Samsung 在 Samsung Health app 中推行了一项颇具争议的新规定，实际上迫使用户在同意将健康数据用于 AI 训练与建模与保留其现有数据之间做出选择。公司在应用设置深处嵌入了一个名为 "Consent to the Use of Health Data for AI training and modelling" 的开关，启用后 Samsung 可利用个人健康指标来优化机器学习算法并改进服务。

如果用户选择退出数据共享，应用会明确警告其将无法再将健康数据与 Samsung account 同步。公司还表示，除非法律另有要求，否则现有健康数据将被删除；若法律要求必须保留，数据将在法定保留期届满后立即被删除。

所涉数据范围相当广泛，主要包括四类：睡眠模式、服药记录、医疗记录和经期追踪详情。 Samsung 还披露，这些数据可能会由人工审阅者查看，审阅者可能包括公司员工和第三方承包商。

此政策调整与平台范围更广的生成式 AI 改造同步推出，正值 Galaxy Watch 9 等新硬件发布之际。由 AI 驱动的新功能包括 Vitals 工具，可监测心率、心率变异性、呼吸频率、皮肤温度和血氧等生理指标，提供健康洞察；应用中还将整合 Heart Health Score 、 Cardio Load 和 Fitness Index 等功能，旨在利用用户数据进行更深入的分析。 Samsung has introduced a controversial new requirement for users of the Samsung Health app, effectively forcing them to choose between consenting to AI data training or losing their stored information. The company has embedded a specific toggle labeled Consent to the Use of Health Data for AI training and modelling deep within the app settings. This setting enables Samsung to use personal metrics to refine its machine learning algorithms and improve overall service functionality.

When users attempt to opt out of this data sharing, the application issues a clear warning that they will no longer be able to sync their health data with their Samsung account. Furthermore, the company states that any existing health data will be deleted, unless it is required to be retained by law. If such a retention period is mandatory, the data will be erased as soon as that window closes.

The scope of data involved is extensive, covering four primary categories, including sleep patterns, medication logs, medical records, and cycle tracking details. Samsung has also disclosed that this data may be subject to review by human reviewers, which potentially includes both company employees and third-party contractors.

This policy shift arrives alongside a broader Generative AI overhaul of the platform, timed for the launch of new hardware like the Galaxy Watch 9. These AI-powered features include the Vitals tool, which tracks biometrics such as heart rate, heart rate variability, respiratory rate, skin temperature, and blood oxygen levels to provide health insights. Additional features integrated into the app include a Heart Health Score, Cardio Load metrics, and a Fitness Index, all designed to leverage user data for deeper analysis.

当 Trump administration 着手削减对 NOAA 网站 Climate.gov 的经费并将其停用时，超过 15 年的重要气候研究与数据面临被抹去的危险。为防止这些信息从公众视野中消失，前员工 Rebecca Lindsey 、她的姐妹 Mary Lindsey 和同事 Anna Eshelman 组成的团队出手相救，成功上线了后续网站 Climate.us，有效保存了大量关键地图、教学资料和科学报告的档案。

这些数据得以保存，主要因为 U.S. government 的信息在法律上被归为公有领域。由于这些资料以宽松许可方式发布，它们可以被合法地抢救并托管到其他地方，而不会完全消失。若无这些开放数据的法律保护，政府当时的举措可能会导致关键资源的永久丧失，其中就包括 Fifth National Climate Assessment——政府关于气候危机最全面的分析之一。

Climate.us 目前已成为一个重要资源，提供追踪 Arctic sea ice 等趋势的交互式仪表板；同时设有数据集图库，收藏了诸如 NOAA oral history archives 等独特的历史记录，记录了受环境变化直接影响者的亲身经历。除原始数据外，网站还为教师和学生提供教育工具，帮助他们更好地探讨气候与能源相关议题。

尽管这次抢救行动取得了成效，项目仍然处于脆弱状态。与最初由政府资助的项目不同，后续网站完全依赖自愿捐款维持运营。私人公民出手相助固然值得赞赏，但对捐款的依赖凸显了政府基础设施的系统性不足。此类档案是做出知情决策的重要工具，理应获得稳定可靠的支持，确保这些关键公共信息持续向所有人开放。 When the Trump administration moved to defund and deactivate the NOAA website Climate.gov, it threatened to erase over 15 years of vital climate research and data. In response, a team consisting of former employees Rebecca Lindsey, her sister Mary Lindsey, and colleague Anna Eshelman took action to prevent this information from being lost to the public. They successfully launched a successor site, Climate.us, effectively preserving a massive archive of essential maps, educational materials, and scientific reports.

The preservation of this data was made possible primarily because U.S. government information is classified as public domain by law. Because this information was available under a permissive license, it could be legally rescued and hosted elsewhere rather than disappearing entirely. Without these open data laws, the administration's actions would have resulted in the permanent destruction of critical resources, including the Fifth National Climate Assessment, which stands as one of the government's most comprehensive analyses of the climate crisis.

Climate.us currently functions as a robust resource, featuring interactive dashboards that track trends such as Arctic sea ice levels. It also provides a dedicated gallery for datasets, including unique historical records like the NOAA oral history archives, which document the experiences of individuals directly impacted by environmental changes. Beyond raw data, the site offers educational tools designed to help teachers and students engage with topics related to climate and energy.

Despite the success of this rescue effort, the project remains in a precarious position. Unlike the original government-funded initiative, this successor site relies entirely on voluntary donations to sustain its operations. While it is commendable that private citizens have stepped in to fill this gap, the reliance on charity highlights a systemic failure of government infrastructure. These archives serve as vital tools for informed decision-making, and they deserve reliable support to ensure that such essential public information remains accessible to everyone.

域名 t.me 最初于 May 20, 2010 注册，历史悠久。当前由注册商 GoDaddy.com, LLC 管理，注册有效期至 May 20, 2035 。根据最近一次于 July 13, 2026 的更新，注册信息显示该域由位于 Arizona, United States 的 Domains By Proxy, LLC 持有。

该域名受到严格限制，可从多项状态码看出：clientDeleteProhibited 、 serverDeleteProhibited 、 clientRenewProhibited 、 clientTransferProhibited 、 serverTransferProhibited 、 clientUpdateProhibited 和 serverUpdateProhibited 。此外，该域目前处于 serverHold 状态，DNSSEC 未签名（unsigned）。

技术上该域由 Google 的云端名称服务器管理，具体为 ns-cloud-b1.googledomains.com 至 ns-cloud-b4.googledomains.com 。虽然原始 Whois 数据列出了这些技术联系人和注册信息，但注册人、管理及技术联系人的大多数个人信息已被屏蔽以保护隐私。

查看此 Whois 记录的用户请注意：所提供的信息仅供参考并用于按需查询。注册局运营商和注册商明令禁止将这些数据用于大批量自动化处理或群发骚扰性通信。记录反映域名的当前状态与管理情况，但并不保证绝对准确，可能会根据注册局政策和隐私法规发生变更。 The domain t.me has a long history, having been originally registered on May 20, 2010. It is currently managed through the registrar GoDaddy.com, LLC, and is set to remain active until May 20, 2035. As of the most recent update on July 13, 2026, the registration details indicate that the domain is held under the organization Domains By Proxy, LLC, which is based in Arizona, United States.

The domain is heavily restricted, as evidenced by a comprehensive set of status codes. It is flagged with multiple prohibitions, including clientDeleteProhibited, serverDeleteProhibited, clientRenewProhibited, clientTransferProhibited, serverTransferProhibited, clientUpdateProhibited, and serverUpdateProhibited. Additionally, the domain is currently in a serverHold status, and the DNSSEC configuration is listed as unsigned.

Technical management of the domain is handled via Google's cloud-based name servers, specifically ns-cloud-b1 through b4.googledomains.com. While the raw Whois data provides these technical points of contact and registry information, most of the specific personal contact details for the registrant, administrative, and technical roles are redacted to protect privacy.

Users accessing this Whois record are reminded that the provided information is intended for informational purposes and query-based access. The registry operator and registrar maintain strict policies against using this data for high-volume automated processes or mass unsolicited communications. While the records offer a snapshot of the domain's current status and oversight, they do not guarantee absolute accuracy and are subject to change according to registry policies and privacy regulations.

California 的立法者正在审议一项可能迫使社交媒体公司为年轻用户重塑平台设计的法案。 Assembly Bill 1709 由 Assemblymember Josh Lowenthal 提出，最初拟禁止 16 岁以下青少年使用社交媒体。但在关于数据隐私、潜在言论自由问题以及可能孤立弱势青少年的担忧提出后，法案被修订为不全面封禁，而是针对特定以提高参与度为目的的功能进行监管。

修订后的提案要求像 Meta 和 Reddit 这样的平台为 16 岁以下用户提供一种替代的、成瘾性更低的信息流。若公司未能调整其软件，将被禁止允许该年龄组保留账户。法案将"成瘾性功能"定义为那些旨在最大化参与度并导致强迫性使用的设计，例子包括无限滚动、自动播放和某些算法推荐。

Assemblymember Josh Lowenthal 将这些功能视为产品设计选择而非受保护的言论，强调立法目标是遏制掠夺性设计，而非限制交流。根据修订方案，公司须在 2028 年前完成相关改造。此外，法案建议成立由专家组成的监督小组，向 California Attorney General's Office 就如何有效管理这些数字安全标准提供咨询。

尽管以 TechNet 为代表的科技行业此前反对全面禁令，但对于修订后的要求，业界已开始展开谨慎对话。法案的倡导者认为，United States 在为数字时代未成年人建立必要保护方面落后于 Australia 等国，立法是既能让青少年保持数字联系又能营造更安全网络环境的一步。

该法案目前已送交 Senate Appropriations Committee 进一步审查。支持者希望，通过将责任交由科技公司去设计更健康的数字体验，可以在不切断青少年所依赖在线社区的前提下，减轻那些以最大化参与度为目标的算法带来的负面影响。 California lawmakers are currently considering legislation that could force social media companies to overhaul the design of their platforms for younger users. Assembly Bill 1709, introduced by Assemblymember Josh Lowenthal, originally sought to prohibit teenagers under 16 from accessing social media. However, after concerns were raised regarding data privacy, potential free speech violations, and the risk of isolating vulnerable youth, the bill was amended to focus on regulating specific engagement-driven features rather than banning access entirely.

The updated proposal mandates that platforms like Meta and Reddit must provide an alternative, less addictive feed for users under 16. If companies fail to adapt their software, they will be prohibited from allowing individuals in that age group to maintain accounts. The bill specifically defines addictive features as those designed to maximize engagement, which result in compulsive use. Key examples mentioned include the infinite scroll, autoplay functions, and certain algorithmic recommendations.

Assemblymember Lowenthal has framed these features as product choices rather than protected speech, emphasizing that the goal is to stop predatory design rather than to restrict communication. Under the revised plan, companies would have until 2028 to implement these changes. Furthermore, the legislation proposes creating an oversight group of experts tasked with advising the California Attorney General's Office on how to manage these digital safety standards effectively.

While the tech industry, represented by groups like TechNet, previously pushed back against a total prohibition, there is now cautious dialogue regarding the amended requirements. Advocates of the bill argue that the United States is currently lagging behind other nations, such as Australia, in establishing necessary protections for minors in the digital age. They contend that the legislation is a vital step toward creating a safer online environment while still allowing teens to maintain their digital connections.

The bill is now moving forward to the Senate Appropriations Committee for further review. Proponents hope the shift in focus, which puts the responsibility on technology firms to design healthier digital experiences, will ultimately mitigate the negative impacts of engagement-maximizing algorithms without cutting teenagers off from the online communities they rely on.

开发和分发 Mac 和 iOS 应用并不需要持续使用 Xcode 的图形界面。虽然必须在系统上安装 Xcode，因为它包含 iOS SDK 、 notarytool 、 devicectl 等必要的命令行工具，但所有构建与部署工作都可以在终端 shell 中完成。唯一需要 GUI 的场景是一次性设置：配置 Apple Developer 账户、生成签名证书并将公证凭据存入 keychain 。

要摆脱传统的 Xcode 工作流，可以用 XcodeGen 管理项目结构。 Xcode 项目文件本质上是复杂的文件夹，变动频繁且容易与版本控制发生冲突。 XcodeGen 允许你在简洁的 YAML 文件中定义配置，并按需重建项目目录。这样，git 仓库只需追踪轻量的配置文件，整个构建过程也能保持可脚本化与可复现。

自动化发布流水线可以用自定义 shell 脚本实现，按序执行归档、签名、公证和安装等标准命令。结合 notarytool 进行恶意软件扫描、用 stapler 附加公证票据，可以复刻 Xcode Organizer 中的步骤。脚本还可以加入签名需求检查与最终 bundle 验证等环节，确保部署过程稳健并能在后台自动完成。

为了让 AI 编码助手无缝对接这套流程，建议编写项目文档（例如 CLAUDE.md），明确说明构建与发布命令。一旦 agent 理解了这些约定及发布脚本的结构，就能独立完成整个构建与测试周期。这既允许你快速进行未签名的本地构建以做初步检查，也能在准备分发时执行完整的签名与公证流程。

总体而言，这种无头方法将复杂的应用开发生命周期转化为一系列可预测、可重复的命令行操作。将这些任务委托给 LLM，可以避免 Xcode 那种不透明界面带来的手动开销。初始基础设施建立完毕后，生成、测试与交付专业级应用往往只需执行一条命令，从而显著简化开发体验。 Developing and distributing Mac and iOS applications does not require constant interaction with the Xcode graphical interface. While Xcode must be installed on your system because it contains essential command-line tools like the iOS SDK, notarytool, and devicectl, all build and deployment tasks can be handled through a standard terminal shell. The only time the GUI is necessary is during a one-time setup process to configure your Apple Developer account, generate signing certificates, and store notarization credentials within your keychain.

To move away from the traditional Xcode workflow, you can utilize XcodeGen to manage your project structure. Because Xcode project files are complex folders that frequently change and cause friction with version control, XcodeGen allows you to define your settings in a clean YAML file, which then recreates the project folder on demand. This approach ensures that only the lightweight configuration file needs to be tracked in your git repository, while the entire build process remains scriptable and reproducible.

Automating your release pipeline is achieved through a custom shell script, which executes a series of standard commands to archive, sign, notarize, and install your application. By incorporating tools like notarytool to handle malware scanning and stapler to attach notarization tickets, you can mirror the steps normally performed by the Xcode Organizer. This script can be further enhanced by implementing checks for signing requirements and verifying the final bundle, ensuring a robust, automated deployment process that functions entirely in the background.

To make this workflow seamless for AI coding assistants, you can create a project documentation file, such as CLAUDE.md, that outlines your build and release commands. Once an agent understands these conventions and the structure of your release script, it can handle the entire build and testing cycle independently. This setup allows you to perform rapid, unsigned local builds for quick sanity checks while reserving the full, signed, and notarized release process for when you are ready to distribute your application.

Ultimately, this headless methodology transforms the complex app development lifecycle into a series of predictable, repeatable command-line operations. By delegating these tasks to an LLM, you avoid the manual overhead typically associated with Xcode's inscrutable interface. Once the initial infrastructure is established, the process of generating, testing, and shipping professional-grade applications becomes a matter of executing a single command, significantly streamlining the development experience.

Sega 32X 是这一系列复古计算项目的最新目标，旨在把 Linux kernel 带到九十年代中期的硬件上。和之前将 Linux 移植到 Atari Jaguar 的工作类似，这次的目标是通过应对遗留硅片、有限资源和非标准架构的各种复杂性，来打磨板级启动（bringup）能力。对于 32X 来说，这意味着要面对一种特殊的双处理器结构：两颗 23 MHz 的 Hitachi SH2，作为 16-bit Genesis 与 32-bit Saturn 之间的过渡设计。

要在 32X 上成功引导 Linux，必须突破严重的内存限制。附加卡上只有 256KB RAM，因此项目依赖现代基于 FPGA 的 flash carts，特别是带有 Extended SSFv2 Mapper 的型号。通过把 cartridge ROM 当作可写的 RAM 使用，扩展了总可用内存，从而能为 Linux kernel 和一个小型 initramfs 留出足够空间。在调整 SDK 示例以支持 ROM-to-RAM 映射之后，工作重心转向用自定义的 sh2eb-elf toolchain 交叉编译内核，并修复各种内部编译器错误与寄存器被破坏的问题。

在实现 Symmetric Multiprocessing (SMP) 支持时遇到了重大挑战。由于 32X 硬件本身缺乏原生同步原语和缓存一致性，要实现真正的并行处理只能靠巧妙的软件方案。开发者采用 Peterson's algorithm 来做 CPU 间的同步，并利用原始的 Genesis 68000 处理器作为仲裁器负责中断路由，这样系统才能引导两个 SH2 核心。为保证稳定性，他们通过禁用共享变量的缓存并为每个处理器分配唯一标识，规避了缓存相关的问题。

整个过程充满反复试验：为解决调度错误和 kernel panics 进行了大量调试，并对代码进行了激进优化，才把 Linux 和 Busybox 装进有限的存储空间。实现了一个基本的控制台显示驱动并利用 ELF FDPIC 支持后，该移植达到了能运行最小用户空间的可用状态。尽管最终系统存在严重的总线争用且性能有限，但能够在 32X 硬件上成功运行 Linux kernel，本身就是一项技术成就，证明只要有足够的毅力和创造性的工程设计，再受限的复古架构也能支持现代软件。 The Sega 32X serves as the latest target in a series of ambitious retro-computing projects aimed at bringing the Linux kernel to hardware from the mid-nineties. Much like the previous effort to port Linux to the Atari Jaguar, the goal here is to refine board bringup skills by navigating the complexities of legacy silicon, limited resources, and non-standard architectures. For the 32X, this meant working with a unique dual-processor setup featuring two 23 MHz Hitachi SH2s, which were intended to act as an intermediary step between the 16-bit Genesis and the 32-bit Saturn.

Successfully booting Linux on the 32X required overcoming significant memory constraints. With only 256KB of RAM available on the addon, the project relied on modern FPGA-based flash carts, specifically those featuring an Extended SSFv2 Mapper. By treating the cartridge ROM as writeable RAM, the total usable memory was expanded, allowing for a configuration that leaves enough room for both the kernel and a modest initramfs. After adjusting the SDK examples to handle ROM-to-RAM mappings, the focus shifted to cross-compiling the kernel using a custom `sh2eb-elf` toolchain and addressing various internal compiler errors and register clobbering issues.

A major challenge arose during the development of Symmetric Multiprocessing (SMP) support. Because the 32X hardware lacks native synchronization primitives and cache coherency, implementing true parallel processing required a clever software-based approach. By leveraging Peterson's algorithm to handle inter-CPU synchronization and utilizing the original Genesis 68000 processor as an arbiter to route interrupts, the system was able to boot both SH2 cores. To ensure stability, the developers bypassed cache-related issues by disabling caching for shared variables and assigning unique identifiers to each processor.

Ultimately, the process involved extensive trial and error to resolve scheduling bugs and kernel panics, as well as aggressive optimization to fit Linux and Busybox into the limited footprint. By implementing a basic console display driver and utilizing ELF FDPIC support, the port reached a functional state capable of running a minimal user space. While the resulting system experiences significant bus contention and modest performance, the ability to successfully execute the Linux kernel across the 32X hardware stands as a technical achievement that proves even the most restrictive retro architectures can support modern software if provided with enough perseverance and creative engineering.

Super Dario 是 Pascal Schuster 开发的一款简单的浏览器平台游戏。玩法直观：用 A 、 D 键或方向键移动，按空格跳跃，按 M 键开关声音。

作为一项极简主义小作，游戏强调可访问性和即开即玩的网页体验，是作者的一次创意练习，注重易上手和即时互动。

该项目由 Pascal Schuster 亲自托管，他通过个人博客和社交媒体分享作品并与玩家交流，体现了独立网页游戏的休闲与个人特色。 Super Dario is a simple, browser-based platform game developed by Pascal Schuster. The game features straightforward mechanics where players navigate a character using either the AD keys or the arrow keys, jump using the spacebar, and manage sound settings with the M key.

Designed as a minimalist project, the game highlights the creator's focus on accessible, quick-to-play web experiences. It serves as a creative exercise in game development, prioritizing ease of access and immediate interactivity for the user.

The project is hosted by its creator, Pascal Schuster, who maintains a presence through his personal blog and various social media platforms. Through these channels, he shares his work and connects with players, emphasizing the casual and personal nature of independent web gaming.

Apple 在 iOS 和 macOS 26 中引入了新的语音识别 API SpeechAnalyzer，实际上取代了旧有的 SFSpeechRecognizer 。由于 Apple 没有公布此次更新的官方准确性基准，开发者和用户对于其相较于 OpenAI 的 Whisper 等已有方案的性能存在疑问。为此，研究者用来自 LibriSpeech 的 5,559 条测试语句，比较了新的 Apple 引擎、旧版 API 以及三种完全在设备端运行的 Whisper 变体，开展了一次全面的基准测试。

结果显示，SpeechAnalyzer 目前是 Apple 平台上最准确的本地语音引擎。它在所有测试的 Whisper 模型（包括 Whisper Small）之上，并且效率显著更高，运行速度约为其三倍。相比之下，旧版 SFSpeechRecognizer 表现最差，其准确度甚至低于最小的 Whisper 模型。对于那些不只依赖简单语音命令的开发者，数据表明迁移到 SpeechAnalyzer 是非常值得的，因为它在转录准确性上有显著提升，且输出文本更干净、带有标点。

在与 Whisper 的对比中，Apple 的新引擎在现代 Apple 硬件上成为英语转录的更优选择。虽然 Whisper 在语言覆盖面上仍有优势，支持 100 多种语言，而 Apple 的 SpeechTranscriber 约支持 30 种，但在英语处理方面两者的差距已大幅缩小。因此，在英语任务中，Whisper 不再是追求最高本地准确度时的默认选项。基准测试方法严谨，采用了标准的 LibriSpeech 语料，并确保 Whisper 的测试结果与 OpenAI 公布的指标一致，从而验证了针对 Apple 引擎所用方法的有效性。

该基准的透明性体现在原始转录的公开以及使用相同的生产代码路径，确保结果反映真实使用场景而非孤立的实验室条件。研究还指出，所有被测引擎的运行速度均远快于实时，但新的 Apple API 在效率上具有明显优势。这项分析已推动实际的软件部署调整，例如促使 Inscribe 应用在支持的语言中优先采用更准确的 SpeechAnalyzer 。

尽管收益显著，结果也有局限：LibriSpeech 主要以英语朗读语音为主，后续需要测试带口音的语音、远场录音或多说话人的会议音频等场景，以评估各引擎的表现。此外，虽然预计这些准确性结论在 Apple Silicon 系列硬件上普遍适用，但运行速度会随具体芯片架构有所波动。总之，对于当前的 iPhone 和 Mac 用户来说，系统内建工具现在能在不牺牲准确性的前提下，提供一流且注重隐私的本地语音转录解决方案。 Apple has introduced a new speech recognition API, SpeechAnalyzer, as part of iOS and macOS 26, effectively replacing the older SFSpeechRecognizer. Because Apple did not provide official accuracy benchmarks for this update, there has been significant uncertainty for developers and users regarding its performance compared to existing solutions like OpenAI's Whisper models. To address this, a comprehensive benchmark was conducted using 5,559 test utterances from the LibriSpeech dataset, comparing the new Apple engine, the legacy API, and three Whisper variants, all running fully on-device.

The results clearly indicate that the new SpeechAnalyzer is the most accurate on-device speech engine currently available on Apple platforms. It outperformed every Whisper model tested, including Whisper Small, while also demonstrating significantly higher efficiency, running approximately three times faster. In contrast, the legacy SFSpeechRecognizer performed the poorest, proving to be substantially less accurate than even the smallest Whisper model. For developers relying on the old API for anything beyond simple voice commands, the data suggests that migrating to SpeechAnalyzer is highly recommended due to the significant gains in transcription accuracy and the output of cleaner, punctuated text.

When evaluated against Whisper, Apple's new engine emerged as the superior choice for English transcription on modern Apple hardware. While Whisper maintains distinct advantages regarding language support, with compatibility for over 100 languages compared to the roughly 30 supported by Apple's SpeechTranscriber, the gap for English processing has closed. Consequently, for English-language tasks, Whisper is no longer the automatic default for those seeking the highest accuracy on-device. The benchmarking process was robust, utilizing standard LibriSpeech corpora and ensuring that the Whisper results aligned with OpenAI's own published metrics, which serves to validate the methodology used for the Apple engine tests.

The transparency of this benchmark is reinforced by the publication of raw transcripts and the use of identical production code paths, ensuring that the findings reflect real-world usage rather than isolated laboratory conditions. The study highlights that all tested engines operated comfortably faster than real time, though the efficiency of the new Apple API provides a distinct performance edge. This analysis has already led to practical changes in software deployment, specifically prompting updates to the Inscribe application to prioritize the more accurate SpeechAnalyzer for supported languages.

Despite the clear benefits, there are noted limitations to these findings, as the LibriSpeech dataset focuses primarily on English read speech. Future testing will be necessary to determine how these engines perform with accented, far-field, or multi-speaker meeting audio. Furthermore, while these accuracy results are expected to hold across Apple Silicon hardware, performance speeds will naturally fluctuate based on the specific chip architecture. Ultimately, for users on current iPhones or Macs, the built-in system tools now provide a best-in-class, privacy-conscious solution for local speech transcription, removing the need to compromise on accuracy for the sake of on-device processing.

Department of Government Efficiency（简称 DOGE）已正式结束运作，该部门曾主导对联邦政府的一次大规模重组。但尽管其影响深远，目前并没有关于具体措施、节省成本或内部决策的公开账目。 Trump administration 一直在积极隐瞒这些记录，这也令人担忧像 Elon Musk 这样的私人个体如何能够影响政府职能这一问题的透明度。

政府为保护这些文件所采取的做法，是把 DOGE 归类为咨询机构而非独立机构。官员们据此主张 DOGE 不受 Freedom of Information Act 的约束，而其记录应适用 Presidential Records Act 。这一分类有效把公众获取信息的时间至少推迟到政府离任后五年，造成事实上的信息真空，使公众无法核实其实际所为。

迹象显示，DOGE 的数字痕迹正在被清除。有报道称包括 National Labor Relations Board 在内的联邦机构已删除了 DOGE 团队的账户，这可能危及针对举报人投诉的正在进行的调查。此类销毁行为可能违反要求保存政府档案的 Federal Records Act 。通过破坏像 FOIA requests 这样的传统透明机制，政府几乎切断了公众了解这场行政整顿实际程度的渠道。

局势因政府最近关于总统记录性质的法律主张而更加复杂。 Justice Department 的一份备忘录提出，Presidential Records Act 本身可能违宪，暗示总统文件属于私人财产而非公共资产。这一激进的政策转向为政府任意销毁或扣留信息提供了法律依据。尽管目前已有诉讼试图反对这一立场并保全记录，政府已表明将上诉，令政府透明度的前景更加不确定。

围绕 DOGE 的透明度缺失，反映出一场更广泛、协同行动的努力——削弱政府问责制。从撤换独立监察长到强制签署限制性保密协议，这些举措都使公众无法看清政府内部运作。批评者担忧，DOGE 的做法把权力外包给私人利益的同时又摧毁审计这些决策的能力，会为未来将私人影响置于民主监督之上的治理模式开创危险先例。 The Department of Government Efficiency, or DOGE, has officially concluded its operations after leading a sweeping restructuring of the federal government. Despite the initiative's significant impact, there is currently no public accounting of its specific actions, cost savings, or internal decisions. The Trump administration has actively worked to keep these records hidden, raising concerns about a lack of transparency regarding how a private figure like Elon Musk was able to influence government functions.

The administration's strategy for shielding these documents is rooted in the classification of DOGE as an advisory body rather than an independent agency. By labeling it this way, officials argue that DOGE is not subject to the Freedom of Information Act, which would typically allow for public oversight. Instead, they contend its records fall under the Presidential Records Act. This classification effectively delays public access for at least five years after the administration leaves office, creating a functional blackout that prevents the public from verifying what was actually done.

Evidence suggests that the digital footprint of DOGE is already being erased. Reports indicate that federal agencies, such as the National Labor Relations Board, have deleted DOGE team accounts, potentially compromising ongoing investigations into whistleblower complaints. This destruction of records may violate the Federal Records Act, which requires the preservation of government files. By sabotaging traditional transparency mechanisms like FOIA requests, the administration has left the public with little recourse to understand the extent of the administrative overhaul.

The situation is further complicated by the administration's recent legal assertions regarding the nature of presidential records. A Justice Department memo has floated the idea that the Presidential Records Act itself is unconstitutional, suggesting that presidential papers are private property rather than public assets. This radical shift in policy provides a legal framework for the administration to destroy or withhold information at will. While legal battles are currently underway to contest this stance and preserve records, the administration has signaled its intent to appeal, leaving the future of government transparency in a state of limbo.

Ultimately, the lack of transparency surrounding DOGE highlights a broader, coordinated effort to dismantle government accountability. From the removal of independent inspectors general to the imposition of restrictive nondisclosure agreements, these actions blind the public to the inner workings of their government. The concern among critics is that the DOGE model, which outsources power to private interests while destroying the ability to audit those decisions, sets a dangerous precedent for future governance that prioritizes private influence over democratic oversight.

Los Angeles Police Department（LAPD）准备结束与 Flock Safety 为期三年的合同。 Flock Safety 以其庞大的车牌识别摄像头网络著称，LAPD 的首席信息官 Dean Gialamas 表示，出于对公民自由和隐私的严重担忧，决定暂停使用该技术，直到就数据存储、安全和信息共享等问题达成更明确的合同条款。

作为美国最大的警察机构之一，LAPD 退出这一平台具有重要意义。其他城市，如 Mountain View, California 和 South Portland, Maine，也已相继与该公司断绝合作。这些地区常以隐私风险和数据可能被未经授权使用为由提出反对，曾有报道称监控数据被联邦移民官员利用，违背了当地的 sanctuary city 政策。

Flock Safety 对合同到期表示惊讶，称部门的担忧源于误解，公司希望予以澄清。但公众和监督组织对其的审查愈发严格。批评者指出，多起因 AI 误报导致警方拘留甚至持枪对准无辜司机的事件；有记者也记录到，在其车辆被系统错误标记为被盗后遭警方跟踪多日的情况。

除了操作失误外，该公司还因屡次出现的安全漏洞受到指责。调查显示，其摄像头有时在互联网上暴露，研究人员还发现执法端登录凭据缺乏强制多因素认证等安全缺陷。这些问题引发了要求联邦调查的呼声，尤其是在有报道称 Drug Enforcement Administration 等联邦机构被指未经授权使用当地警察密码监控与移民相关的嫌疑人之后。

抵制情绪已在社区层面显现，一些居民直接拆除摄像头或用垃圾袋遮挡。随着公众对这些监控能力认识的提高，执法部门对自动化追踪工具的需求与公众对隐私的期待之间的紧张关系不断加剧。

随着更多管辖区重新审视对该类技术的依赖，Flock Safety 是否能有效应对日益严峻的法律与社会挑战，仍有待观察。 The Los Angeles Police Department is set to end its three-year contract with Flock Safety, a prominent surveillance company known for its vast network of license plate reader cameras. The department's chief information officer, Dean Gialamas, cited serious concerns regarding civil liberties and privacy as the primary drivers behind the decision. The LAPD intends to pause its use of the technology until it can establish clearer contractual terms regarding data storage, security, and information sharing.

As one of the largest police departments in the United States, the LAPD's departure from the platform marks a significant shift. Other municipalities, such as Mountain View, California, and South Portland, Maine, have similarly cut ties with the company. These regions have frequently raised objections centered on privacy and the potential for unauthorized data usage, including reports that surveillance information was being leveraged by federal immigration officials in contradiction of local sanctuary city policies.

Flock Safety has expressed surprise at the contract expiration, maintaining that the department's concerns are rooted in misconceptions the company hopes to clarify. Despite this pushback, the firm has faced mounting scrutiny from the public and watchdog groups. Critics have pointed to numerous instances of police detaining or even pointing weapons at innocent motorists due to AI-driven false positives. Journalists have also documented being tracked for days by police after their vehicles were incorrectly flagged as stolen by the system.

Beyond operational errors, the company is under fire for recurring security lapses. Investigations have revealed that the platform's cameras have at times been publicly exposed to the internet, and researchers have identified vulnerabilities where law enforcement login credentials lacked mandatory multi-factor authentication. These issues have prompted calls for federal investigations, particularly after reports emerged that federal agencies, such as the Drug Enforcement Administration, allegedly utilized local police passwords without authorization to monitor immigration-related suspects.

The growing backlash has manifested in community-level resistance, with some residents taking direct action to dismantle cameras or obscure them with trash bags. As public awareness of these surveillance capabilities grows, the tension between law enforcement's desire for automated tracking tools and the public's expectation of privacy continues to intensify. Whether Flock Safety can successfully navigate these mounting legal and social challenges remains to be seen as more jurisdictions reconsider their reliance on the technology.