@alexalbert__
· 2026-07-25 01:09
@alexalbert__
Alex Albert · 0 followers
欢迎来到这个世界，Opus 5。
10今日信号
🔥 Opus 5 发布
1 条最高 9.4
🔥 Claude Opus 5 发布
1 条最高 8.0
@alexalbert__
· 2026-07-25 01:14
分享一些 Opus 5 发布会中我最喜欢的图表。我们投入了大量工作，旨在让模型在提升智能水平的同时，实现跨领域的高 Token 效率利用……
事实Opus 5 的核心改进之一在于降低跨领域的 Token 消耗。
观点模型厂商正在加速将竞争焦点转向“推理效率与成本控制”。对 CEO 而言，API 成本的降低将显著提升复杂多步骤 Agent 的商业可行性，应重新审视长上下文及高频调用的应用架构。
Some of my favorite graphs from Opus 5 launch. We put a ton of work into making this model token efficient across domains while still raising the inte...
🔥 Claude Cowork 整合发布
1 条最高 7.6
@alexalbert__
· 2026-09-17 02:48
团队在这次整合中表现出色。融合后的用户体验比单独使用 Chat 或 Cowork 要好得多。而且全新的幻灯片/文档/设计……
纯对话界面正在被混合工作台取代。
事实Anthropic 将 Chat 与 Cowork 模式整合并增强文档与设计生成能力。
观点
对你意味着不可只做孤立的问答层，深度融合任务上下文与多模态画布的工作流才是 AI 原生应用的核心壁垒。
The team did an amazing job with this merge. The resulting UX feels so much better than chat or Cowork did on their own. And the new slides/docs/desig...
🔥 Fable 5.1 EFS 发布
1 条最高 7.6
@alexalbert__
· 2026-09-02 04:45
在随 Fable 5.1 一同宣布的内容中，有一项非模型相关的更新让我非常高兴，那就是企业级前沿安全防护（EFS）。即使你不是 CEO 或 CISO，这篇博客...
事实Anthropic 随 Fable 5.1 推出了企业级前沿安全防护（EFS）。
观点言外之意，前沿 AI 的竞争正在从单纯的模型能力扩展至企业级安全与合规保障。
对你意味着构建企业级 AI 应用时，需将安全合规架构提至核心战略高度。
One non-model thing I'm very excited we announced today with Fable 5.1 is Enterprise Frontier Safeguards (EFS). Even if you're not a CEO/CISO, this bl...
🔥 Claude 速率限制重置
1 条最高 7.4
@alexalbert__
· 2026-07-10 02:25
更多 Fable！Claude 开发者通知：我们已重置所有用户的 5 小时和每周速率限制。
对你意味着——
事实速率限制重置通常发生在系统扩容后，或对高需求事件（如 Fable 上线流量激增）的响应。Fable 被 Replit CEO 和 Anthropic 同步推广，说明这是一次协同上线，而非各自独立的公告。
观点如果你的产品依赖 Claude API，此次限制重置是一个评估窗口——更宽松的配额意味着现在是密集测试 Claude 能力边界、收集基准数据的好时机。
More Fable!ClaudeDevs: We've reset 5-hour and weekly rate limits for all users.
🔥 Claude Tag 发布
1 条最高 6.6
@alexalbert__
· 2026-06-24 01:19
这彻底改变了我与Claude的工作方式。感觉不再像在使用一个工具，更像是在管理一个团队。
对你意味着"管理团队"vs"使用工具"这组对比是Claude Tag体验设计的核心意图——Anthropic正在将AI助手从单次请求响应转变为持续协作的团队成员关系。
观点如果这个体感成立，意味着你在Slack中委派给Claude的任务可以更复杂、更持续，适合多轮信息汇集和跨时间段的任务跟进。
观点
This has completely changed how I work with Claude. It feels less like using a tool and more like managing a team.
🔥 Claude Fable 5 恢复上线
1 条最高 4.4
@alexalbert__
· 2026-07-02 03:39
欢迎 Fable 回归世界！！
对你意味着Alex Albert 使用「Welcome back to the world」而非简单的产品更新语气，暗示 Fable 5 的下线不只是技术维护——语气中有明显的如释重负。
观点当 AI 公司因政策原因暂停模型能力、恢复时进行公开庆祝，往往是在向开发者社区重建信任；这种「重返仪式」本身是 AI 公司在政策合规期间的品牌管理动作，一定程度上反映了下线期间的内部压力强度。
Welcome back to the world Fable!!
📘 模型公司战略
2 条最高 8.4
@alexalbert__
· 2026-07-25 03:08
仅仅 6 个多月后，Opus 5 现在能够生成接近超人类水平的电子表格，以及达到专业咨询顾问制作水平的演示文稿。事情正在...
事实Anthropic 正将 Opus 5 的能力重点提升至复杂表格与分析汇报等高阶白领工作流。
观点纯套壳类办公 AI 辅助工具的壁垒将快速失效，AI 驱动型公司应直接接入顶级原生模型重新设计内部商业分析与决策汇报链条。
Just over 6 months later, Opus 5 now produces near-superhuman level spreadsheets and slide decks that match what a consultant would make. Things are c...
@alexalbert__
· 2026-09-23 00:39
使用 Opus 5.5 的感觉就像开车回家路上遇到的每个路口都是绿灯，口渴时喝到一杯冰水，以及在漫长一天后冲了一个热水澡……
事实Anthropic 内部已在测试 Opus 5.5。
观点核心人员公开造势预示发布临近。
对你意味着切勿在当前模型短板上堆砌补丁，为新底座保留即插即用的弹性。
Using Opus 5.5 feels like hitting every green light on your drive home, drinking a cold glass of water when you're thirsty, and a hot shower after a l...
📘 AI 产品化
1 条最高 6.0
@alexalbert__
· 2026-09-13 00:41
这个周末很值得花几分钟完整阅读这篇文章。在科技行业，“嵌入式评估员（Embedded evaluators）”听起来可能有些罕见，但在其他行业中却相当普遍……
事实传统行业普遍设有专职质检角色，而科技界此前较为罕见。
观点
对你意味着，随着 AI 落地深水区，纯自动化 Eval 难以应对复杂业务场景；应尽早在各业务线内嵌专业评估人员，将质控沉淀为组织核心壁垒。
Worth spending a few minutes this weekend to read this essay in full. Embedded evaluators might sound unusual in tech but it's pretty normal in other ...
尚未被主编选入事件/主题的原始推文（共 49 条，按时间倒序）
@alexalbert__09-23 00:31
RT Claude: Re One more thing: we’re increasing five-hour usage limits on Pro, Max, and Team plans. We’re also providing subscription users a rate li...
🔗 原推@alexalbert__09-02 02:35
RT ClaudeDevs: With Fable 5.1 out today, we've also reset 5-hour and weekly limits for all users.
🔗 原推@alexalbert__09-02 02:03
RT Claude: Re Fable 5.1 excels at complex, long-running tasks. And its research capabilities offer an early glimpse of how AI models will contribute t...
🔗 原推@alexalbert__08-15 03:16
RT Anthropic: We’ve written an FAQ to answer some of the questions we've received about watermarking. In summary: • We’re implementing watermarking...
🔗 原推@alexalbert__07-08 01:36
RT Claude: We're extending access to Claude Fable 5 on all paid plans through July 12.
🔗 原推@alexalbert__07-02 05:16
RT ClaudeDevs: Now that Fable 5 is ready to build (again), we've reset everyone's 5-hour and weekly rate limits.
🔗 原推@alexalbert__07-01 02:00
RT Claude: Introducing Claude Sonnet 5, our most agentic Sonnet yet. It makes plans, uses tools like browsers and terminals, and runs autonomously at ...
🔗 原推@alexalbert__06-24 06:26
RT Andrej Karpathy: This is a new paradigm for interacting with Claude that is significantly more "inline" with all the other human activity org-wide....
🔗 原推@alexalbert__06-13 08:50
RT Anthropic: The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Myt...
🔗 原推@alexalbert__06-13 03:09
RT Epoch AI: Claude Fable 5 scores very well on FrontierMath: Tiers 1–4 (v2), reaching 87% on Tiers 1–3 and 88% on Tier 4. This continues a streak o...
🔗 原推@alexalbert__06-11 13:56
RT ClaudeDevs: We’re rolling out changes to make Fable 5’s safeguards for frontier LLM development visible. Starting this week, flagged requests wil...
🔗 原推@alexalbert__06-11 02:48
RT Dario Amodei: Today I'm publishing a new essay, Policy on the AI Exponential. AI is progressing extremely fast—much faster than the policy process...
🔗 原推@alexalbert__06-10 06:00
We've reset usage limits across our products! For those just starting to test Fable, here's four tips for using it more effectively: 1. Give it bigger...
🔗 原推@alexalbert__06-10 01:09
I've been at Anthropic through every model launch. There's been a few cases I can remember of a launch that stands out and marks a step-change in how ...
🔗 原推@alexalbert__06-10 01:08
RT Claude: Re Releasing a model this capable comes with risks. Without safeguards, Fable 5’s capabilities in areas like cybersecurity could be misuse...
🔗 原推@alexalbert__06-06 06:01
RT Claude: We've doubled usage limits in Claude Cowork for the next month. Delegate bigger, more complex tasks to Claude.
🔗 原推@alexalbert__06-05 01:01
We just published internal data on how much of Claude's development is already being done by Claude: - Over 80% of all code merged into our codebase i...
🔗 原推@alexalbert__06-02 01:35
RT ClaudeDevs: We've reset 5-hour and weekly rate limits for all users on Pro and Max plans. We fixed an issue that caused some Claude Code sessions t...
🔗 原推@alexalbert__05-29 01:35
We put a lot of work into calibrating thinking effort for Opus 4.8. As you're trying out the model, if you do run into any examples of it still over/u...
🔗 原推@alexalbert__05-29 01:13
Fast mode for Opus 4.8 is much more affordable now. Try it out in Claude Code, I've found it changes how I use Claude. Fast mode for interactive work ...
🔗 原推@alexalbert__05-29 01:01
RT Dan Shipper 📧: BREAKING: Anthropic just dropped Opus 4.8—and it is a MONSTER We've been testing for about a week @every and our verdict is they...
🔗 原推@alexalbert__05-29 00:59
Excited to release Opus 4.8 today! We heard your feedback on 4.7 and have made many fixes for 4.8. 4.8 understands nuances better, feels much more nat...
🔗 原推@alexalbert__05-26 23:40
RT levent: over the weekend i checked the obvious thing, which is whether mythos is able to solve the erdos unit distance problem, aka erdos problem #...
🔗 原推@alexalbert__05-17 22:30
RT Peter Yang: Here's my new episode with @alexalbert__, who shared an inside look at how Anthropic is building the next Claude. We talked about how t...
🔗 原推@alexalbert__05-16 02:00
RT ClaudeDevs: Happy Friday! We've reset everyone's 5-hour and weekly rate limits.
🔗 原推@alexalbert__05-14 03:07
RT ClaudeDevs: Claude Code weekly limits are increasing 50%, now through July 13. Live now for all Pro, Max, Team, and seat-based Enterprise users.
🔗 原推@alexalbert__05-13 02:23
RT ClaudeDevs: Fast mode for Claude Opus 4.7 is now available in research preview on the API and in Claude Code.
🔗 原推@alexalbert__05-09 07:54
An early Claude Mythos Preview snapshot we provided METR has a time horizon of more than 2x the next best model on their 80% success rate benchmark
🔗 原推@alexalbert__05-09 03:02
RT Peter Wildeford🇺🇸🚀: PALO ALTO NETWORKS on MYTHOS: "In our testing, three weeks of model-assisted analysis matched a full year of manual pe...
🔗 原推@alexalbert__05-07 00:44
I'll be at Code with Claude all day today so come find me and let's chat about Claude! I'll also be giving a talk on the main stage at 530pm PT so tun...
🔗 原推@alexalbert__04-24 01:44
RT ClaudeDevs: Over the past month, some of you reported Claude Code's quality had slipped. We investigated, and published a post-mortem on the three ...
🔗 原推@alexalbert__04-21 01:19
Jack's young money blog had a big impact on me when I was in college and navigating what I wanted to do post-grad. If you are in your teens/20s and tr...
🔗 原推@alexalbert__04-17 23:40
Everyone with a vision can produce very high-quality designs now (with a lil help from Claude)
🔗 原推@alexalbert__04-17 23:14
A lot of bugs that folks may have hit yesterday when first trying Opus 4.7 are now fixed. Thanks for bearing with us🙏
🔗 原推@alexalbert__04-17 08:55
RT Jeremy Howard: Wow I can already say after just 5 hours using @AnthropicAI Opus 4.7 that this is the first model that "gets" what I'm doing when I'...
🔗 原推@alexalbert__04-16 22:53
RT Vals AI: The new Opus 4.7 model places #1 on our Vibe Code Benchmark, at 71%. When we first released the benchmark 4.5 months ago, no model scored ...
🔗 原推@alexalbert__04-16 22:44
Some of my favorite things in Opus 4.7: - Very good at async work and following instructions - Effort levels are far more predictable for token contro...
🔗 原推@alexalbert__04-16 22:09
RT ClaudeDevs: For the developers building with Claude, a direct line from the team. Follow for changelogs, API releases, community updates, and deep ...
🔗 原推@alexalbert__04-15 04:01
New version of Claude Code in the desktop app dropped today with tons of new features and performance improvements. Between Cowork and Code, I've foun...
🔗 原推@alexalbert__04-11 02:25
RT Claude: Claude for Word is now in beta. Draft, edit, and revise documents directly from the sidebar. Claude preserves your formatting, and edits ap...
🔗 原推@alexalbert__04-10 03:49
Allowing Sonnet to "phone a friend" (i.e. call Opus) increases performance while also reducing total cost since it reduces tokens spent trying to solv...
🔗 原推@alexalbert__04-09 02:10
I've found Managed Agents to somehow be both the fastest way to hack together a weekend agent project and the most robust way to ship one to millions ...
🔗 原推@alexalbert__04-08 02:07
Glasswing is possibly the most consequential event in the AI industry I've seen up close since joining Anthropic almost 3 years ago. It feels like we'...
🔗 原推@alexalbert__03-24 08:05
Highly recommend reading this guest post by theoretical physicist Matthew Schwartz to get a sense of how AI is helping science. He found Opus 4.5 to b...
🔗 原推@alexalbert__03-24 07:42
The future where I never have to open up my laptop to get work done is becoming real very fast
🔗 原推
生成于 2026-09-25 22:18（北京时间）· v0.2 CEO 战略情报版