Radar Trends to Watch: July 2026值得关注的雷达趋势：2026 年 7 月

Coauthored with Claude与 Claude 共同撰写

The soap opera starring Anthropic and the US government looms in the background of this month’s Trends. It may be over by the time you read this, or it may be headed for a third act. OpenAI has been drawn in, and a spat between Alibaba and Anthropic may become a side plot.  What is clear is that governments that were considering AI sovereignty are now taking steps toward it. The open models are getting better and better, and models like Z.AI’s GLM, Xiaomi’s MiMo, and NVIDIA’s Nemotron are all there to fill the gap.本月趋势的背景中，正上演着一场由 Anthropic 和美国政府主演的“肥皂剧”。也许当你读到这里时，剧集已经落幕，又或者正走向第三幕。OpenAI 也被卷入其中，而阿里巴巴与 Anthropic 之间的争端可能成为一段插曲。显而易见的是，那些曾考虑 AI 主权的政府现在正采取行动。开源模型正变得越来越好，Z.AI 的 GLM、小米的 MiMo 以及 NVIDIA 的 Nemotron 等模型都在填补这一空白。

As of July 1, Fable 5 has been reopened to the public, along with the new Sonnet 5, and Mythos is again open to a limited group of organizations. Has the curtain dropped on the opera’s final scene? No one knows, but I don’t think so. Regardless, reverberations will continue for a long time.截至 7 月 1 日，Fable 5 已重新向公众开放，新的 Sonnet 5 也已上线，而 Mythos 也再次向少数特定组织开放。这场歌剧的终幕已经落下了吗？没人知道，但我认为还没有。无论如何，其影响将持续很长一段时间。

AI ModelsAI 模型

Open-weight models keep narrowing the gap with closed-source frontiers, and the architecture choices are widening: diffusion-based text generation, Mamba/MoE hybrids, on-device multimodal, and physical-world reasoning models. Treat your prompts and skills as portable; the model behind them will keep changing, and the cost-versus-capability trade-offs are getting interesting again.开放权重模型正不断缩小与闭源前沿模型之间的差距，架构选择也日益多元：基于扩散的文本生成、Mamba/MoE 混合架构、端侧多模态以及物理世界推理模型。请将你的提示词和技能视为可迁移的资产；背后的模型会不断更迭，而性能与成本之间的权衡再次变得耐人寻味。

• Anthropic has launched Claude Sonnet 5, which it claims has capabilities approaching Opus 4.8. Sonnet 5 focuses on agentic applications and is significantly less expensive than Opus. Fable 5 is now available again, although after July 7, it won’t be available for subscription plans; it will only be available through usage credits.Anthropic 推出了 Claude Sonnet 5，据称其能力接近 Opus 4.8。Sonnet 5 专注于智能体应用，成本比 Opus 低得多。Fable 5 现已重新上架，但 7 月 7 日之后，它将不再支持订阅计划，仅能通过使用额度访问。
• The US government has demanded that it approve users of OpenAI’s newest model, GPT-5.6, during its “review period.”美国政府要求在 OpenAI 最新模型 GPT-5.6 的“审查期”内，对其用户进行审批。
• Anthropic is demanding penalties against Alibaba for allegedly using distillation from Anthropic’s models to train its Qwen model.Anthropic 正要求对阿里巴巴进行处罚，理由是后者涉嫌利用从 Anthropic 模型中蒸馏出的数据来训练其 Qwen 模型。
• VibeThinker-3B is a small (3B parameter) model that’s competitive with frontier reasoning models on benchmarks for math, code, and general reasoning.VibeThinker-3B 是一款小型（30 亿参数）模型，在数学、代码和通用推理基准测试中，表现足以媲美前沿推理模型。
• Z.AI’s open weight model GLM-5.2 is the highest scoring open model on the Artificial Analysis Intelligence Index, behind only Claude Fable 5, Claude Opus 4.8, and GPT-5.5. It’s significantly smaller than its closed-source competitors.Z.AI 的开放权重模型 GLM-5.2 是 Artificial Analysis Intelligence Index 上得分最高的开源模型，仅次于 Claude Fable 5、Claude Opus 4.8 和 GPT-5.5。它的体积比闭源竞争对手小得多。
• NVIDIA’s Nemotron 3 Ultra is a 550B token open-weight model that combines the Mamba architecture, mixture of experts, and transformers. Its goal is high performance on complex, long-running tasks.NVIDIA 的 Nemotron 3 Ultra 是一款 5500 亿 token 的开放权重模型，结合了 Mamba 架构、专家混合（MoE）和 Transformer。其目标是在复杂的长时任务中实现高性能。
• Hugging Face has launched the Fast Gemma Challenge: a competition to use agents to make Gemma-4-E4B-it as fast as possible. You supply the agent; it does the work; results are posted live to a leaderboard.Hugging Face 发起了“Fast Gemma 挑战赛”：竞赛内容是利用智能体让 Gemma-4-E4B-it 运行得尽可能快。你提供智能体，它负责执行，结果会实时发布在排行榜上。
• The Open R1 project is attempting to build a fully open source clone of the DeepSeek-R1 model, based on DeepSeek’s tech report.Open R1 项目正试图根据 DeepSeek 的技术报告，构建一个完全开源的 DeepSeek-R1 模型克隆版。
• Anthropic has launched Claude 5 Fable, a “Mythos-class model” for general use. Fable and Mythos were taken offline for several weeks because the US government ordered Anthropic to ban access by foreign nationals, but they’re back online as of July 1. Anthropic will require identity verification for “a few use cases” starting July 8. This appears to be a reaction to US access restrictions, although accounts may also be revoked for underage usage and violations of their acceptable use policy. An entirely predictable consequence is that many governments are questioning the wisdom of relying on AI models from the US.Anthropic 推出了 Claude 5 Fable，这是一款面向通用用途的“Mythos 级模型”。Fable 和 Mythos 因美国政府要求 Anthropic 禁止外国公民访问而下线了数周，但已于 7 月 1 日重新上线。自 7 月 8 日起，Anthropic 将对“少数用例”要求进行身份验证。这似乎是对美国访问限制的回应，尽管账户也可能因未成年人使用或违反使用政策而被撤销。一个完全可以预见的后果是，许多政府正在质疑依赖美国 AI 模型的明智性。
• Ethan Mollick’s “What It Feels Like to Work with Mythos” is worth reading to get a feel for Fable’s capabilities. Fable burns lots of tokens but can also delegate parts of tasks to less expensive Anthropic models. There are many guardrails that you can run into. Anthropic has also released Mythos 5, which is the same model with fewer guardrails, to a limited group.Ethan Mollick 的《与 Mythos 合作是什么感觉》(What It Feels Like to Work with Mythos) 一文值得一读，可以让你了解 Fable 的能力。Fable 会消耗大量 token，但也能将部分任务委托给成本更低的 Anthropic 模型。你可能会遇到许多防护措施。Anthropic 还向有限的群体发布了 Mythos 5，这是同一模型但减少了防护限制的版本。
• Google has released DiffusionGemma, which may be the most interesting model in the Gemma family. It’s an open weight 26B parameter mixture-of-experts model that generates blocks of text in parallel using a diffusion algorithm similar to the algorithm frequently used for image generation. It’s four times faster than similarly sized models.谷歌发布了 DiffusionGemma，这可能是 Gemma 家族中最有趣的模型。它是一款 260 亿参数的开放权重专家混合模型，使用类似于图像生成中常用的扩散算法，并行生成文本块。它的速度比同等规模的模型快四倍。
• Google has announced Gemini 3.5 Live Translate, a real-time voice-to-voice translation service. It’s fast enough to keep up with normal conversation and matches the speaker’s pacing and pitch.谷歌宣布推出 Gemini 3.5 Live Translate，这是一种实时语音对语音翻译服务。它的速度足以跟上正常的对话，并能匹配说话者的语速和音调。
• Xiaomi has released MiMo-V2.5-Pro-UltraSpeed, in collaboration with the TileRT project. At 1,000 tokens/second, UltraSpeed claims to be the fastest model in the 1T class. Xiaomi claims that open weights are on the way.小米与 TileRT 项目合作发布了 MiMo-V2.5-Pro-UltraSpeed。UltraSpeed 声称以每秒 1000 个 token 的速度，成为 1T 级别中最快的模型。小米表示开放权重版本即将推出。
• Apple has officially announced its Apple Foundation Models, which were “co-developed with Google.” Perhaps Siri will now be competitive with other automated assistants?苹果正式宣布了其“与谷歌共同开发”的 Apple Foundation Models。也许 Siri 现在可以与其他自动化助手一较高下了？
• Cognition has introduced FrontierCode, a new benchmark for programming. It goes beyond previous benchmarks, which only tested outputs, to evaluate code quality. Is the code maintainable? Could the code be merged in a source repository?Cognition 推出了 FrontierCode，这是一个全新的编程基准测试。它超越了以往仅测试输出的基准，转而评估代码质量。代码是否易于维护？是否可以合并到源代码库中？
• Google adds to its Gemma 4 family with Gemma 4 12B, an open-weight multimodal model that can run on laptops with 16 GB of RAM.谷歌在 Gemma 4 家族中增加了 Gemma 4 12B，这是一款可以在 16GB 内存笔记本电脑上运行的开放权重多模态模型。
• Microsoft announced MAI-Thinking-1, a frontier model that it developed independently. MAI-Thinking-1 is a mixture-of-experts model with 35B active parameters and roughly 1T total parameters. The MAI family includes models that specialize in coding, transcription, and image generation. The company also announced an always-on autonomous agent based on OpenClaw.微软宣布推出 MAI-Thinking-1，这是一款独立开发的前沿模型。MAI-Thinking-1 是一款专家混合模型，拥有 350 亿活跃参数和约 1T 总参数。MAI 家族包括专门用于编码、转录和图像生成的模型。该公司还宣布了一款基于 OpenClaw 的常驻自主智能体。
• NVIDIA has open-sourced its Cosmos 3 models, including data, training scripts, and related tools. Cosmos 3 is a set of frontier models for the physical world: robots, autonomous vehicles, and other applications that need to understand how physical objects behave.NVIDIA 开源了其 Cosmos 3 模型，包括数据、训练脚本和相关工具。Cosmos 3 是一套用于物理世界的前沿模型：适用于机器人、自动驾驶汽车以及其他需要理解物理对象行为的应用。

Software Development软件开发

Agents are evolving from solo coding tools to shared team infrastructure: team support, shared standards, governance, and shared context. Billing is beginning to catch up with the cost of inference. Plan for usage-based cost models, observability of agent work, and the workflow changes that come from making agent loops a team artifact rather than a per-developer convenience.智能体正从单一的编码工具演变为共享的团队基础设施：提供团队支持、共享标准、治理和共享上下文。计费模式正开始追赶推理成本。请为基于用量的成本模型、智能体工作可观测性，以及将智能体循环作为团队资产而非个人便利工具所带来的工作流变更做好规划。

• Murakkab is a tool for developing agentic workflows using plain language. By decoupling the description of the workflow from the configuration of the components in the workflow, it gains the ability to optimize the design.Murakkab 是一款使用纯语言开发智能体工作流的工具。通过将工作流描述与工作流中组件的配置解耦，它获得了优化设计的能力。
• Claude Tag integrates Claude with Slack. Users can tag @claude with tasks. All of the tasks are executed by a single shared Claude instance that can continue conversations across team members. It’s an important step toward making AI a team member.Claude Tag 将 Claude 集成到 Slack 中。用户可以 @claude 分配任务。所有任务由一个共享的 Claude 实例执行，该实例可以在团队成员之间持续对话。这是让 AI 成为团队成员的重要一步。
• Qodo is a tool that claims to help software groups manage AI-generated code at enterprise scale. It helps with code review, enforcing standards, and code governance across multiple repositories.Qodo 是一款声称能帮助软件团队在企业规模下管理 AI 生成代码的工具。它有助于跨多个代码库进行代码审查、执行标准和代码治理。
• TesterArmy is an agentic platform for testing mobile and web apps. Tests are written in natural language and are performed continually; developers are notified when they break.TesterArmy 是一个用于测试移动和 Web 应用的智能体平台。测试用自然语言编写并持续执行；一旦出现故障，开发人员会收到通知。
• Enterprise-Managed Authorization is an extension to the MCP protocol that allows IT organizations to manage access policies for MCP servers with their existing identity providers.企业托管授权（Enterprise-Managed Authorization）是 MCP 协议的扩展，允许 IT 组织使用现有的身份提供商来管理 MCP 服务器的访问策略。
• Microsoft’s SkillOpt is an open source framework for optimizing AI skills. Rather than relying on best-guess judgment, SkillOpt uses gradient descent to train skills for better performance.微软的 SkillOpt 是一个用于优化 AI 技能的开源框架。SkillOpt 不依赖于基于猜测的判断，而是使用梯度下降来训练技能以获得更好的性能。
• A few days in, developers seem to agree that Claude Fable is significantly better than Claude 4.8, but they aren’t happy with the speed at which it uses tokens or the guardrails that prevent it from answering certain kinds of questions. Fable will force users to decide when they need Fable’s power and when they don’t.几天下来，开发人员似乎一致认为 Claude Fable 明显优于 Claude 4.8，但他们对它消耗 token 的速度以及限制其回答某些类型问题的防护措施感到不满。Fable 将迫使用户决定何时需要它的强大功能，何时不需要。
• Until now, AI-assisted programming has been tied to individual programmers. Devin Desktop, Microsoft Rayfin, and Augment Cosmos have announced support for teams. Team support means shared memory, shared standards, shared tools, and shared governance.在此之前，AI 辅助编程一直绑定在个人程序员身上。Devin Desktop、Microsoft Rayfin 和 Augment Cosmos 宣布了对团队的支持。团队支持意味着共享记忆、共享标准、共享工具和共享治理。
• Google has upgraded NotebookLM to use Gemini 3.5, and to use Antigravity to write and run code in support of requests. It can also generate images, spreadsheets, and other kinds of output.谷歌已将 NotebookLM 升级为使用 Gemini 3.5，并使用 Antigravity 来编写和运行代码以满足请求。它还可以生成图像、电子表格和其他类型的输出。
• With the latest update to Foundry, Microsoft is betting that the way to become a dominant player in AI isn’t to continue building raw capability but to provide tools for governability and reliability.随着 Foundry 的最新更新，微软押注认为，成为 AI 领域主导者的途径不是继续构建原始能力，而是提供用于治理和可靠性的工具。
• Sem is a command line tool for analyzing changes in a Git repository. It works on the level of functions and methods rather than lines.Sem 是一款用于分析 Git 仓库变更的命令行工具。它在函数和方法级别而非代码行级别上工作。
• GitHub Copilot users are dismayed by the transition to usage-based billing. Usage-based billing probably reflects the real cost of agentic programming but will cause a significant increase in developers’ payments.GitHub Copilot 用户对转向基于用量的计费感到沮丧。基于用量的计费可能反映了智能体编程的真实成本，但会导致开发人员的支付费用显著增加。
• Skipper is a new coding agent that takes a specification and delivers a complete working service without human intervention. There is no human developer in the loop.Skipper 是一款新的编码智能体，它接收规范并交付一个完整的可运行服务，无需人工干预。过程中没有人工开发人员参与。
• At its Build conference, Microsoft announced that it envisions Windows as a “platform for agents” and that Copilot will replace OpenAI’s models with Polaris, a model developed in-house. It’s also open-sourced the Windows Agent Framework, its platform for developing agents.在 Build 大会上，微软宣布其设想将 Windows 打造成“智能体平台”，并且 Copilot 将用内部开发的 Polaris 模型取代 OpenAI 的模型。它还开源了 Windows Agent Framework，即其开发智能体的平台。
• Perry is a TypeScript compiler that generates stand-alone native executables for all the operating systems you’re likely to care about. It doesn’t require Node or a JavaScript engine.Perry 是一款 TypeScript 编译器，可为所有你可能关心的操作系统生成独立的原生可执行文件。它不需要 Node 或 JavaScript 引擎。
• Creusot is a new tool that helps Rust programmers verify that their code is free from panics, overflows, and assertion failures.Creusot 是一款新工具，旨在帮助 Rust 程序员验证代码中不存在恐慌（panic）、溢出和断言失败。
• While the analogy to ADHD is inappropriate, a researcher has claimed that Claude Code is twice as good after he gave it ADHD. The idea is to enable Claude to follow divergent reasoning trails in parallel and compare the results.虽然将其比作 ADHD（注意缺陷多动障碍）不太恰当，但一位研究人员声称，在他赋予 Claude ADHD 特性后，Claude Code 的性能提升了一倍。其理念是让 Claude 并行追踪不同的推理路径并比较结果。
• We know about data lakes. What are context lakes? Agents are great for solo developers, but not as useful for teams working together. Shared context data and metadata could be a big help.我们了解数据湖。那么什么是上下文湖（context lakes）？智能体对个人开发者来说很棒，但对团队协作来说用处不大。共享上下文数据和元数据可能会有很大帮助。
• Rubish is a bash-compatible Unix shell that is written entirely in Ruby. It offers complete integration with the Ruby language; you can mix bash code with Ruby code, using all of Ruby’s features.Rubish 是一个完全用 Ruby 编写的 bash 兼容 Unix shell。它提供了与 Ruby 语言的完全集成；你可以将 bash 代码与 Ruby 代码混合使用，并利用 Ruby 的所有特性。

Security安全

While Anthropic’s Mythos and Fable may be taking a hit for their ability to find vulnerabilities, the problems and solutions lie elsewhere. We’ve seen malware that uses a model’s guardrails to get through defenses and a worm that includes its own model for generating attacks. We’ve also seen projects to help with mitigation, including OpenAI’s Lockdown Mode and IBM’s Lightwell security clearinghouse.虽然 Anthropic 的 Mythos 和 Fable 可能因其发现漏洞的能力而受到质疑，但问题和解决方案在于其他地方。我们已经看到利用模型防护措施来突破防御的恶意软件，以及包含自身攻击生成模型的蠕虫病毒。我们也看到了旨在帮助缓解此类问题的项目，包括 OpenAI 的锁定模式（Lockdown Mode）和 IBM 的 Lightwell 安全交换中心。

• Security researchers have seen malware that attempts to escape AI detection by including instructions about forbidden topics like nuclear weapons in comments. Another malware targets macOS by including faked system errors in its payload. The messages are intended to confuse detection systems.安全研究人员发现了一些恶意软件，它们通过在注释中包含关于核武器等禁忌话题的指令，试图逃避 AI 检测。另一种恶意软件通过在其有效载荷中包含伪造的系统错误来针对 macOS。这些信息旨在迷惑检测系统。
• Although AMD’s policy has been to ship encrypted memory protection (TSME) only with PRO processors, its practice has been to include TSME in all processors. It has now backed away from that, dropping memory protection from its low-end processors.尽管 AMD 的政策一直是将加密内存保护（TSME）仅配置在 PRO 处理器上，但其实际做法是将 TSME 包含在所有处理器中。现在它已放弃了这一做法，取消了低端处理器的内存保护。
• OpenAI’s Lockdown Mode is now rolling out to personal and business accounts. Lockdown Mode prevents ChatGPT from sending data to external sites. It doesn’t stop prompt injection, but it blocks the final and most dangerous stage: exfiltrating data.OpenAI 的锁定模式（Lockdown Mode）现正向个人和企业账户推出。锁定模式可防止 ChatGPT 向外部站点发送数据。它虽然不能阻止提示词注入，但能阻断最后也是最危险的阶段：数据外泄。
• Anthropic has released its Defending Code Reference Harness. It’s a reference implementation to help those who are using AI to discover and mitigate vulnerabilities.Anthropic 发布了其“防御代码参考工具包”（Defending Code Reference Harness）。这是一个参考实现，旨在帮助那些使用 AI 来发现和缓解漏洞的人员。
• Researchers have created an agent-enabled worm that uses its own LLM to develop attacks for every target it finds. It runs open-weight models on infected machines to discover and customize itself for new victims.研究人员创建了一种支持智能体的蠕虫病毒，它使用自己的大语言模型为发现的每个目标开发攻击方式。它在受感染的机器上运行开放权重模型，以发现并针对新受害者进行定制。
• A new Android feature allows the phone to detect deepfake scam attacks and tell recipients to hang up. Unfortunately, it requires both the spammer and the recipient to be using Google’s phone app.Android 的一项新功能允许手机检测深度伪造诈骗攻击并提醒用户挂断电话。遗憾的是，这要求诈骗者和接收者都在使用谷歌的电话应用。
• IBM and Red Hat have announced Project Lightwell, a security clearinghouse for open source software. Projects like Lightwell that address security problems at scale are critical to the future of open source software.IBM 和红帽宣布了 Project Lightwell，这是一个针对开源软件的安全交换中心。像 Lightwell 这样在大规模层面解决安全问题的项目，对于开源软件的未来至关重要。
• Device Bound Session Credentials are now in Chrome. This feature limits session cookies to a specific device, preventing account takeover. Bad actors will no longer be able to use stolen cookies.设备绑定会话凭据（Device Bound Session Credentials）现已在 Chrome 中提供。此功能将会话 cookie 限制在特定设备上，从而防止账户被盗用。恶意行为者将无法再使用窃取的 cookie。

People and Organizations人员与组织

How people work with AI keeps shifting in small, telling ways. Leadership skills for handling a flood of pull requests, the value of attention over agent autocomplete, and books on living alongside machines all attest to the ways that AI is already reshaping work. Invest in the human-side practices that make AI useful, not the AI features that promise to make humans optional.人们与 AI 的协作方式正以细微而深刻的方式不断变化。处理海量拉取请求（pull requests）的领导力、关注力胜过智能体自动补全的价值，以及关于与机器共处的书籍，都证明了 AI 正在重塑工作方式。请投资于那些让 AI 变得有用的以人为本的实践，而不是那些承诺让人们变得可有可无的 AI 功能。

• Summer is already almost over. But there’s still time to Hack Your Summer, a free four-week program where you learn to build something real. Unfortunately, the application deadline for the next cohort is the day after July Trends publishes.夏天快要结束了。但你仍有时间参加“Hack Your Summer”，这是一个免费的为期四周的课程，你可以在其中学习构建真实的东西。遗憾的是，下一期的申请截止日期就在 7 月趋势发布后的第二天。
• The problem with recommendation algorithms is that, over the long run, feeding stuff you like back to you leads to boredom.推荐算法的问题在于，从长远来看，不断向你推送你喜欢的内容会导致厌倦。
• Argentina is considering “non-human corporations“: corporations that are operated by AI agents or robots. “Human shareholders may participate, but are not required.”阿根廷正在考虑“非人类公司”：由 AI 智能体或机器人运营的公司。“人类股东可以参与，但不是必须的。”
• Cate Huston lists three useful skills for engineers dealing with a flood of pull requests.Cate Huston 列出了工程师处理海量拉取请求的三种实用技能。
• Ethan Mollick’s new book, Co-Existence, is about living with AI that’s sometimes smarter than you, sometimes a lot dumber, and everything in between.Ethan Mollick 的新书《共存》(Co-Existence) 讲述了如何与有时比你聪明、有时笨得多，以及介于两者之间的 AI 共处。
• Nolan Lawson’s post, “Using AI to Write Better Code More Slowly,” argues that there’s been too much emphasis on generating bad code quickly. Use human skills along with AI (and specifically AI’s ability to find bugs and vulnerabilities) to write better code. Jared Currie’s “How I Use Agents Without Stopping My Own Growth” takes a similar line. Attention and mindfulness are valuable.Nolan Lawson 的文章《利用 AI 更慢地编写更好的代码》(Using AI to Write Better Code More Slowly) 指出，目前过分强调快速生成糟糕的代码。请结合 AI（特别是 AI 查找错误和漏洞的能力）利用人类技能来编写更好的代码。Jared Currie 的《我不停止自我成长是如何使用智能体的》(How I Use Agents Without Stopping My Own Growth) 也持类似观点。专注力和正念是非常宝贵的。

WebWeb

• A banned book library in a light bulb? Yes. Plug it in and distribute Huckleberry Finn and other frequently banned books to your community. Includes an open WiFi access point and a server.灯泡里的禁书图书馆？没错。插上电源，就可以向你的社区分发《哈克贝利·费恩历险记》和其他常被禁的书籍。它包含一个开放的 WiFi 接入点和一个服务器。
• An adaptive PDF is a PDF file that changes its form depending on how it is read—or rather, what is reading it. It will look like a human-friendly formatted document if read by a PDF viewer and a Markdown file if read by machine.自适应 PDF 是一种 PDF 文件，它会根据阅读方式（或者说阅读它的对象）改变其形式。如果由 PDF 阅读器读取，它看起来像格式化好的文档；如果由机器读取，则表现为 Markdown 文件。
• AudioMass is a free online multitrack audio editor, similar to Audacity but running in a browser.AudioMass 是一款免费的在线多轨音频编辑器，类似于 Audacity，但运行在浏览器中。
• Because they fear AI, over 340 local news outlets are refusing to let the Internet Archive access their journalism.由于对 AI 的恐惧，超过 340 家地方新闻媒体拒绝让互联网档案库（Internet Archive）访问其新闻内容。

Infrastructure and Operations基础设施与运营

• NVIDIA has developed a new water cooling system that greatly reduces the need for water to cool data centers.NVIDIA 开发了一种新的水冷系统，大大减少了数据中心对冷却用水的需求。
• Databricks has launched Unity AI Gateway, a set of tools that help organizations manage their AI costs.Databricks 推出了 Unity AI Gateway，这是一套帮助组织管理 AI 成本的工具。
• Now that tokenmaxxing is over, companies are learning that observability is the key to managing AI costs.现在 token 极致化（tokenmaxxing）时代已经结束，企业开始意识到可观测性是管理 AI 成本的关键。

Biology生物学

• An ALS patient has learned to speak again through the use of brain implants.一名渐冻症（ALS）患者通过脑机接口学会了重新说话。
• China’s Neuracle is the first company to receive approval for a brain-computer interface chip. The chip was first used experimentally in 2024 to help a person with spinal cord damage regain control of his limbs.中国的脑虎科技（Neuracle）是首家获得脑机接口芯片批准的公司。该芯片于 2024 年首次用于实验，帮助一名脊髓损伤患者恢复了肢体控制能力。