibit limited Mode Adaptiveness, while the gains produced by tool use on hard examples are largely offset by the harm introduced on easy examples that the models can already solve. Motivated by these observations, we propose Beacon, a novel agentic visual reasoning model that achieves stronger overall performance, improved Mode Adaptiveness, and genuine tool-induced performance gains. At the core of Beacon are the Necessity-Aware Adaptive Reward and the Hint-Guided Capability Expansion mechanism in the reinforcement learning stage, which respectively encourage adaptive tool invocation based on task necessity and strengthen the model's tool-use capability on the most challenging problems. Extensive experiments across diverse benchmarks demonstrate the strong overall performance of Beacon and its substantial improvements in both Mode Adaptiveness and Tool Effect.

Retrieval-augmented generation (RAG) spans lexical and dense retrieval, graph-based indexing, and agentic search, but these paradigms are usually evaluated on different benchmarks at one corpus size, leaving their accuracy-cost scaling unclear. To bridge this gap, we present a controlled study that varies corpus size along 28 strictly nested tiers spanning roughly 450-fold, while holding questions and a fixed bedrock of relevant and adversarial documents unchanged. Under one reader model and one judging protocol, we measure official accuracy, construction and query tokens, and latency. The results reveal a scale-dependent crossover rather than an unconditional winner. File-System Agent leads at the smallest shared tiers, but its sequential exploration costs 39 times more query tokens at the bedrock and becomes less effective as the search space grows. Around 10 million corpus tokens, BM25 overtakes it and leads at every larger shared tier, with a margin approaching 20 points at full scale. BM25 also anchors the low-cost end of the Pareto frontier without LLM-based construction. Dense retrieval remains efficient but less accurate, whereas graph-based RAG encounters construction walls before deployment scale and its scalable variants remain below BM25 at shared tiers. Overall, corpus growth increasingly favors global candidate ranking: lexical retrieval is the strongest scalable default, while agentic reasoning works best after ranked discovery rather than in place of it.

Large language model training in open-ended domains lacks verifiable rewards, making task preferences difficult to formalize as effective supervision. Contexts can convey such preferences, yet provide little additional supervision once distilled into the student, motivating contexts that evolve with student performance. However, directly using evolving contexts as in-training supervision results in an unstable distillation target and conflicting distributions, requiring mechanisms to stabilize target and downweight conflicts. In this paper, we analyze the effect of contexts through a decomposition of the reverse KL objective, revealing two findings: the student is distilled toward the geometric mean of context-conditioned teachers, and the objective contains a conflict term that measures conflicts among these teachers. Based on this decomposition, we propose Flux-OPD, an OPD paradigm that uses evolving contexts as in-training supervision to capture task preferences in open-ended domains. Flux-OPD treats the differences between context-conditioned and context-free teachers as contextual difference signals, injects them as contextual corrections into the context-free teacher anchor, and weights their correction strength using the conflict term as an indicator. Experiments on open-ended tasks show that Flux-OPD outperforms existing OPD paradigms, highlighting the potential to combine teacher supervision with evolving contexts.

Text-to-image and personalized editing models now synthesize high-fidelity single-subject images with ease. Yet placing multiple named people into shared contact actions such as embrace, carry, or grapple still exposes major failures: fused limbs, invented extremities, and interpenetrating bodies. Existing evaluations largely overlook these anatomical and geometric issues, and VLM-as-a-judge checklists often saturate on Interaction while the errors remain obvious to humans. We introduce MPIE-Bench, a 2,500-sample benchmark of video-mined editing triplets spanning 405 scenes, 14 interaction categories, and four contact densities (C0-C3). We also propose MPIE-Eval, whose two new axes score contact-time geometry from a frozen public multi-person mesh reconstruction. Anatomy asks whether every human-like mass is explained by a complete set of reconstructed bodies, and Interaction asks whether the penetration and surface distance between those bodies match the contact the instruction asked for. Across ten editors, mesh Anatomy tops out at 0.65 and mesh Interaction at 0.72 on two different models, so no single editor is strong on both, while VLM checklists rate the same images above 0.95. A five-rater study confirms that both axes track human judgement more closely than a zero-shot VLM judge, and the rankings hold under ablation of every weight and threshold.

Embodied intelligence faces a fundamental data bottleneck. Models must capture how first-person perception, whole-body motion, dexterous manipulation, object state, sound, and touch evolve together as humans pursue goals over time. Existing datasets fragment this experience across viewpoints, modalities, or spatial scales, leaving the full perception-action loop only partially observed. We introduce the Ambient Capture Engine (ACE), a human-centric data engine that transforms real home environments into spatially calibrated, temporally synchronized recording studios. ACE operates at two complementary scales: a table-scale configuration resolves hand-object manipulation, while a room-scale configuration captures whole-body motion, locomotion, and interactions across a furnished home. ACE records egocentric and multi-view exocentric video, full-body and articulated hand motion, object geometry and 6-DoF trajectories, audio, and tactile signals as a unified multisensory stream. Using ACE, we build ACE-Data-0, comprising 150 hours and 17M video frames across 200 task categories, performed by 50 participants in 2 environments, for a total of 75,000 interaction episodes. The dataset spans atomic manipulation, long-horizon chains of household activities, and human-scene interaction, while preserving natural behavioral variation through goal-level rather than step-by-step instructions. We further introduce a hierarchical benchmark that progresses from signals to scene components and then to interactions. Evaluations of state-of-the-art methods expose substantial gaps under contact, occlusion, egomotion, and long temporal horizons. ACE-Data-0 provides synchronized human demonstrations with aligned perceptual, kinematic, and contact supervision, offering a scalable foundation for imitation learning, world models, vision-language-action systems, and embodied AI.

Role-playing agents (RPAs) have become one of the most important consumer applications of large language models. Users engage in multi-turn conversations with RPAs for experiences such as emotional comfort, making reliable evaluation essential for measuring capability, comparing systems, and guiding further improvement. Existing benchmarks, however, typically require an RPA to continue a fixed dialogue history and then evaluate the continuation using a fixed rubric detached from the user. We identify and empirically demonstrate two limitations of this design. First, an RPA's output is shaped by the preceding dialogue history, preventing a scientifically grounded assessment of its role-playing ability in real multi-turn settings. Second, user experience varies substantially across individuals, and conventional fixed rubrics need not align with user satisfaction. We therefore introduce PALATE (Person-Aligned LLM-Simulated-User Assessment with Tailored Evaluation), a scalable RPA benchmark built on user simulators. PALATE is accompanied by a pool of 300 character profiles. Its main evaluation trains five per-user simulators and lets them engage candidate RPAs in free-form, multi-turn conversations over a pre-frozen panel of character profiles. Alongside a general quality rubric, we construct personalized rubrics to measure user satisfaction; on held-out annotated data, the personalized rubrics show higher agreement with human judgments than the general rubric. In the main evaluation of 16 candidates, PALATE separately characterizes generic turn quality, long-horizon session capability, and per-user experience on multi-turn trajectories co-constructed by each candidate. It thereby produces interpretable evaluations of specific user-RPA pairs rather than compressing systems into a single user-independent ranking.

Existing video captioning models generate natural descriptions of video content but cannot explicitly ground local visual elements to multiple reference images. We introduce multi-reference image-grounded video captioning, a new task requiring factual video descriptions with phrase-level reference grounding, and propose RefCaptioner, a two-stage post-training framework for this task. RefCaptioner combines mixed-data SFT with Hierarchical Coverage-Discounted GRPO to jointly improve reference selection, phrase-level binding, distractor rejection, and cross-reference consistency while preserving general video-captioning ability. To support training, we construct a corpus containing 20,000 videos and 171,354 reference images. We further introduce MRVBench, a benchmark for evaluating caption factuality and multi-reference grounding on both real-world and AI-generated videos. Experiments show that RefCaptioner achieves the best overall performance among the open-source models while remaining competitive on standard video captioning benchmarks. Human evaluation further confirms that its captions are preferred by annotators and enable more source-faithful video reconstruction with both open-source and proprietary video generators.

Multimodal large language models increasingly use sketches, annotations, tools, and intermediate images during reasoning, but it remains unclear whether they truly rely on these visual states. Existing benchmarks are limited both by task collections with narrow coverage or partially text-solvable samples and by evaluations that emphasize final answers without diagnosing how intermediate visual states are generated, rendered, and used. We introduce See2Think, a unified evaluation framework comprising See2ThinkBench and Visual Action-of-Thought (VAoT). See2ThinkBench contains 1,200 open-ended, visually dependent problems across 12 task categories spanning 2D structured, 3D scene, and real-world reasoning. VAoT records textual thoughts, visual actions, rendered states, and subsequent reasoning under four controlled inference settings. Evaluating representative proprietary and open-source multimodal models, we find that visual reasoning is strongly model- and environment-dependent, with no single setting consistently dominating across tasks. Process analysis further shows that models usually select relevant visual operations, while faithful rendering remains the clearest bottleneck and high feedback uptake does not necessarily translate into accuracy gains. Under task-relevant corrupted feedback, models exhibit behavioral dependence on visual states, with accuracy dropping by over 10 percentage points in controlled interventions.

Financial Times : Filing: Amazon has completed its $50B investment in OpenAI; source: OpenAI received the final tranche this week, taking Amazon's position to roughly 5% —  Ecommerce giant has roughly 5% stake in AI lab after equity deal  —  Amazon has completed a $50bn investment in OpenAI …

Spencer Soper / Bloomberg : Amazon shares closed up 15.32% on Friday, the biggest one-day jump since April 2012, after the company reported accelerating AWS revenue —  Amazon.com Inc. shares surged the most since 2012 after the company reported accelerating cloud-computing revenue, easing investors' concerns …

Kit Norton / Barron's Online : Roblox shares close down ~27%, its worst-ever daily percentage decline, after forecasting falling bookings and slowing revenue growth; RBLX is down 40%+ YTD —  Choose Barron's as a preferred source of financial news  —  Key Points  —  Shares of Roblox tumbled Friday as Wall Street responded harshly …

Reuters : Sources: OpenAI has discovered other instances where AI agents escaped containment; none of the agents were thought to have left OpenAI's network —  OpenAI has discovered other instances in which autonomous agents have escaped containment as the company expands its investigation …

James Rundle / Wall Street Journal : FBI and EPA say water and wastewater utilities in at least seven states reported cyberattacks this week, with some causing flooding and other operational issues —  Water plants in at least seven U.S. states have reported attacks this week, with some affecting operations

Ben Szalinski / Capitol News Illinois : Illinois Governor JB Pritzker signs the Children's Social Media Safety Act, a bill designed to limit children's access to “addictive” social media algorithms —  Cracking down on social media algorithms one of governor's top priorities this year  —  Article Summary

Wall Street Journal : An in-depth account of the past two days inside Situational Awareness, as Aschenbrenner assured investors he would focus on changes to his portfolio management —  Leopold Aschenbrenner was hailed as the ‘Nostradamus of AI.’ But his Situational Awareness hedge fund took on too much leverage, leading to a crash he never saw coming.

Geoff Brumfiel / NPR : Google rolls back an image generation tool in Google Earth to add “stronger guardrails” after concerns arose it can be used to create deepfake satellite imagery —  A day after Google unveiled a new feature that allowed users to create AI-generated satellite images with the click …

Mark Bergen / Bloomberg : Index Ventures has raised $2B in new capital: a $900M venture vehicle, $400M for seed financing, and an addition of $700M to a 2024 $1.5B growth fund —  Index Ventures, an investor in cybersecurity firm Wiz and fintech Revolut, has raised $2 billion in new capital to back startups across multiple stages of development.

Lauren Forristal / TechCrunch : Snap says its recommendation systems will be adjusted so only videos created by real people, not AI-generated ones, are eligible for Spotlight recommendations —  Snapchat is the latest company to strengthen its stance against AI slop.  The social network platform will no longer reward fully …

Oliver Barnes / Financial Times : Sources: NXP is in talks to buy Ambarella, which has a market cap of ~$3.3B and makes image-processing chips for security cameras and self-driving cars —  Deal would land amid consolidation wave sweeping semiconductors as they position for AI revolution  —  NXP Semiconductors …

Katherine Hamilton / Wall Street Journal : OpenAI says its models now have more than 1B active users and are used by more than 2M businesses —  The announcement comes after OpenAI said earlier this week it was cutting the price of several of its models  —  OpenAI has more than one billion active users and two million businesses using its artificial intelligence models.

Reuters : Trump says he does not think Iran is behind a cyberattack on Minnesota water systems: “I think I blame it on Minnesota because they're grossly incompetent” —  U.S. President Donald Trump at Camp David told reporters that he does not think Iran is behind a cyberattack on Minnesota water systems …

Todd Bishop / GeekWire : Filing: Microsoft's product R&D roles declined for the second straight year as the company's total FY2026 headcount fell for the first time since 2016 —  The number of product research and development roles at Microsoft declined for the second straight year, according …

Tim De Chant / TechCrunch : SpaceXAI says it will remove 69 gas turbines used to power the Colossus data centers near Memphis by July 2027, as it moves to a 1.2 GW natural gas power plant —  SpaceX said on Thursday that it will remove the unpermitted turbines powering its xAI data centers near Memphis as it transitions to a permanent …

前一段时间未知身份的攻击者创建了一系列新账号，然后通过这些账号接管无人维护的 Arch User Repository(AUR)孤儿包，植入恶意代码，推送恶意更新。Arch Linux DevOps 团队现在宣布关闭 AUR 孤儿包的领养和修改功能。AUR 中有逾 107,000 个软件包，其中近 14,000 个无人维护可供认领，此前任何注册用户都可以认领和修改这些无人维护的软件包。在攻击发生之后，Arch Linux 关闭了新用户注册功能，后注册功能于 7 月 13 日恢复。对恶意负荷的分析显示，攻击者的目的是通过植入远程访问木马窃取用户数据。

sroorz 写道： 在初代 Apple Silicon 发布六年后，CrossOver 宣布推出原生 Apple Silicon 支持，并发布了首个预览版。该软件旨在在 macOS 上运行 Windows 游戏和软件。 苹果于 2020 年推出 Apple Silicon，并内置 Rosetta 2 转译层以兼容旧有 Intel 软件。CrossOver 和其他基于 Wine 的模拟器一直依赖 Rosetta 2 在 Apple Silicon 上运行 x86 Windows 游戏。然而，苹果计划逐步淘汰 Rosetta 2，并于 macOS 28 完全停止支持。CrossOver 的原生 Apple Silicon 支持正是为了应对这一变化，实现独立于 Rosetta 2 的 x86 Windows 游戏支持。 CrossOver 产品经理 Meredith 在发布公告中回顾了 CodeWeavers 开发团队过去四年如何通过改进 Wine 和相关开源组件、工具链实现这一目标。她也指出，这是首个预览版，仍存在一些问题，目前仅推荐测试使用。此外，由于 ARM64 版 D3DMetal 尚未就绪，目前仅能通过 DXMT ARM64 运行 DirectX 11 游戏。但所有这些问题都将在正式版 CrossOver 27 中得到解决。 Wine 是一个开源项目，是除虚拟机外，在非 Windows 操作系统上运行 Windows 游戏的唯一解决方案。基于 Wine 衍生出的项目中最著名的是 Steam Proton。CodeWeavers 是 Wine 项目的长期贡献者和 Proton 开发者，其员工贡献了 Wine 90% 以上的代码，CrossOver 是其基于 Wine 开发的商业版。

Google 官方博客称，LLM 正在解锁自动化漏洞发现的能力，使得漏洞发现的速度以及数量都远超过去，因此需要采用新的方法应对这一新形势。它的一项应对之策是开发名为动态补丁（dynamic patching）的新功能。该功能可以在大多数情况下让 Chrome 浏览器无需重启就打上安全补丁。利用 Chrome 的多进程架构，动态补丁可以动态地将后台子进程如渲染器和 GPU 的二进制文件逐一替换为更新后的版本。

科学家首次发现了基因构成在影响癌症风险上发挥巨大作用的直接证据，遗传的基因与后天获得的基因突变相互作用塑造了癌症的演变方式。研究人员培育出四种不同肝癌易感性的小鼠，然后暴露在单剂量肝致癌物二乙基亚硝胺(diethylnitrosamine，DEN)下，DEN 存在于烟草烟雾和部分加工食品中，会损伤肝细胞 DNA，导致基因突变，引发肿瘤生长。所有小鼠都在相同年龄且在严格控制的条件下接受了相同剂量的 DEN，研究人员得以消除环境因素的干扰。结果显示，致癌突变几乎所有小鼠都是一致的，激活了相同的促癌信号通路 MAPK 通路。但遗传背景的差异使得癌症风险显著不同。研究首次揭示了遗传背景对突变过程和肿瘤发展通路的影响程度。

根据发表在《科学》期刊上的一项研究，对智利木乃伊 DNA 的分析揭示了美洲已知最早的天花古代毒株。这些发现提供了迄今为止最清晰的基因学证据：欧洲殖民将这种疾病带到了新大陆。天花是欧洲殖民时期传入美洲的最致命疾病之一。该疾病在北美、中美和南美的迅速蔓延估计导致三至四百万人丧生，尤其是在那些此前对该病没有免疫力的原住民群体中。研究人员在智利北部出土的两具印加至早期殖民时代原住民的木乃伊体内发现了古天花病毒 DNA；这些个体生活于公元 1492 年至 1631 年间。这两个古病毒基因组几乎完全相同，表明这些个体很可能在同一次疫情爆发期间被感染，或被同一种流行毒株感染；这为殖民时期南美洲流行的某天花毒株提供了罕见且直接的分子证据。

JAXA 宣布，探测器“隼鸟2号”抵近小行星鸟船（Torifune）表面约 400 米处，实施了高速掠过并进行观测。这一距离创下人类历史上对太阳系天体的最接近纪录。隼鸟2号于7月5日18时30分成功对距离地球约1亿公里的鸟船实施了飞掠探测。以约5.3公里/秒的速度（相对速度）掠过。JAXA 分析确认, 飞掠时探测器与小行星表面最近距离约 400 米，距离小行星中心约 744 米。此前飞掠探测的最近纪录被认为由中国探测器“嫦娥二号”保持，当时嫦娥二号抵达距小行星表面约 770 米处。对观测数据的分析结果显示，推测鸟船长轴约 840 米、短轴约 340 米，平均直径约为 540 米。

边开车边听广播；边做早餐边和家人聊天；边吃三明治边看书。我们的日常生活建立在多任务处理之上，但大脑真的能同时做两件事吗？还是在两个任务之间快速切换？研究发现，对于高认知任务，大脑是在任务之间切换；而对于部分熟练任务，大脑会进入某种自动驾驶模式，也就是接近多任务处理。有十年驾龄的老司机，其大脑不会像以前那样注重每一个细节。根据发表在《Journal of Cognitive Neuroscience》期刊上的新研究，在经过大量练习之后，大脑能重路由部分任务的回路，减轻精神负荷，改进多任务处理能力。前额叶皮层是大脑主要的计算中心，但通常一次只能处理一件事，这种现象被称为“额叶瓶颈（frontal bottleneck）”。在实验中，研究人员训练参与者完成分类任务，在训练前后扫描了大脑。任务一开始参与者激活了前额叶皮层。但在训练之后部分任务被转移到了颞叶皮层。颞叶皮层负责记忆编码和目标识别。在随后展开的双任务实验中，将更多分类任务转移到颞叶皮层的参与者表现更出色。研究人员说，大脑通过改变回路能同时做两件事。

科学家已经确认，现代人类/智人大约 5 万年前走出非洲后，曾与尼安德特人和丹尼索瓦人杂交，他们的部分 DNA 至今仍留在现代人体内。这些发现得益于从化石中提取测序的古 DNA。但基因组中还有更早的混血痕迹，而那些已经灭绝的古人类，没有留下任何可供测序的遗传物质，他们对基因的贡献难以辨认。加州伯克利团队开发出通过祖先重组图估计追踪古老贡献（TRACE）的技术。它的原理是利用世界各地现代个体的全基因组数据，重建基因组中的家谱关系，即“祖先重组图”，描绘 DNA 片段如何经由共同祖先一路传递至今。研究人员在现代人类基因组中定位了来自两个未知远古祖先的 DNA 区域——一个是分化于约 80 万年前的“幽灵祖先”，“幽灵祖先”存在于所有现代人类之中，而非非洲人独有——每个现代人约有 0.5%-1% 的基因组来自这一谱系，规模与尼安德特人的贡献大体相当。另一个是距今 180 万年的“超级古老祖先”，它并没有直接与现代人类杂交，而是先与欧亚大陆的丹尼索瓦人混血。丹尼索瓦人后来又与智人通婚，将其中一小部分超古老 DNA 带进了现代人类。

2026年7月27日，The Peterman Pod主持人瑞安·彼得曼（Ryan Peterman）发布了对朱迪亚·珀尔（Judea Pearl）的长篇访谈。话题从他的科学启蒙、超导存储器与早期人工智能一路延伸到贝叶斯网络、因果阶梯、大语言模型（LLM）和通用人工智能（AGI）。珀尔没有从模型榜单和产品能力出发评价LLM。他把问题放回了一条延续数十年的研究线索：机器究竟怎样把观察变成判断，又怎样从相关性走向行动与解释。

为了减少电子垃圾，欧盟的新法律将要求在欧洲销售的电子产品能更换电池，这一法律将于 2027 年 2 月生效。为了遵守该法律，任天堂已经宣布在欧洲销售的 Switch 2 游戏机以及手柄将允许更换电池。现在 PC 外设制造商罗技透露在欧洲销售的新鼠标将可以更换锂电池。支持更换电池的新鼠标将只在欧洲销售。罗技新任游戏业务负责人 Robin Piispanen 声称，新款欧洲版鼠标不仅制造成本更高，而且更重、更结实，以最大限度降低意外刺伤的风险。

知名游戏设计师 Ron Gilbert 宣布他已经开始着手开发《银莲花公园（Thimbleweed Park）》的续作，预计 2028 年初推出。目前 Steam 版的《银莲花公园》正在打折，续作的页面已经上线。Ron Gilbert 是最有影响力的游戏设计师之一，他在 LucasArts 工作期间设计了 SCUMM 引擎，在此基础上开发了一系列经典冒险游戏。他在 1992 年离开 LucasArts，创办了自己的游戏公司。2014 年他与 Gary Winnick 等人在 Kickstarter 上发起了《银莲花公园》的众筹，尝试复兴 LucasArts 风格的点击冒险游戏。在成功众筹后游戏于 2017 年推出。

根据中国国家能源局公布的数据，今年前 6 个月煤电发电量为 2.5 万亿千瓦时，占总发电量的比重降至 49.7%，半年发电量占比首次低于 50%；可再生能源发电量占全部发电量的 41.2%，占比首次超四成。其中风电、太阳能发电总量突破 1.2 万亿千瓦时，约占全社会用电量的四分之一。2025 年中国用电量历史性突破突破 10 万亿千瓦时，其规模已是美国的 2 倍多，高于欧盟、俄罗斯、印度、日本总和。

科学家通过激光雷达（LiDAR）航空勘测技术检测到了亚马孙地区西南部 432 处被称为“土墩”的地表遗存。这些发现的地理位置此前曾是古代亚马孙西南部 Aquiry 文明的定居地，这一地区可能有 24,000-30,000 处土墩。古代 Aquiry 文明留下的大型地面几何形沟槽围合结构，称为亚马孙地雕，人们认为其功能兼具仪式、政治和公共集会中心。这些大多单独分布的结构由沟渠和堤坝组成，占地在 0.35-14 公顷不等，已知最大的地雕面积近 50 公顷。最新检测结合现有卫星研究，研究者估计 Aquiry 文明在公元 100-300 年鼎盛时期可能养活了约 125-300 万人口，意味着它对亚马孙的土壤、森林结构和生物多样性产生的影响比此前认为的更为深远。 Aquiry文明所占区域不到大亚马孙地区的 3%。

炸薯条是广受欢迎的美食，但并不健康，因为薯条在油炸过程中吸收了大量油脂，增加了脂肪和热量，而经常食用高脂肪食物会导致肥胖和高血压等健康问题。现在 Illinois Urbana-Champaign 大学的研究人员发表了两篇论文，报告了结合微波油炸和传统油炸让薯条更健康的方法。这种方法缩短了烹饪时间，减少了吸油量，同时能保持传统炸薯条酥脆外皮和口感。研究人员将土豆冲洗去皮切成条状，然后焯水、撒盐，在加热至 180 摄氏度的豆油中炸制。他们发现，光靠微波油炸无法形成酥脆的表面，需要结合传统油炸。研究人员建议将两者结合起来，用微波减少油脂吸收，用传统方法保持表面酥脆。

辣椒的活性成分辣椒素能与监测极端环境热源的受体 TRPV1 结合，模拟强烈的灼烧感，欺骗神经系统，激发身体强烈反应，但并不会真的在身体上造成什么伤痕。医生已在利用这种局部化学反应去缓解慢性疼痛。根据发表在《Physiology & Behavior》期刊上的一项研究，重庆师范大学的研究人员发现辣椒有助于减轻极端疼痛。研究人员从学校招募了 48 名无痛史的健康年轻人，让学生口含掺有辣椒粉的明胶块五分钟，然后用激光红外热脉冲照射手背，体会炙热的疼感。激光强度有轻重之分。结果发现，相比对照组，辣椒对轻度照射没产生有区别的影响，但显著降低了重度照射产生的疼痛感。