OrangeBot.AI Digest — 2026-07-16

AI account executives that demo and close B2B deals

Turn any goal into a personalized, adaptive learning path.

Build AI-driven workflows across 1,000+ apps

Tactile controls for your Codex agents

Your private home for every lab result

Email for AI agents. They sign up, send and reply.

Turn a Reddit complaint into your next company

Agentic AI for Healthcare Triage and Care Automation

Analytics that lives inside your AI assistant

Turn company operations into a living map for agents

Hybrid academic search over 236M papers, built for agents

Think out loud and watch it become a living map.

Your context, available to every agent.

Build and hire autonomous AI employees from a single prompt

Empty your backlog. Pay only when it ships

The capability of a modern AI agent depends not only on its foundation model but also on its harness, which constructs prompts, manages state, invokes tools, and coordinates execution. As models, APIs, environments, and requirements evolve, the harness must be continually modified. Before such a change can be made, a developer or coding agent must identify all code locations that implement the target behavior. This is difficult because production harnesses are large, tightly coupled, and behaviorally distributed, while modification requests describe what the system should do and repositories are organized by files and modules. Code search, repository indexing, and long-context processing ease inspection, but still leave this behavior-to-code mapping to be recovered by hand. Behavior localization is therefore a central bottleneck in harness evolution. We introduce the Harness Handbook, a behavior-centric representation synthesized automatically from a harness codebase via static analysis and LLM-assisted structuring, linking each behavior to its corresponding source. We also introduce Behavior-Guided Progressive Disclosure (BGPD), which guides agents from high-level behaviors to relevant implementation details and verifies candidate locations against the current source. On diverse modification requests from two open-source harnesses, Handbook-Assisted planning improves behavior localization and edit-plan quality while using fewer planner tokens, with the largest gains on scattered sites, rarely executed paths, and cross-module interactions. Evolving complex agentic systems thus depends not only on generating edits, but also on determining where those edits should be made.

We introduce Boogu-Image-0.1, an open-source unified multimodal understanding and generation model family, comprising Base, Turbo, Edit, and Edit-Turbo variants. It delivers competitive performance in high-quality text-to-image generation, fast inference, instruction-based editing, and bilingual (Chinese-English) text rendering. Closed-source multimodal systems like Nano-Banana-Pro and GPT-Image-2 achieve strong performance through system-level integration rather than a single model, yet their internal practices remain largely undisclosed. In this work, we demonstrate that targeted improvements in model understanding, data quality, and training pipelines, coupled with agentic inference-time scaling, can substantially enhance generation and editing performance even under highly constrained compute budgets. Comprehensive evaluations show that Boogu-Image-0.1 consistently matches or surpasses other open-source models across standard benchmarks, and achieves results approaching leading closed-source systems. Notably, this is accomplished with only 208.62 million unique images. The base model's theoretical training cost is only approximately \$400K. We share practical discussions that we believe are valuable to the broader research community, and release weights, code, and recipes under Apache 2.0 to advance the open ecosystem for unified multimodal understanding and generation. Our code is available here: https://github.com/Boogu-Project/Boogu-Image.

Reinforcement learning with verifiable rewards without human-annotated data, often referred to as zero RL, has emerged as a powerful paradigm for eliciting chain-of-thought reasoning. However, due to computational constraints, existing studies are largely restricted to small models, leaving the training dynamics and emergent capabilities at a large scale unexplored. To meaningfully explore this frontier, we aim to elicit high-quality reasoning behaviors from the model. However, we find that naive scaling often suffers from poor readability, token redundancy, and a lack of adaptive reasoning depth. To address these challenges, we present a stable and efficient training pipeline, incorporating algorithmic and system optimizations such as clipped importance sampling, training-inference ratio correction, and mixed-precision control. Our experiments offer three key findings that validate the "bitter lesson" of scaling: (1) scaling to 1T parameters significantly enhances sample efficiency and performance ceilings; (2) the training process progresses sequentially through an initial discovery phase followed by a sharpening phase; and (3) the model spontaneously develops advanced cognitive behaviors, including anthropomorphism, structured formatting, self-verification, parallel reasoning, and context anxiety, rendering hand-crafted heuristics redundant. Evaluated on seven mathematical benchmarks, Ring-2.5-1T-Zero achieves competitive performance. Additionally, to assess CoT quality beyond final-answer correctness, we propose a structured evaluation framework across three dimensions: comprehensibility, reproducibility, and efficiency, where our model demonstrates clear advantages in producing structured and concise reasoning traces. By sharing our observed emergent phenomena, we hope to provide the community with deeper insights into scaling behaviors, particularly at the 1-trillion scale.

OpenClaw has emerged as a leading agent framework for complex task automation, yet it faces insufficient cross-platform GUI interaction support and a well-built self-evolution mechanism. These flaws limit its adaptation to diverse device ecosystems and prevent performance improvements through continuous learning from execution experience. To resolve these issues, we propose the Know Deeply, Act Perfectly paradigm for personal assistants, which holds that accumulated user interaction and task-running experience directly improve execution accuracy and efficiency, unifying cognitive comprehension and operational execution. Based on this paradigm, we introduce KnowAct-GUIClaw, a novel Know-Route-Act-Reflect framework designed to address OpenClaw's GUI manipulation deficits and break through its cross-platform and recursive self-improvement constraints. First, the host agent leverages accumulated interaction experience and task-relevant knowledge for long-horizon task decomposition and allocation (Know). Second, a pluggable GUI subagent with an experience-attributable memory system (Know) and self-evolving skill library (Act), enabling seamless cross-platform migration and fast-path integration. Especially, this framework continuously stores user profiles and feedback to improve the accuracy of task decomposition and tool calls. Extensive experiments across Android, iOS, HarmonyOS and Windows show that KnowAct-GUIClaw achieves superior efficiency, accuracy and cross-platform adaptability. Especially, the GUIClaw with open-source Kimi-2.6 models achieves the best performance (64.1%) on the long-horizon MobileWorld benchmark, beating all agentical frameworks and closed-source agentical models, e.g., Seed-2.0-Pro and GPT-5.5. Additionally, the knowledgeable memory and execution skills supported by our framework are transferable across diverse base models, improving by 8.5% with Kimi-2.6.

We introduce OvisOCR2, a 0.8B document parsing model. OvisOCR2 is designed as an end-to-end parser: given a document page image, it generates a Markdown representation in natural reading order, covering text, formulas, tables, and visual regions. We build a data engine that combines filtered real-document annotations with synthetic pages whose rendered images and Markdown targets are derived from the same HTML source. The training recipe includes supervised fine-tuning, reinforcement learning on a 4B branch with a multi-component reward design, on-policy distillation into the 0.8B model, and model fusion. On OmniDocBench v1.6, OvisOCR2 achieves a state-of-the-art overall score of 96.58, placing an end-to-end model at the top of this leaderboard previously dominated by pipeline methods and highlighting the potential of end-to-end document parsing. On PureDocBench, OvisOCR2 also achieves the highest Avg3 score of 75.06. Beyond these two public benchmarks, we evaluate OvisOCR2 on an in-house benchmark designed to cover a broader set of long-tail and challenging scenarios. OvisOCR2 obtains the best overall performance among the compared methods, providing further evidence of its generalization and robustness. OvisOCR2 is available at https://huggingface.co/ATH-MaaS/OvisOCR2.

Image guardrails are typically trained and evaluated under a fixed safety policy, implicitly treating safety as an intrinsic property of an image. Real deployments are different: the same image may be allowed in one product, restricted in another, and newly disallowed when a policy boundary changes. We study policy-adaptive image guardrailing, where a model must decide whether an image violates the currently supplied policy and generalize to held-out policy definitions. We introduce PolicyShiftBench, a comprehensive benchmark with 2,000 policy-discriminative instances over 265 images, where each image is paired with 7.55 policy-conditioned prompts on average to test whether models adapt to the active policy rather than relying on image-level safety priors. We then propose PolicyShiftGuard, a compact policy-conditioned guardrail trained with a two-stage training recipe that combines Randomized Policy SFT (RP-SFT) with Boundary-Pair Policy Adaptation (BP-Adapt). BP-Adapt trains matched prompts for the same image and risk category using standard label supervision and a pairwise comparison loss that separates blocking policies from passing policies. Experiments show that existing VLMs and specialized guardrails remain brittle under policy shifts, while PolicyShiftGuard substantially improves policy-sensitive performance. The 7B model achieves SOTA performance of 76.9 Avg. F1 and 72.1 Avg. PSS on PolicyShiftBench, transfers well to UnSafeBench and SafeEditBench, and improves the latency-performance trade-off with a concise output format. Ablations confirm that matched pass/block boundary pairs are essential for stable policy adaptation.

Current visual generation models are capable of producing high-quality content, yet they lack a coherent perception of the spatial structure. Existing generative novel view synthesis methods typically introduce explicit geometry priors, which enforce spatial consistency but inherently restrict generalization in large view changes. In contrast, recent interactive generative methods favor implicit scene modeling, offering greater flexibility at the cost of precise camera control and geometry consistency. In this paper, we propose MetaView, a diffusion-based monocular novel view synthesis framework that enables rendering under large view changes from a single image. Our key insight is to combine implicit geometry modeling with minimal yet essential explicit 3D cues: we incorporate implicit geometry priors from a feed-forward geometry perception network to regularize structure without imposing restrictive reconstruction pipelines, while leveraging metric depth to anchor the generation to a metric scale. This design allows MetaView to achieve both geometry consistency and precise controllability. Extensive experiments demonstrate that, under challenging monocular large viewpoint changes, MetaView significantly outperforms existing methods and exhibits superior generalization. Our code is publicly available at https://github.com/KlingAIResearch/MetaView.

World Action Models (WAMs) improve robot policy learning by jointly modeling actions and future visual observations, using future scene evolution as dense supervision for physically grounded action generation. However, a common design in existing WAMs is to explicitly generate future videos at inference time, incurring substantial computational overhead and hindering real-time closed-loop deployment. GigaWorld-Policy addresses this issue with an action-centered formulation, where future visual dynamics are used during training while action-only decoding is used at inference time. Building upon this framework, we present GigaWorld-Policy-0.5, an enhanced action-centered WAM designed for more efficient robot control. During pretraining, GigaWorld-Policy-0.5 adopts a mixed Action-Conditioned World Modeling (AC-WM) and WAM training strategy. This strengthens the coupling between visual dynamics and robot actions and improves the transferability of action representations for downstream policy learning. For efficient inference, GigaWorld-Policy-0.5 introduces a Mixture-of-Transformers architecture that separates visual dynamics modeling and action generation into specialized experts, reducing active computation during action-only inference and achieving 85 ms inference latency on a local RTX 4090 setup. In addition, we employ an agent-based AutoResearch pipeline to systematically search training configurations, enabling more efficient identification of optimal experimental setups while reducing the time and manual intervention required for hyperparameter tuning. Experiments and ablations show that GigaWorld-Policy-0.5 preserves the training benefits of future visual dynamics while improving inference efficiency for robot control.

Vision Transformers (ViTs) are known to exhibit high-norm patch-token outliers that degrade feature map quality, a problem effectively mitigated by register tokens. As diffusion models increasingly adopt transformer architectures and move toward pixel-space training, they become closer in form to ViTs, raising the question of whether register tokens are also useful for Diffusion Transformers (DiTs). In this work, we show that DiTs differ from ViTs in a key respect: they do not exhibit patch-token outliers but still benefit from registers. Interestingly, registers are more effective in pixel-space DiTs than in latent-space DiTs. By analyzing intermediate representations, we find that register tokens produce cleaner feature maps at high noise levels, which may contribute to their effectiveness in pixel-space generation. We further observe that recent pixel-space DiT architectures implicitly incorporate register-like mechanisms, which may partially account for their strong empirical performance. Motivated by these observations, we propose Register Guidance, a technique that amplifies the contribution of register tokens responsible for improving visual structure and coherence.

Self-improving autonomous agents are moving from research prototypes to deployed systems. The primary goal is controllable evolution, or adaptation, from experience with minimal or even no human input. This survey frames modern self-improving agents as adaptive systems that convert experience into accumulated capability gains. We offer a system-level framework that represents a modern agent as a configuration coupling a foundation model with an operational scaffold of prompts, memory, tools, and control logic. Within this framework, self-improvement is formalized as a self-induced update operator that obtains and commits updates to model parameters or scaffold components. We organize prior work by update target and by the signals that drive change, then review applications and discuss evaluation, before closing with open problems and future directions. For convenience, we track technical updates on https://github.com/selfimproving-agent/awesome-Self-Improving-Agents.

While recent advances in 3D generation have enabled impressive visual synthesis, existing methods often rely on 2D diffusion supervision without explicit mechanisms for geometric consistency, leading to spatial hallucinations such as duplicated structures and misaligned geometry. These issues become more severe in 4D generation, where maintaining consistency across viewpoints and temporal evolution introduces additional challenges, including jitter, identity flicker, and structural drift. We present Hallo4D, a unified and model-agnostic framework for mitigating spatiotemporal hallucinations in 3D and 4D content generation. Hallo4D introduces a generation-detection-correction paradigm that leverages large multimodal language models (LMMs) to identify and summarize spatial and temporal inconsistencies from multi-view and multi-frame renderings. These insights guide a consensus-driven image-space consistency optimization, where an LMM-based selector evaluates candidate corrections through multi-model voting, without requiring retraining or architectural modifications. To further improve temporal consistency and optimization efficiency, Hallo4D incorporates motion-aware keyframe sampling, LMM-guided initialization, and appearance alignment. We additionally introduce exposure-aware optimization and visibility pruning to enhance robustness under challenging viewpoints. Extensive experiments demonstrate that Hallo4D consistently outperforms strong baselines across diverse 3D and 4D generation settings, providing a scalable and generalizable solution for consistency-aware content generation.

Structured pruning is a hardware-friendly way to compress LLMs, but it is mostly validated on multiple-choice recognition tasks, while the same compressed checkpoints can collapse on the free-form generation that deployment actually requires. Two observations trace this gap. First, greedy pass@1 nearly vanishes after compression, yet pass@k recovers substantially under repeated sampling: useful generations are demoted, not erased. Second, the recoverable regime fails mainly through suffix repetition. Recovery should therefore train on the compressed model's own on-policy states with dense token-level supervision, which On-Policy Distillation (OPD) provides by reusing the pre-compression model as a frozen teacher. However, long on-policy rollouts spend early recovery budget on low-information repetitive suffixes, delaying loss descent. To mitigate this waste, we propose \shortopd, a short-to-long OPD schedule that detects teacher-confirmed repetitive suffixes, treats the surviving prefix as each rollout's effective length, and allocates future rollout budgets to the effective lengths the policy can currently use. Across math, code, and open-ended generation, \shortopd\ raises the compressed model's score to about 9times its unrecovered value and 1.6--4.4times standard recovery recipes (SFT w/o KD, KD, and SeqKD), and it matches a fixed 8192-token rollout horizon within two points using a quarter of the training time (8.5 vs.\ 35.9 hours) and 71% fewer rollout tokens. We hope this recipe helps move structured pruning beyond marginal gains on perplexity and multiple-choice benchmarks, a step closer to deployment-ready generation quality.

As Large Language Models (LLMs) evolve into autonomous agents, the need for unified evaluation infrastructure becomes critical. However, current evaluation pipelines remain highly fragmented and tightly coupled, hindering reproducibility and causing redundant engineering. To address this, we introduce AgentCompass, an open-source, lightweight, and extensible infrastructure for evaluating LLM-based agents. AgentCompass organizes the evaluation process around three independent components, namely Benchmark, Harness, and Environment, thereby enabling flexible configurations without requiring the reimplementation of complex execution logic. Furthermore, it features a fault-tolerant asynchronous runtime and comprehensive trajectory analysis tools to transparently diagnose nuanced failure modes like reward-hacking. Natively supporting over 20 benchmarks across five capability dimensions, AgentCompass provides the community with a scalable and reproducible infrastructure for advancing agent research.

When should an intelligent assistant speak up without being asked? Continuous egocentric video offers rich, evolving context that enables a new form of assistance: one that is proactive rather than merely reactive. Yet existing approaches either wait passively for user queries or treat every detected event as requiring a response, without considering the user's history, current activity, or whether assistance would actually be welcome. We reframe proactive assistance as a context-dependent decision problem: the agent must not only perceive what is happening, but reason over accumulated temporal context to determine when and whether to intervene. To this end, we present Vinci2, a proactive egocentric assistance system that advances the on-device assistant Vinci from reactive response toward proactivity. On the evaluation side, we present EgoServe, the first large-scale benchmark for proactive assistance in continuous egocentric video. EgoServe comprises over 3,000 service instances organized along 4 temporal memory horizons, ranging from immediate safety alerts to long-term habit coaching, across 10 service categories. On the modeling side, we propose EgoMemo, a training-free, memory-augmented agent that maintains three complementary memory representations: multi-scale temporal summaries, a semantic knowledge graph, and visual embedding archives. At each timestep, EgoMemo performs retrieval-augmented reasoning to determine whether assistance is warranted and, if so, produces contextually grounded responses. Experiments demonstrate that EgoMemo establishes strong baselines on EgoServe while remaining competitive on existing egocentric benchmarks. Our benchmark and code are publicly available at https://sitonggong.github.io/EgoServe-page/{Vinci2}.

Failure attribution for LLM-based agentic systems, i.e., identifying which steps in a failure trajectory caused the task to fail, is critical for debugging and improving these systems. Existing approaches either rely on prompting-based pipelines, which are computationally expensive, or require post-training on failure trajectories with step-level error annotations, which are costly to collect and difficult to scale. We argue that a practical failure attribution model should be lightweight and trainable without step-level supervision on failure data. To this end, we address unsupervised failure attribution, i.e., training exclusively on successful trajectories and identifying error steps at inference time given a failure trajectory. We propose OAT, which casts this problem as one-class learning with neural controlled differential equations, modeling the dynamical pattern of successful trajectories in latent space. At inference time, each step in a failure trajectory is assigned an anomaly score based on its deviation from the dynamics learned on successful trajectories, which is then used to form a set of error steps. With training on only 100 successful trajectories, experiments show that OAT is 200--5000 times faster than prompting-based baselines, and, at the same time, consistently outperforms them in both in-domain and out-of-distribution datasets with +20% and +7% F1 scores, respectively, demonstrating that OAT is a promising and efficient direction for diagnosing agentic system failures.

Lorelei Smillie / Bloomberg : US Homeland Security says it seized 30k+ SIM cards in June and July in a nationwide operation it claims dismantled infrastructure used in telephone fraud —  US Homeland Security Investigations recently seized more than 30,000 mobile phone SIM cards from sites around the country …

Emma Roth / The Verge : Netflix says roughly 300 titles used generative AI this year, mostly in post-production, “to deliver higher quality output more quickly and at a lower cost” —  Some titles used AI to create ‘enhanced crowds, historical battle sequences, and worldbuilding establishing shots.’

Kashmir Hill / New York Times : An NYT reporter finds AI-generated, unauthorized biographies of herself and other journalists on Amazon, where AI-made books with elusive “authors” proliferate —  Recently, I received a strange text from a new acquaintance.  “You have your own biography???” it read.

Lillian Rizzo / CNBC : Netflix reports Q2 revenue up 13% YoY to $12.56B, vs. $12.59B est., and says it will give fewer engagement updates starting in 2027; NFLX drops 7%+ after hours —  Netflix reported second-quarter revenue and earnings that were in line with analyst estimates on Thursday as Wall Street is keeping …

Michael Nuñez / VentureBeat : Kimi-K3 is now #1 on the Frontend Code Arena benchmark, surpassing Claude Fable 5; the model scored 88.3 on Terminal Bench 2.1, only below GPT-5.6 Sol's 88.8 —  Moonshot AI, the Beijing-based artificial intelligence startup backed by Alibaba, on Thursday released Kimi K3 …

Sarah Perez / TechCrunch : DoorDash launches a limited beta of DoorDash CLI, which lets users place orders via an AI agent, available by waitlist to macOS developers in the US and Canada —  Sudo make me a sandwich.  The future has arrived!  DoorDash just introduced a limited beta of DoorDash CLI …

Sarah Perez / TechCrunch : Google updates its AI video editor Vids with Gemini Omni and a feature that lets users create custom digital avatars using a selfie and voice recording —  OpenAI's Sora may have shut down, but Google apparently thinks there's still interest in a tool that lets you star in your own AI videos.

Kimi : Moonshot AI releases Kimi K3, a 2.8T-parameter AI model that it says rivals Opus 4.8 and GPT-5.5, and plans to release model weights by July 27 —  Today, we are introducing Kimi K3 — our most capable model.  Kimi K3 is a 2.8T-parameter model built on our Kimi Delta Attention and Attention Residuals …

Tibo / @thsottiaux : After reports of GPT-5.6 deleting files, OpenAI says the issue most often occurs in full-access mode without sandboxing and it is working to mitigate the risk —  On file deletions. We've investigated a handful of reports where GPT-5.6 unexpectedly deleted files. What we have found is that this most commonly occurs when: - Full access mode is enabled and codex is run without sandboxing protections, including without auto review being

CoinDesk : Crypto.com raised $400M from Citadel Securities in its first institutional funding round, valuing the Singapore-based crypto exchange at $20B —  Crypto.com secured a $400 million strategic investment from market maker Citadel Securities in a deal that values the crypto exchange at $20 billion …

Bloomberg : Sources: Google is months behind schedule on delivering Gemini 3.5 Pro because the company has been trying to improve its capabilities, particularly in coding —  Alphabet Inc.'s Google is months behind schedule on delivering Gemini 3.5 Pro, its most powerful flagship AI model …

Ivan Mehta / TechCrunch : Beehiiv launches Community, which lets subscribers chat with each other, adds programmatic ads and an AI tool that helps creators manage their audiences —  Newsletter platform Beehiiv is expanding into new avenues of engagement by launching a feature called Community, which lets subscribers of a creator chat with each other.

Karissa Bell / Engadget : Meta's Oversight Board says top AI models may be restricting free expression in its first evaluation of LLMs, as it seeks to expand its influence beyond Meta —  The group is trying to extend its influence beyond Meta.  —  The Oversight Board, the independent content moderation organization created …

Aisha Malik / TechCrunch : Google says users in the US can now link to and interact with some apps in AI Mode, including Instacart, Canva, and YouTube Music —  Google announced on Thursday that it now allows users to link and interact with some of their go-to apps right in AI Mode, the tech giant's conversational search experience.

Jay Peters / The Verge : Roblox unveils Build, which lets users create basic games through text prompts directly within its mobile app, powered by Roblox's models and open-source models —  It's the latest push into AI from Roblox, which has embraced the technology. … Roblox is about to let people make games …

xAI 在 GitHub 上公开了其辅助编程智能体 Grok Build 的源代码，此举可能是某种重新赢得用户信任的补救措施。因为在这之前它被发现存在严重的隐私安全问题，会上传用户的完整代码库。Grok Build 被发现在读取或处理文件时，该文件的内容未经任何编辑就被传输到 xAI 使用的 Google Cloud Storage 中。Grok Build 的数据保留远超 Claude Code、Gemini 和 Codex 等类似工具。有 Grok Build 用户报告，包含 SSH 密钥、密码管理器数据库等的完整用户目录都被上传了。Elon Musk 表示该公司将彻底删除此前上传到服务器上的用户数据。

发表在 PLOS One 上的一项新研究发现，马不依赖声音、气味或过往经验就能识别屏幕上的捕食者。传感器显示，马看到屏幕上的狼后心跳会加速，但面无表情十分镇定。研究显示，马没有摇头，没有摇尾巴，其目光也没有锁定屏幕以表明大脑在处理威胁信息。论文第一作者 Zeynep Benderlioglu 称，马在评估潜在威胁时展现出惊人的认知克制而不是惊吓。在实验中，马看到屏幕上的袋熊时心率平稳，但看到屏幕上的狼——无论狼是在攻击还是梳理——时其心率会显著加快，其中雄性马的反应更强烈。Benderlioglu 说，马的认知处理能力出乎意料的高。它们高度警惕，但威胁并没有出现，所以它们没有表现出任何异常行为，它们正在进行认知评估。”

大型图书出版商 Hachette、Cengage 和 Elsevier 以及作家 Scott Turow 指控 Google 在训练 Gemini 模型过程中未经许可使用了数百万受版权保护图书，声称这是历史上最严重的版权侵犯事件之一。出版商称，Google 挪用了用于 Google Books、Google Play Books 和 Google Scholar 等服务的图书，这些服务允许 Google 以特定方式使用相关图书——如显示可搜索的片段或销售电子书，但无权将这些图书用于训练商业 AI 产品。诉讼书称：“为了维持其在互联网领域的统治地位，Google 放弃了其早期‘不作恶’的座右铭，犯下了历史上最严重的版权侵权事件之一。”诉讼书称，Google 内部早已认识到该问题，它可能会面临“100-1000 亿美元的潜在罚款”，但仍然未经许可将这些图书用于训练 Gemini。

深度求索（DeepSeek）计划启动新一轮融资，目标融资高达 500 亿元人民币，估值高达 4800 亿人民币。该公司也已开始就可能在上交所科创板上市进行初步探讨。该公司的内部目标是计划在今年提交 IPO 申请。今年 6 月，DeepSeek 完成了成立以来的首轮外部融资，募资总额逾 500 亿元人民币，融资后估值约为 3380 亿人民币。首轮融资中，梁文锋个人出资 200 亿元人民币，腾讯控股和电池巨头宁德时代分别出资 100 亿元人民币和 50 亿元人民币，成为最大的外部股东。京东、网易及 IDG 资本各出资 30 亿元，国家人工智能产业投资基金出资 10 亿元。DeepSeek 如此迅速的融资节奏，源于其对资本支出的预期增加。该公司计划建设自有数据中心并采购更多 AI 芯片。

对部分戴尔电脑用户而言，Windows 的周二例行安全更新之后紧跟着的是周三的电脑出问题了。微软确认，由于意外关机、性能下降、发热增加和电池耗电过快，它停止向部分配备英特尔处理器的戴尔电脑推送更新。但微软和戴尔尚未披露受影响设备的型号。微软只是表示正与戴尔合作，防止受影响型号出现问题，计划在未来几天内发布针对受影响设备的解决方案。

在 Sam Altman 住所遭遇纵火未遂事件五天后，一名男子尝试尾随 Anthropic 员工进入公司大楼，保安及时阻止了他，该男子对保安说要去警告一名 Anthropic 高管，有人要杀他。随着对 AI 的反对声浪日益高涨，科技公司高管开始加强个人安保。提供安全服务的 Liferaft 公司称，针对 AI 公司高管和数据中心的网络威胁数量增长了七倍，6 月的威胁数量略有下降。Equilar 对公开文件的分析显示，2025 年标普 500 指数成分股公司中有 38.1% 的公司披露了高管安保支出，而 2021 年这一比例为 26.8%。其中 Palantir Technologies 高管安保支出在一年内增长了 150%；甲骨文安保支出增长了 85.5%——主要是保护 Larry Ellison 的住所；Salesforce 安保支出增至 400 万美元，比 2024 年多约 100 万美元。硅谷安保公司 JPT Security 的客户关系副总裁 Dakota Dominguez 称由于面临强烈反对，科技公司越来越多的要求配备武装保镖，不同于明星和政客喜欢身材魁梧的保镖，科技公司高管通常要求的是低调的安保措施。AI 公司也日益不鼓励员工身穿带有公司标识的服装。

IBM 公布了不及预期的季度业绩初步报告，称客户将原计划采购 Z 系列大型机的费用用于囤积 AI 硬件，包括服务器、存储和内存，导致大型机的收入下滑。这一消息导致 IBM 的股价周二暴跌 25% 以上。IBM 包括大型机的基础设施业务收入下滑 7%，而大型机销售疲软连带导致配套的交易处理软件销售收入下滑。IBM CEO Arvind Krishna 将这一切归因于客户竞相采购 AI 硬件。

美国众议院以 308 票赞成 117 票反对通过永久性夏令时法案 Sunshine Protection Act。该法案旨在结束一年两次调整时钟。但该法案在参议院的前景不明，已有共和党参议员表示反对该法案。研究已经发现，一年两次调整时间并不利于公众健康和安全。部分睡眠专家在标准时间和夏令时之间倾向于选择标准时间，认为标准时间与人体的昼夜周期节律更为一致，也更有利于冬季早晨的安全。

特朗普政府禁止身处刚果民主共和国的美国公民回国，该国最近爆发了埃博拉疫情。目前身处刚果民主共和国或近期曾前往该国的美国公民已被列入“禁止登机”名单。他们须在第三国停留 21 天后才能返回美国。24 名原定于周二登机回国的美国公民已被新规阻止。目前尚不清楚该禁令是否也适用于政府工作人员。美国疾控中心（CDC）至少有 24 名员工在刚果民主共和国工作。专家批评此类限制措施不仅无效而且有害。这些措施会阻碍公开疫情和疾病风险信息，损害经济，造成污名化。埃博拉病毒不像呼吸道病毒那样容易传播，它通过接触患者或近期死亡者的体液传播。人们不会因为坐在咳嗽者旁边而感染埃博拉病毒。截至 7 月 14 日，刚果民主共和国报告了 1963 例病例和 719 例死亡。

发表在《Nature Communications》期刊上的新研究显示，第二次怀孕会以与第一次怀孕熟悉而不同的方式改变大脑。早期的研究表明，第一次怀孕会重塑大脑，在此基础上，研究人员发现，每次怀孕都会在母亲的大脑上留下自己独特的印记。研究小组对 110 名女性进行了长期跟踪调查。一些人期待着他们的第一个孩子，一些人怀了第二个孩子，还有一些人仍然没有孩子。通过反复进行脑部扫描，研究人员追踪了整个研究过程中大脑的变化。研究人员发现，第一次怀孕会使大脑默认模式网络的结构和活动发生最大的变化，这是一个涉及自我反思、社会思维等功能的系统。在第二次怀孕期间，同样的网络再次发生变化，尽管程度较轻。最明显的变化发生在负责引导注意力和对感官信息做出反应的大脑网络中。

纽约州成为美国第一个暂停建造大型数据中心的州。这一禁令适用于用电量 50 MW 或更高的数据中心。随着美国民众越来越担心污染风险、能源成本上涨和水资源短缺，美国各地要求停建数据中心的呼声日益高涨。佛蒙特州参议员 Bernie Sanders 和纽约州民主党众议员 Alexandria Ocasio-Cortez 已提出立法，寻求在全国范围内禁止数据中心建设。但特朗普政府不支持此类禁令。

Linus Torvalds 在内核邮件列表上强调，Linux 内核项目不是一个反 AI 的项目，也从来不是一个“社会正义战士”项目。AI 因为对社会产生越来越大的影响而日益受到争议，Linux 内核作为一个大型开源项目，参与者中也有不少人持反 AI 立场。Linus 说，如果有人想要在内核推动反 AI 议题，那么他们最好创建分支或离开。AI 是一种有用的工具，AI 相关问题不涉及它是否“有用”，它确实也可能会对维护者带来痛苦，但解决方法不是像鸵鸟那样将头埋在沙里，而是确保 AI 工具能真正帮助维护者。内核不会阻止开发者使用 AI 工具，而 AI 和人类的自然智能一样都不完美。开源项目的社会意义从来都是附带的，参与开源项目主要是为了获得更好的技术，不是出于宗教信仰。内核项目的决策主要基于技术优势，而非对新工具的恐惧。

微软在本周二释出了 7 月例行安全更新，修复了创纪录的 570 个漏洞。微软将补丁数量的激增归于 AI 辅助的漏洞发现。其中近 60 个 Bug 的风险等级归类为高危级。微软还修复了 3 个 0day，其中 2 个正被利用。2 个被利用的 0day 都是提权漏洞，包括 CVE-2026-56155 和 CVE-2026-56164。第三个 0day 则能绕过 Windows BitLocker 的安全功能，微软表示尚未发现该漏洞被利用。

从尤里·加加林开始，宇航员进入太空已有 65 年历史，如今他们首次在太空中完成了诊断质量的 X 光照射。在这之前空间站上唯一可用的医学成像技术是超声波。X 射线是现代医学最强大的诊断工具之一，因为它速度快、精度高，不到一分钟就能确诊疑似骨折。2025 年发射的 Fram2 私人太空任务携带了一台便携式 X 光系统，宇航员在太空飞行期间拍摄了身体的 X 光照片。在太空拍摄 X 光照片的难点是如何定位患者、侦测器和 X 射线源，以及保持静止足够长时间以获得清晰的图像。手和手臂是最容易成像的身体部位，因为很容易保持静止。胸部、腹部和盆腔的成像难度更大，但其图像仍然足以进行诊断。

26 名 Meta 员工提起诉讼，指控社交巨人利用 AI 驱动的评分和监控系统筛选裁员对象，并且特别针对了休病假、家庭事务假、产假或育儿假的员工。这起诉讼被认为是美国首例针对科技巨头在裁员中使用 AI 的案例。诉讼指控，Meta 不是通过熟悉员工的中层经理制定裁员名单，而是使用了一系列内部 AI 系统。其中一种评分标准是员工使用 AI 工具的频率。Meta 内部系统会据此将员工分类为——AI Native、AI First 和 AI Enabled。诉讼称，Meta 的系统没有考虑到残疾员工以及休假员工所产生的差异，导致残疾和休假员工更高比例的进入裁员名单。Meta 回应称指控缺乏依据，称它的裁员决定是由人类而不是 AI 做出的。