\u65e5\u671f\uff1a2026-05-03<\/p>\n
\u672c\u671f\u805a\u7126\uff1a\u91cd\u70b9\u5173\u6ce8\u6a21\u578b\u53d1\u5e03\u4e0e release notes\u3001\u5b98\u65b9 engineering blog\u3001AI coding \/ agent \/ SRE\u3001\u8bc4\u6d4b\u699c\u5355\u53d8\u5316\u3001\u5f00\u53d1\u8005\u5b9e\u8df5\u535a\u5ba2\u3001\u6846\u67b6\u751f\u6001\u3001\u5f00\u6e90\u6a21\u578b\u4e0e\u771f\u5b9e\u7528\u6237\u89c6\u89d2\uff1b\u5f53 HN\u3001Reddit\u3001Hugging Face \u7b49\u793e\u533a\u6e90\u53ef\u8bbf\u95ee\u65f6\u4f18\u5148\u7eb3\u5165\u3002<\/p>\n
Artificial Analysis \u6700\u65b0\u6a21\u578b\u6392\u540d\u89c2\u5bdf<\/strong>\uff08Artificial Analysis\uff09<\/p>\n \u4e2d\u6587\u6458\u8981\uff1a<\/strong>Artificial Analysis \u53d1\u5e03\u4e86\u6700\u65b0\u7684 AI \u6a21\u578b\u7efc\u5408\u6392\u540d\uff0c\u4ece\u667a\u80fd\u3001\u901f\u5ea6\u3001\u4ef7\u683c\u548c\u4e0a\u4e0b\u6587\u7a97\u53e3\u56db\u4e2a\u7ef4\u5ea6\u5bf9\u4e3b\u6d41\u6a21\u578b\u8fdb\u884c\u4e86\u8bc4\u4f30\u3002\u5728\u667a\u80fd\u6307\u6570\u65b9\u9762\uff0cGPT-5.5 (xhigh) \u4ee5 60 \u5206\u4f4d\u5c45\u699c\u9996\uff0cClaude Opus 4.7 (max) \u548c Gemini 3.1 Pro Preview \u5e76\u5217\u7b2c\u4e09\uff0857 \u5206\uff09\u3002\u901f\u5ea6\u65b9\u9762\uff0cMercury 2 \u4ee5 778 tokens\/\u79d2 \u9886\u5148\uff0cIBM Granite 4.0 H Small \u4ee5 400 tokens\/\u79d2 \u7d27\u968f\u5176\u540e\u3002\u4ef7\u683c\u7ef4\u5ea6\u4e0a\uff0cQwen3.5 0.8B \u4ee5\u6bcf\u767e\u4e07 tokens \u4ec5 0.02 \u7f8e\u5143\u6210\u4e3a\u6700\u4fbf\u5b9c\u6a21\u578b\u3002\u4e0a\u4e0b\u6587\u7a97\u53e3\u65b9\u9762\uff0cLlama 4 Scout \u652f\u6301 1000 \u4e07 tokens\uff0cGrok 4.20 \u652f\u6301 200 \u4e07 tokens\u3002\u8be5\u6392\u540d\u91c7\u7528 Intelligence Index v4.0\uff0c\u7efc\u5408 GDPval-AA\u3001Terminal-Bench Hard\u3001Humanity's Last Exam \u7b49 10 \u9879\u8bc4\u6d4b\u3002<\/p>\n English Summary:<\/strong> Artificial Analysis released its latest AI model rankings across intelligence, speed, price, and context window metrics. GPT-5.5 (xhigh) leads the Intelligence Index with a score of 60, followed by Claude Opus 4.7 (max) and Gemini 3.1 Pro Preview tied at 57. Mercury 2 tops speed at 778 tokens\/s, while Qwen3.5 0.8B is the most affordable at $0.02 per million tokens. Llama 4 Scout offers the largest context window at 10 million tokens. Rankings use Intelligence Index v4.<\/p>\n \u539f\u6587\u94fe\u63a5<\/a><\/p>\n<\/li>\n Introducing Claude Opus 4.7<\/strong>\uff08Anthropic News\uff09<\/p>\n \u4e2d\u6587\u6458\u8981\uff1a<\/strong>Anthropic \u6b63\u5f0f\u53d1\u5e03 Claude Opus 4.7\uff0c\u8fd9\u662f Opus 4.6 \u7684\u663e\u8457\u5347\u7ea7\u7248\u672c\uff0c\u5728\u9ad8\u7ea7\u8f6f\u4ef6\u5de5\u7a0b\u4efb\u52a1\u4e0a\u8868\u73b0\u5c24\u4e3a\u7a81\u51fa\u3002\u65b0\u6a21\u578b\u5728\u590d\u6742\u957f\u5468\u671f\u4efb\u52a1\u4e2d\u5c55\u73b0\u51fa\u66f4\u9ad8\u7684\u4e25\u8c28\u6027\u548c\u4e00\u81f4\u6027\uff0c\u80fd\u591f\u81ea\u4e3b\u9a8c\u8bc1\u8f93\u51fa\u7ed3\u679c\u3002\u89c6\u89c9\u80fd\u529b\u5927\u5e45\u63d0\u5347\uff0c\u652f\u6301\u6700\u9ad8 2576 \u50cf\u7d20\uff08\u7ea6 375 \u4e07\u50cf\u7d20\uff09\u7684\u9ad8\u5206\u8fa8\u7387\u56fe\u50cf\u5904\u7406\u3002\u5728\u591a\u9879\u57fa\u51c6\u6d4b\u8bd5\u4e2d\uff0cOpus 4.7 \u8d85\u8d8a\u4e86\u524d\u4ee3\uff1aSWE-bench Verified \u5f97\u5206 76.4%\uff08\u5bf9\u6bd4 Opus 4.6 \u7684 68.6%\uff09\uff0cTerminal-Bench 2.0 \u5f97\u5206 82.0%\uff08\u5bf9\u6bd4 69.6%\uff09\u3002\u65b0\u589e xhigh \u52aa\u529b\u7ea7\u522b\uff0c\u63d0\u4f9b\u66f4\u7cbe\u7ec6\u7684\u63a8\u7406\u63a7\u5236\u3002\u5b9a\u4ef7\u7ef4\u6301\u4e0d\u53d8\uff1a\u8f93\u5165 5 \u7f8e\u5143\/\u767e\u4e07 tokens\uff0c\u8f93\u51fa 25 \u7f8e\u5143\/\u767e\u4e07 tokens\u3002\u540c\u65f6\u5f15\u5165\u4e86\u7f51\u7edc\u5b89\u5168\u9632\u62a4\u63aa\u65bd\uff0c\u81ea\u52a8\u68c0\u6d4b\u548c\u963b\u6b62\u9ad8\u98ce\u9669\u7f51\u7edc\u653b\u51fb\u8bf7\u6c42\u3002<\/p>\n English Summary:<\/strong> Anthropic officially released Claude Opus 4.7, a significant upgrade from Opus 4.6 with notable improvements in advanced software engineering. The model demonstrates greater rigor and consistency on complex long-running tasks, with enhanced vision capabilities supporting images up to 2,576 pixels (~3.75 megapixels). Benchmarks show major gains: SWE-bench Verified at 76.4% (vs 68.6% for Opus 4.6), Terminal-Bench 2.0 at 82.0% (vs 69.6%). A new xhigh effort level provides finer control over reasoning. Pricing remains $5\/M input and $25\/M output tokens. The release includes cyber safeguards to automatically detect and block high-risk cybersecurity requests.<\/p>\n \u539f\u6587\u94fe\u63a5<\/a><\/p>\n<\/li>\n Featured An update on recent Claude Code quality reports<\/strong>\uff08Anthropic Engineering\uff09<\/p>\n \u4e2d\u6587\u6458\u8981\uff1a<\/strong>Anthropic \u5de5\u7a0b\u56e2\u961f\u53d1\u5e03 Claude Code \u8d28\u91cf\u95ee\u9898\u7684\u590d\u76d8\u62a5\u544a\uff0c\u8ffd\u6eaf\u5e76\u89e3\u51b3\u4e86\u8fc7\u53bb\u4e00\u4e2a\u6708\u7528\u6237\u53cd\u9988\u7684\u4e09\u4e2a\u72ec\u7acb\u95ee\u9898\u3002\u7b2c\u4e00\uff0c3 \u6708 4 \u65e5\u5c06\u9ed8\u8ba4\u63a8\u7406\u52aa\u529b\u7ea7\u522b\u4ece high \u6539\u4e3a medium \u4ee5\u964d\u4f4e\u5ef6\u8fdf\uff0c\u4f46\u5f71\u54cd\u4e86\u667a\u80fd\u8868\u73b0\uff0c\u5df2\u4e8e 4 \u6708 7 \u65e5\u56de\u6eda\uff0cOpus 4.7 \u9ed8\u8ba4\u8bbe\u4e3a xhigh\u3002\u7b2c\u4e8c\uff0c3 \u6708 26 \u65e5\u5b9e\u65bd\u7684\u7f13\u5b58\u4f18\u5316\u5b58\u5728 bug\uff0c\u5bfc\u81f4\u8d85\u8fc7\u4e00\u5c0f\u65f6\u7a7a\u95f2\u7684\u4f1a\u8bdd\u5728\u6062\u590d\u540e\u6bcf\u8f6e\u90fd\u4f1a\u6e05\u9664\u5386\u53f2\u601d\u8003\u5185\u5bb9\uff0c\u4f7f Claude \u663e\u5f97\u5065\u5fd8\u548c\u91cd\u590d\uff0c\u5df2\u4e8e 4 \u6708 10 \u65e5\u4fee\u590d\u3002\u7b2c\u4e09\uff0c4 \u6708 16 \u65e5\u6dfb\u52a0\u7684\u51cf\u5c11\u5197\u957f\u8f93\u51fa\u7684\u7cfb\u7edf\u63d0\u793a\u4e0e prompt \u53d8\u66f4\u7ed3\u5408\u540e\u5f71\u54cd\u4e86\u7f16\u7801\u8d28\u91cf\uff0c\u5df2\u4e8e 4 \u6708 20 \u65e5\u56de\u6eda\u3002Anthropic \u5df2\u91cd\u7f6e\u6240\u6709\u8ba2\u9605\u8005\u7684\u4f7f\u7528\u9650\u989d\uff0c\u5e76\u627f\u8bfa\u6539\u8fdb\u5185\u90e8\u6d4b\u8bd5\u6d41\u7a0b\u548c\u7cfb\u7edf\u63d0\u793a\u53d8\u66f4\u63a7\u5236\u3002<\/p>\n English Summary:<\/strong> Anthropic's engineering team published a postmortem on recent Claude Code quality issues, tracing user complaints to three separate changes. First, a March 4 change lowering default reasoning effort from high to medium was reverted April 7 after users reported reduced intelligence. Second, a March 26 caching optimization bug caused thinking history to be cleared on every turn after idle sessions, making Claude appear forgetful; fixed April 10. Third, an April 16 system prompt change to reduce verbosity harmed coding quality and was reverted April 20. Anthropic reset usage limits for all subscribers and committed to improving internal testing and system prompt change controls.<\/p>\n \u539f\u6587\u94fe\u63a5<\/a><\/p>\n<\/li>\n Scaling Managed Agents: Decoupling the brain from the hands<\/strong>\uff08Anthropic Engineering\uff09<\/p>\n \u4e2d\u6587\u6458\u8981\uff1a<\/strong>Anthropic \u5de5\u7a0b\u535a\u5ba2\u53d1\u5e03\u300aScaling Managed Agents: Decoupling the brain from the hands\u300b\uff0c\u9610\u8ff0\u4e86\u6258\u7ba1\u4ee3\u7406\uff08Managed Agents\uff09\u7684\u67b6\u6784\u8bbe\u8ba1\u7406\u5ff5\u3002\u6838\u5fc3\u601d\u60f3\u662f\u5c06\u4ee3\u7406\u7684"\u5927\u8111"\uff08Claude \u53ca\u5176 harness\uff09\u4e0e"\u53cc\u624b"\uff08\u6c99\u76d2\u548c\u6267\u884c\u5de5\u5177\uff09\u4ee5\u53ca"\u4f1a\u8bdd"\uff08\u4e8b\u4ef6\u65e5\u5fd7\uff09\u89e3\u8026\uff0c\u901a\u8fc7\u865a\u62df\u5316\u62bd\u8c61\u5b9e\u73b0\u7ec4\u4ef6\u72ec\u7acb\u6f14\u8fdb\u3002\u8fd9\u79cd\u67b6\u6784\u89e3\u51b3\u4e86\u65e9\u671f\u5355\u5bb9\u5668\u8bbe\u8ba1\u7684"\u5ba0\u7269\u670d\u52a1\u5668"\u95ee\u9898\u2014\u2014\u5f53\u5bb9\u5668\u6545\u969c\u65f6\u4f1a\u8bdd\u4e22\u5931\u3002\u89e3\u8026\u540e\uff0charness \u901a\u8fc7 execute(name, input) \u63a5\u53e3\u8c03\u7528\u6c99\u76d2\uff0c\u5bb9\u5668\u53d8\u4e3a\u53ef\u66ff\u6362\u7684" cattle"\uff1b\u4f1a\u8bdd\u65e5\u5fd7\u72ec\u7acb\u5b58\u50a8\uff0charness \u5d29\u6e83\u540e\u53ef\u901a\u8fc7 wake(sessionId) \u6062\u590d\u3002\u8be5\u8bbe\u8ba1\u4f7f p50 \u9996 token \u65f6\u95f4\u964d\u4f4e\u7ea6 60%\uff0cp95 \u964d\u4f4e\u8d85 90%\uff0c\u5e76\u652f\u6301\u591a brain \u548c\u591a hand \u7684\u7075\u6d3b\u7ec4\u5408\uff0c\u4e3a\u957f\u671f\u8fd0\u884c\u7684\u81ea\u4e3b\u4ee3\u7406\u63d0\u4f9b\u53ef\u9760\u3001\u5b89\u5168\u7684\u57fa\u7840\u8bbe\u65bd\u3002<\/p>\n English Summary:<\/strong> Anthropic's engineering blog published "Scaling Managed Agents: Decoupling the brain from the hands," explaining the architecture design for Managed Agents. The core concept decouples the "brain" (Claude and its harness) from "hands" (sandboxes and tools) and "session" (event logs) through virtualization abstractions. This solves the early single-container "pet server" problem where session loss occurred on container failure. After decoupling, harnesses call sandboxes via execute(name, input), making containers replaceable "cattle"; session logs are stored independently allowing harness recovery via wake(sessionId). This architecture reduced p50 time-to-first-token by ~60% and p95 by over 90%, supporting flexible multi-brain and multi-hand configurations for reliable long-running autonomous agents.<\/p>\n \u539f\u6587\u94fe\u63a5<\/a><\/p>\n<\/li>\n AI-generated actors and scripts are now ineligible for Oscars<\/strong>\uff08TechCrunch AI\uff09<\/p>\n \u4e2d\u6587\u6458\u8981\uff1a<\/strong>\u7f8e\u56fd\u7535\u5f71\u827a\u672f\u4e0e\u79d1\u5b66\u5b66\u9662\uff08\u5965\u65af\u5361\u4e3b\u529e\u65b9\uff09\u53d1\u5e03\u65b0\u89c4\uff0c\u660e\u786e\u89c4\u5b9a AI \u751f\u6210\u7684\u6f14\u5458\u8868\u6f14\u548c\u5267\u672c\u5c06\u6ca1\u6709\u8d44\u683c\u83b7\u5f97\u5965\u65af\u5361\u5956\u3002\u6839\u636e\u65b0\u89c4\u5219\uff0c\u53ea\u6709"\u5728\u7535\u5f71\u6cd5\u5b9a\u6f14\u804c\u5458\u8868\u4e2d\u7f72\u540d\u3001\u4e14\u7531\u4eba\u7c7b\u5b9e\u9645\u8868\u6f14\u5e76\u83b7\u5f97\u5176\u540c\u610f"\u7684\u8868\u6f14\u624d\u5177\u5907\u53c2\u8bc4\u8d44\u683c\uff1b\u5267\u672c\u5219\u5fc5\u987b\u662f"\u4eba\u7c7b\u521b\u4f5c"\u3002\u5b66\u9662\u4fdd\u7559\u8981\u6c42\u63d0\u4f9b\u5f71\u7247 AI \u4f7f\u7528\u60c5\u51b5\u548c\u4eba\u7c7b\u521b\u4f5c\u8bc1\u660e\u7684\u66f4\u591a\u4fe1\u606f\u7684\u6743\u5229\u3002\u8fd9\u4e00\u89c4\u5219\u53d8\u5316\u6b63\u503c AI \u751f\u6210\u6f14\u5458\u5f15\u53d1\u4e89\u8bae\u4e4b\u9645\u2014\u2014\u5305\u62ec\u4e00\u90e8\u4f7f\u7528 AI \u7248 Val Kilmer \u7684\u72ec\u7acb\u7535\u5f71\u6b63\u5728\u5236\u4f5c\u4e2d\uff0c\u4ee5\u53ca AI "\u6f14\u5458" Tilly Norwood \u6301\u7eed\u5f15\u53d1\u5173\u6ce8\u3002AI \u95ee\u9898\u4e5f\u662f 2023 \u5e74\u6f14\u5458\u548c\u7f16\u5267\u7f62\u5de5\u7684\u4e3b\u8981\u4e89\u8bae\u70b9\u4e4b\u4e00\u3002\u9664\u597d\u83b1\u575e\u5916\uff0c\u81f3\u5c11\u6709\u4e00\u90e8\u5c0f\u8bf4\u56e0\u7591\u4f3c\u4f7f\u7528 AI \u88ab\u51fa\u7248\u793e\u64a4\u56de\uff0c\u5176\u4ed6\u4f5c\u5bb6\u56e2\u4f53\u4e5f\u58f0\u660e AI \u751f\u6210\u7684\u4f5c\u54c1\u4e0d\u5f97\u53c2\u8bc4\u5956\u9879\u3002<\/p>\n English Summary:<\/strong> The Academy of Motion Picture Arts and Sciences released new Oscar rules stating that AI-generated actor performances and screenplays are now ineligible for Academy Awards. Only performances "credited in the film's legal billing and demonstrably performed by humans with their consent" qualify, and screenplays must be "human-authored." The academy reserves the right to request additional information about AI usage and human authorship. The rule change comes amid controversy over AI-generated actors, including an independent film using an AI version of Val Kilmer and AI "actress" Tilly Norwood making headlines. AI was a major sticking point in the 2023 actors' and writers' strikes. Outside Hollywood, at least one novel has been pulled by its publisher over AI concerns, and writers' groups are declaring AI-generated work ineligible for awards.<\/p>\n \u539f\u6587\u94fe\u63a5<\/a><\/p>\n<\/li>\n [AINews] AI Engineer World's Fair \u2014 Autoresearch, Memory, World Models, Tokenmaxxing, Agentic Commerce, and Vertical AI Call for Speakers<\/strong>\uff08Latent Space\uff09<\/p>\n \u4e2d\u6587\u6458\u8981\uff1a<\/strong>AI Engineer World's Fair \u5927\u4f1a\u5f00\u542f\u7b2c\u4e8c\u6ce2\u6f14\u8bb2\u8005\u62db\u52df\uff0c\u65b0\u589e\u591a\u4e2a\u524d\u6cbf\u6280\u672f\u4e13\u9898\u8d5b\u9053\uff0c\u5305\u62ec\u81ea\u4e3b\u7814\u7a76\uff08Autoresearch\uff09\u3001\u8bb0\u5fc6\u4e0e\u5b66\u4e60\uff08Memory\uff09\u3001\u4e16\u754c\u6a21\u578b\uff08World Models\uff09\u3001Token\u6548\u7387\u4f18\u5316\uff08Tokenmaxxing\uff09\u3001\u667a\u80fd\u4f53\u5546\u4e1a\uff08Agentic Commerce\uff09\u4ee5\u53ca\u6cd5\u5f8b\u3001\u533b\u7597\u3001GTM\u548c\u91d1\u878d\u7b49\u5782\u76f4AI\u9886\u57df\u3002\u6b64\u5916\uff0c\u5927\u4f1a\u8fd8\u7279\u522b\u4e3a\u673a\u5668\u4eba\u6f14\u793a\u9884\u7559\u514d\u8d39\u5c55\u89c8\u7a7a\u95f4\uff0c\u5e76\u65b0\u589e\u521d\u521b\u4f01\u4e1aBattlefield\u73af\u8282\uff0c\u4e3aPre-A\u8f6e\u516c\u53f8\u63d0\u4f9b\u5411\u9876\u7ea7VC\u5c55\u793a\u7684\u673a\u4f1a\u3002<\/p>\n English Summary:<\/strong> AI Engineer World's Fair announced Wave 2 Call for Speakers with new tracks covering Autoresearch, Memory, World Models, Tokenmaxxing, Agentic Commerce, and Vertical AI in Law, Healthcare, GTM and Finance. The event will also allocate free expo floor space for robotics demos and introduce a Startup Battlefield for pre-series A companies to pitch to top VCs.<\/p>\n \u539f\u6587\u94fe\u63a5<\/a><\/p>\n<\/li>\n DuckLake 1.0: Data Lake Format with SQL Catalog Metadata<\/strong>\uff08InfoQ AI\/ML\uff09<\/p>\n \u4e2d\u6587\u6458\u8981\uff1a<\/strong>DuckDB Labs \u6b63\u5f0f\u53d1\u5e03 DuckLake 1.0\uff0c\u8fd9\u662f\u4e00\u79cd\u65b0\u578b\u6570\u636e\u6e56\u683c\u5f0f\uff0c\u5c06\u8868\u5143\u6570\u636e\u5b58\u50a8\u5728 SQL \u6570\u636e\u5e93\u4e2d\u800c\u975e\u5206\u6563\u5728\u5bf9\u8c61\u5b58\u50a8\u7684\u591a\u4e2a\u6587\u4ef6\u4e2d\u3002DuckLake \u89e3\u51b3\u4e86\u4f20\u7edf\u6e56\u4ed3\u683c\u5f0f\uff08\u5982 Iceberg\u3001Delta Lake\uff09\u4e2d\u5143\u6570\u636e\u64cd\u4f5c\u590d\u6742\u3001\u534f\u8c03\u56f0\u96be\u548c\u5c0f\u6587\u4ef6\u6cdb\u6ee5\u7b49\u95ee\u9898\u30021.0 \u7248\u672c\u652f\u6301\u6570\u636e\u5185\u8054\uff08Data Inlining\uff09\u4ee5\u5904\u7406\u5c0f\u578b\u589e\u5220\u6539\u64cd\u4f5c\u3001\u6392\u5e8f\u8868\u3001\u6876\u5206\u533a\u3001\u51e0\u4f55\u6570\u636e\u7c7b\u578b\u6539\u8fdb\u4ee5\u53ca\u4e0e Iceberg \u517c\u5bb9\u7684\u5220\u9664\u5411\u91cf\u3002DuckLake \u5ba2\u6237\u7aef\u5df2\u652f\u6301 Apache DataFusion\u3001Spark\u3001Trino \u548c Pandas\u3002<\/p>\n English Summary:<\/strong> DuckDB Labs released DuckLake 1.0, a data lake format that stores table metadata in a SQL database rather than across many files in object storage. It addresses metadata coordination complexity and the small file problem in traditional lakehouse formats. Features include data inlining for small updates, sorted tables, bucket partitioning, and Iceberg-compatible deletion vectors. Clients are available for Apache DataFusion, Spark, Trino, and Pandas.<\/p>\n \u539f\u6587\u94fe\u63a5<\/a><\/p>\n<\/li>\n AWS Transform now automates BI migration to Amazon Quick in days<\/strong>\uff08AWS ML Blog\uff09<\/p>\n \u4e2d\u6587\u6458\u8981\uff1a<\/strong>AWS Transform \u63a8\u51fa\u81ea\u52a8\u5316 BI \u8fc1\u79fb\u529f\u80fd\uff0c\u53ef\u5c06\u4f20\u7edf BI \u5de5\u5177\uff08\u5982 Power BI \u548c Tableau\uff09\u7684\u4eea\u8868\u677f\u8fc1\u79fb\u81f3 Amazon QuickSight\uff0c\u5c06\u539f\u672c\u9700\u8981\u6570\u6708\u7684\u8fc1\u79fb\u5de5\u4f5c\u7f29\u77ed\u81f3\u6570\u5929\u3002\u8be5\u65b9\u6848\u901a\u8fc7 AWS Marketplace \u63d0\u4f9b\u7684\u4e13\u4e1a\u4ee3\u7406\uff08Analyzer \u548c Converter\uff09\u5b9e\u73b0\uff0c\u91c7\u7528\u4e24\u6b65\u6d41\u7a0b\uff1a\u9996\u5148\u5206\u6790\u73b0\u6709 BI \u73af\u5883\u7684\u5143\u6570\u636e\u548c\u4f9d\u8d56\u5173\u7cfb\uff0c\u7136\u540e\u81ea\u52a8\u8f6c\u6362\u6570\u636e\u96c6\u3001\u8ba1\u7b97\u5b57\u6bb5\u3001\u53ef\u89c6\u5316\u56fe\u8868\u7b49\u5230 QuickSight\u3002\u6574\u4e2a\u8fc7\u7a0b\u5728\u5ba2\u6237 AWS \u8d26\u6237\u5185\u5b8c\u6210\uff0c\u786e\u4fdd\u6570\u636e\u5b89\u5168\u3002<\/p>\n English Summary:<\/strong> AWS Transform now automates BI migration to Amazon QuickSight, reducing migration timelines from months to days. The solution uses specialized agents (Analyzer and Converter) available through AWS Marketplace to migrate from Power BI and Tableau. The two-step process involves analyzing existing BI metadata and dependencies, then converting datasets, calculated fields, and visualizations to QuickSight. All operations run within the customer's AWS account for security.<\/p>\n \u539f\u6587\u94fe\u63a5<\/a><\/p>\n<\/li>\n [AINews] Agents for Everything Else: Codex for Knowledge Work, Claude for Creative Work<\/strong>\uff08Latent Space\uff09<\/p>\n \u4e2d\u6587\u6458\u8981\uff1a<\/strong>AI \u667a\u80fd\u4f53\u6b63\u4ece\u7f16\u7a0b\u9886\u57df"\u7834\u5708"\u6269\u5c55\u81f3\u66f4\u5e7f\u6cdb\u7684\u77e5\u8bc6\u5de5\u4f5c\u548c\u521b\u610f\u5de5\u4f5c\u573a\u666f\u3002OpenAI \u7684 Codex \u53d1\u5e03\u91cd\u5927\u66f4\u65b0\uff0c\u5b9a\u4f4d\u4e3a\u9762\u5411\u6240\u6709\u4eba\u7684\u901a\u7528\u8ba1\u7b97\u673a\u4efb\u52a1\u52a9\u624b\uff0c\u65b0\u589e\u52a8\u6001 UI\u3001\u54cd\u5e94\u5f0f\u6d4f\u89c8\u5668\u3001\u4e0e Microsoft\/Google\/Salesforce \u529e\u516c\u5957\u4ef6\u7684\u96c6\u6210\uff0c\u4ee5\u53ca\u9488\u5bf9\u6587\u6863\u3001\u5e7b\u706f\u7247\u3001\u8868\u683c\u7b49\u975e\u7f16\u7801\u4efb\u52a1\u7684\u652f\u6301\u3002\u4e0e\u6b64\u540c\u65f6\uff0cAnthropic \u7684 Claude \u63a8\u51fa\u5b89\u5168\u4ee3\u7801\u5ba1\u67e5\u5de5\u5177 Claude Security\uff0c\u5e76\u6269\u5c55\u4e86\u5bf9 Blender\u3001Adobe Creative Cloud\u3001Ableton \u7b49\u521b\u610f\u5de5\u5177\u7684\u652f\u6301\uff0c\u5f62\u6210"Codex \u4e3b\u653b\u77e5\u8bc6\u5de5\u4f5c\u3001Claude \u4e3b\u653b\u521b\u610f\u5de5\u4f5c"\u7684\u53cc\u96c4\u683c\u5c40\u3002<\/p>\n English Summary:<\/strong> AI agents are expanding beyond coding into knowledge work and creative domains. OpenAI's Codex received a major update positioning it as a general computer-use agent for everyone, featuring dynamic UI, responsive browser, integration with Microsoft\/Google\/Salesforce suites, and support for documents, slides, and spreadsheets.<\/p>\n \u539f\u6587\u94fe\u63a5<\/a><\/p>\n<\/li>\n GitHub Copilot CLI for Beginners: Interactive v. non-interactive mode<\/strong>\uff08GitHub AI\/ML\uff09<\/p>\n \u4e2d\u6587\u6458\u8981\uff1a<\/strong>GitHub \u53d1\u5e03 Copilot CLI \u521d\u5b66\u8005\u6307\u5357\uff0c\u8be6\u7ec6\u4ecb\u7ecd\u4ea4\u4e92\u5f0f\uff08interactive\uff09\u548c\u975e\u4ea4\u4e92\u5f0f\uff08non-interactive\uff09\u4e24\u79cd\u6a21\u5f0f\u7684\u4f7f\u7528\u573a\u666f\u4e0e\u533a\u522b\u3002\u4ea4\u4e92\u5f0f\u6a21\u5f0f\u63d0\u4f9b\u7c7b\u4f3c\u804a\u5929\u7684\u6765\u56de\u5bf9\u8bdd\u4f53\u9a8c\uff0c\u9002\u5408\u63a2\u7d22\u6027\u3001\u6df1\u5ea6\u534f\u4f5c\u7684\u5de5\u4f5c\uff1b\u975e\u4ea4\u4e92\u5f0f\u6a21\u5f0f\u5219\u901a\u8fc7\u547d\u4ee4\u884c\u76f4\u63a5\u4f20\u9012\u5355\u4e2a\u63d0\u793a\u8bcd\u83b7\u53d6\u5feb\u901f\u56de\u7b54\uff0c\u9002\u5408\u4ed3\u5e93\u6458\u8981\u3001\u4ee3\u7801\u7247\u6bb5\u751f\u6210\u6216\u81ea\u52a8\u5316\u5de5\u4f5c\u6d41\u7b49\u4e00\u6b21\u6027\u4efb\u52a1\u3002\u7528\u6237\u8fd8\u53ef\u901a\u8fc7 `\/resume` \u6216 `–resume` \u547d\u4ee4\u6062\u590d\u4e4b\u524d\u7684\u4f1a\u8bdd\uff0c\u4fdd\u7559\u5b8c\u6574\u4e0a\u4e0b\u6587\u3002<\/p>\n English Summary:<\/strong> GitHub published a beginner's guide for Copilot CLI explaining interactive and non-interactive modes. Interactive mode offers a chat-like back-and-forth experience for exploratory, hands-on work, while non-interactive mode allows passing a single prompt directly from the command line for quick one-shot tasks like repository summarization or code generation. Users can resume previous sessions with `\/resume` or `–resume` commands to retain full context.<\/p>\n \u539f\u6587\u94fe\u63a5<\/a><\/p>\n<\/li>\n Introducing Advanced Account Security<\/strong>\uff08OpenAI News\uff09<\/p>\n \u4e2d\u6587\u6458\u8981\uff1a<\/strong>OpenAI \u63a8\u51fa Advanced Account Security\uff08\u9ad8\u7ea7\u8d26\u6237\u5b89\u5168\uff09\u529f\u80fd\uff0c\u4e3a ChatGPT \u548c Codex \u7528\u6237\u63d0\u4f9b\u53ef\u9009\u7684\u5f3a\u5316\u5b89\u5168\u4fdd\u62a4\u3002\u8be5\u529f\u80fd\u8981\u6c42\u4f7f\u7528\u901a\u884c\u5bc6\u94a5\u6216\u7269\u7406\u5b89\u5168\u5bc6\u94a5\u767b\u5f55\uff0c\u7981\u7528\u5bc6\u7801\u767b\u5f55\u548c\u90ae\u4ef6\/SMS \u6062\u590d\u65b9\u5f0f\uff0c\u7f29\u77ed\u4f1a\u8bdd\u6709\u6548\u671f\uff0c\u5e76\u81ea\u52a8\u6392\u9664\u5bf9\u8bdd\u6570\u636e\u7528\u4e8e\u6a21\u578b\u8bad\u7ec3\u3002OpenAI \u4e0e Yubico \u5408\u4f5c\u63d0\u4f9b\u4f18\u60e0\u7684\u5b89\u5168\u5bc6\u94a5\u5957\u88c5\uff0c\u8be5\u529f\u80fd\u4e3b\u8981\u9488\u5bf9\u8bb0\u8005\u3001\u653f\u6cbb\u4eba\u7269\u3001\u7814\u7a76\u4eba\u5458\u7b49\u9ad8\u98ce\u9669\u7528\u6237\u7fa4\u4f53\u3002\u4ece 2026 \u5e74 6 \u6708 1 \u65e5\u8d77\uff0cTrusted Access for Cyber \u9879\u76ee\u7684\u6210\u5458\u5fc5\u987b\u542f\u7528\u6b64\u529f\u80fd\u3002<\/p>\n English Summary:<\/strong> OpenAI introduces Advanced Account Security, an opt-in feature providing enhanced protections for ChatGPT and Codex accounts. It mandates passkeys or physical security keys for login, disables password-based authentication and email\/SMS recovery, shortens session durations, and automatically excludes conversations from model training. OpenAI partnered with Yubico to offer discounted security key bundles.<\/p>\n \u539f\u6587\u94fe\u63a5<\/a><\/p>\n<\/li>\n Where the goblins came from<\/strong>\uff08OpenAI News\uff09<\/p>\n \u4e2d\u6587\u6458\u8981\uff1a<\/strong>OpenAI \u53d1\u5e03\u6280\u672f\u535a\u5ba2\u6df1\u5165\u89e3\u6790 GPT-5 \u7cfb\u5217\u6a21\u578b\u4e2d\u9891\u7e41\u51fa\u73b0"\u5730\u7cbe\/\u5c0f\u5996\u7cbe"\uff08goblin\/gremlin\uff09\u7b49\u5947\u5e7b\u751f\u7269\u9690\u55bb\u7684\u73b0\u8c61\u3002\u8c03\u67e5\u8ffd\u6eaf\u53d1\u73b0\uff0c\u8be5\u884c\u4e3a\u6e90\u81ea\u4e3a"Nerdy"\u4e2a\u6027\u5b9a\u5236\u529f\u80fd\u8bbe\u8ba1\u7684\u5f3a\u5316\u5b66\u4e60\u5956\u52b1\u4fe1\u53f7\u2014\u2014\u8be5\u5956\u52b1\u673a\u5236\u5bf9\u5305\u542b\u5947\u5e7b\u751f\u7269\u7684\u9690\u55bb\u8f93\u51fa\u7ed9\u4e88\u9ad8\u5206\uff0c\u5bfc\u81f4\u6a21\u578b\u5728\u8bad\u7ec3\u4e2d\u5c06\u6b64\u98ce\u683c\u6cdb\u5316\u5230\u5176\u4ed6\u573a\u666f\u3002\u5c3d\u7ba1"Nerdy"\u4e2a\u6027\u4ec5\u5360 ChatGPT \u6d41\u91cf\u7684 2.5%\uff0c\u5374\u4ea7\u751f\u4e86 66.7% \u7684"goblin"\u63d0\u53ca\u3002OpenAI \u5df2\u4e8e 3 \u6708\u4e0b\u7ebf\u8be5\u4e2a\u6027\u8bbe\u7f6e\uff0c\u5e76\u5f00\u53d1\u4e86\u65b0\u7684\u6a21\u578b\u884c\u4e3a\u5ba1\u8ba1\u5de5\u5177\u3002<\/p>\n English Summary:<\/strong> OpenAI published a technical blog investigating why GPT-5 models frequently used goblin and gremlin metaphors. The root cause was traced to a reinforcement learning reward signal designed for the "Nerdy" personality customization feature, which inadvertently favored outputs containing creature metaphors. Though the Nerdy personality accounted for only 2.5% of ChatGPT traffic, it generated 66.7% of all "goblin" mentions. OpenAI retired the Nerdy personality in March and developed new tools for auditing and fixing emergent model behaviors.<\/p>\n \u539f\u6587\u94fe\u63a5<\/a><\/p>\n<\/li>\n Reading today's open-closed performance gap<\/strong>\uff08Interconnects\uff09<\/p>\n \u4e2d\u6587\u6458\u8981\uff1a<\/strong>Interconnects AI \u535a\u5ba2\u5206\u6790\u5f53\u524d\u5f00\u6e90\u4e0e\u95ed\u6e90\u5927\u6a21\u578b\u4e4b\u95f4\u7684\u6027\u80fd\u5dee\u8ddd\u53ca\u5176\u590d\u6742\u6027\u3002\u6587\u7ae0\u6307\u51fa\uff0c\u5355\u4e00\u7684\u7efc\u5408\u8bc4\u6d4b\u5206\u6570\u63a9\u76d6\u4e86\u6a21\u578b\u5728\u4e0d\u540c\u80fd\u529b\u7ef4\u5ea6\u4e0a\u7684\u771f\u5b9e\u8868\u73b0\u5dee\u5f02\u3002\u5f53\u524d\u884c\u4e1a\u7126\u70b9\u5df2\u4ece\u6570\u5b66\u548c\u7b80\u5355\u4ee3\u7801\u8f6c\u5411\u590d\u6742\u7f16\u7a0b\u548c\u667a\u80fd\u4f53\u4efb\u52a1\uff0c\u95ed\u6e90\u5b9e\u9a8c\u5ba4\u5728\u8fd9\u4e9b\u9886\u57df\u6295\u5165\u5de8\u8d44\u3002\u4f5c\u8005\u8ba4\u4e3a\uff0c\u968f\u7740\u4efb\u52a1\u590d\u6742\u5ea6\u63d0\u5347\uff0c\u5f00\u6e90\u6a21\u578b\u5728\u83b7\u53d6\u9ad8\u8d28\u91cf\u8bad\u7ec3\u73af\u5883\u548c\u6570\u636e\u65b9\u9762\u5c06\u9762\u4e34\u66f4\u5927\u6311\u6218\uff0c\u4f46\u5f00\u6e90\u6a21\u578b\u901a\u8fc7\u84b8\u998f\u7b49\u65b9\u5f0f\u5feb\u901f\u8ffd\u8d76\u7684\u80fd\u529b\u4e5f\u4e0d\u5bb9\u4f4e\u4f30\u3002\u8bc4\u6d4b\u57fa\u51c6\u7684\u53ef\u4fe1\u5ea6\u6b63\u5728\u4e0b\u964d\uff0c\u771f\u5b9e\u4e16\u754c\u8868\u73b0\u4e0e\u57fa\u51c6\u5206\u6570\u4e4b\u95f4\u7684\u76f8\u5173\u6027\u53d8\u5f97\u66f4\u4e3a\u590d\u6742\u3002<\/p>\n English Summary:<\/strong> Interconnects AI analyzes the nuanced performance gap between open and closed AI models, arguing that composite benchmark scores obscure important capability differences. As industry focus shifts from math and simple coding to complex programming and agentic tasks, closed labs invest heavily in these domains. The author notes that open models face increasing challenges in accessing high-quality training environments and data as tasks grow more complex, though their ability to catch up through distillation remains significant.<\/p>\n \u539f\u6587\u94fe\u63a5<\/a><\/p>\n<\/li>\n Building an emoji list generator with the GitHub Copilot CLI<\/strong>\uff08GitHub AI\/ML\uff09<\/p>\n \u4e2d\u6587\u6458\u8981\uff1a<\/strong>GitHub \u535a\u5ba2\u4ecb\u7ecd\u5982\u4f55\u4f7f\u7528 GitHub Copilot CLI \u6784\u5efa\u4e00\u4e2a\u8868\u60c5\u7b26\u53f7\u5217\u8868\u751f\u6210\u5668\u3002\u8be5\u9879\u76ee\u5728 Rubber Duck Thursday \u76f4\u64ad\u4e2d\u5f00\u53d1\uff0c\u4f7f\u7528 @opentui\/core \u6784\u5efa\u7ec8\u7aef\u754c\u9762\u3001@github\/copilot-sdk \u63d0\u4f9b AI \u80fd\u529b\u3001clipboardy \u5b9e\u73b0\u526a\u8d34\u677f\u529f\u80fd\u3002\u7528\u6237\u53ef\u5728\u7ec8\u7aef\u7c98\u8d34\u6216\u8f93\u5165\u5217\u8868\uff0c\u6309 Ctrl+S \u540e AI \u81ea\u52a8\u4e3a\u6bcf\u884c\u6dfb\u52a0\u76f8\u5173\u8868\u60c5\u7b26\u53f7\u5e76\u590d\u5236\u5230\u526a\u8d34\u677f\u3002\u5f00\u53d1\u8fc7\u7a0b\u5c55\u793a\u4e86 Copilot CLI \u7684 Plan \u6a21\u5f0f\u3001Autopilot \u6a21\u5f0f\u3001\u591a\u6a21\u578b\u5de5\u4f5c\u6d41\uff08Claude Sonnet 4.6 \u548c Opus 4.7\uff09\u4ee5\u53ca GitHub MCP \u670d\u52a1\u5668\u7b49\u7279\u6027\u3002\u9879\u76ee\u5df2\u5f00\u6e90\u3002<\/p>\n English Summary:<\/strong> GitHub's blog demonstrates building an emoji list generator using the GitHub Copilot CLI during a live Rubber Duck Thursday stream. The project uses @opentui\/core for terminal UI, @github\/copilot-sdk for AI capabilities, and clipboardy for clipboard access. Users paste or type a list in the terminal, press Ctrl+S, and AI automatically adds relevant emojis to each line before copying to clipboard. The development showcased Copilot CLI's Plan mode, Autopilot mode, multi-model workflow (Claude Sonnet 4.6 and Opus 4.7), and GitHub MCP server.<\/p>\n