[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"article-ai-weekly-2026-w32-en":3,"article-related-ai-weekly-2026-w32-en":28,"series-industry-2d727daf-559d-44b3-bc2b-3ff53137afef":71},{"id":4,"slug":5,"title":6,"content":7,"summary":8,"source":9,"source_url":10,"author":11,"image_url":12,"cover_image":12,"category":13,"language":14,"translated_content":11,"related_article_id":15,"keywords":16,"key_takeaways":11,"views":25,"created_at":26,"published_at":27,"topic_cluster_id":11},"2d727daf-559d-44b3-bc2b-3ff53137afef","ai-weekly-2026-w32-en","AI Weekly: 2026-07-27 ~ 2026-08-03","\u003Cp>Frontier AI this week split into two tracks: cheaper, more aggressive model releases on one side, and sharper scrutiny of safety and reliability on the other. The practical winner was agent tooling, where small product fixes and workflow framing kept moving from demos toward daily use.\u003C\u002Fp>\u003Ch2>Trend Radar\u003C\u002Fh2>\u003Ctable>\u003Cthead>\u003Ctr>\u003Cth>Dimension\u003C\u002Fth>\u003Cth>Signal\u003C\u002Fth>\u003Cth>This Week\u003C\u002Fth>\u003Cth>What's at Stake\u003C\u002Fth>\u003C\u002Ftr>\u003C\u002Fthead>\u003Ctbody>\u003Ctr>\u003Ctd>Models\u003C\u002Ftd>\u003Ctd>Strong\u003C\u002Ftd>\u003Ctd>Anthropic’s Claude Opus 5 and Kimi K3 pushed pricing, capability, and open-weight debate back to center stage.\u003C\u002Ftd>\u003Ctd>Model competition is shifting from raw benchmark bragging to cost, access, and trust.\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>Agents\u003C\u002Ftd>\u003Ctd>Medium\u003C\u002Ftd>\u003Ctd>Grok Build 0.2.111 added live previews, rewind fixes, and queue handling improvements.\u003C\u002Ftd>\u003Ctd>Agent products are becoming easier to debug, which matters more than flashy autonomy claims.\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>Open Source\u003C\u002Ftd>\u003Ctd>Medium\u003C\u002Ftd>\u003Ctd>Kimi K3’s release intensified attention on open weights and cross-border model strategy.\u003C\u002Ftd>\u003Ctd>Open release strategy is now a competitive weapon, not a side note.\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>Compute & Infra\u003C\u002Ftd>\u003Ctd>Weak\u003C\u002Ftd>\u003Ctd>AI reliability and infrastructure stayed in focus after last week’s outage and the Georgia campus plan continued to frame capacity debates.\u003C\u002Ftd>\u003Ctd>Training and serving capacity remain a strategic bottleneck, especially for frontier labs.\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>Applications\u003C\u002Ftd>\u003Ctd>Weak\u003C\u002Ftd>\u003Ctd>Prompt Engineering vs Loop Engineering vs Graph Engineering reframed how teams think about orchestration layers.\u003C\u002Ftd>\u003Ctd>Builders are moving from prompt craft to system design, which changes team structure and tooling choices.\u003C\u002Ftd>\u003C\u002Ftr>\u003Ctr>\u003Ctd>Policy & Regulation\u003C\u002Ftd>\u003Ctd>Medium\u003C\u002Ftd>\u003Ctd>Anthropic’s safety audit results for recent models kept model evaluation under the microscope.\u003C\u002Ftd>\u003Ctd>Independent safety scoring is becoming part of product positioning, not just compliance theater.\u003C\u002Ftd>\u003C\u002Ftr>\u003C\u002Ftbody>\u003C\u002Ftable>\u003Ch2>Key Stories\u003C\u002Fh2>\u003Ch3>Claude Opus 5 shows that safety is now a product metric\u003C\u002Fh3>\u003Cp>\u003Cstrong>What happened.\u003C\u002Fstrong> Anthropic’s Claude Opus 5 was reported to post the lowest safety audit score among four recent Anthropic models, with a 2.30 result in a deceptive or harmful behavior audit.\u003C\u002Fp>\n\u003Cfigure class=\"my-6\">\u003Cimg src=\"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1785730828896-l1kg.png\" alt=\"AI Weekly: 2026-07-27 ~ 2026-08-03\" class=\"rounded-xl w-full\" loading=\"lazy\" \u002F>\u003C\u002Ffigure>\n\n\u003Cfigure class=\"my-6\">\u003Cimg src=\"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1785730820794-zd9i.png\" alt=\"AI Weekly: 2026-07-27 ~ 2026-08-03\" class=\"rounded-xl w-full\" loading=\"lazy\" \u002F>\u003C\u002Ffigure>\n\u003Cp>\u003Cstrong>Why it matters.\u003C\u002Fstrong> That kind of number changes the conversation from “how capable is the model?” to “how much risk comes with the capability?” For enterprise buyers, safety evaluation is becoming part of procurement math, especially when models are deployed in agentic workflows that can act, not just answer.\u003C\u002Fp>\u003Cp>\u003Cstrong>Who's affected and next to watch.\u003C\u002Fstrong> Anthropic, enterprise AI buyers, and compliance teams should watch whether the company publishes more detail on audit methodology and whether customers start asking for comparable third-party scores in RFPs.\u003C\u002Fp>\u003Ch3>Anthropic’s Opus 5 pricing push raises the pressure on frontier labs\u003C\u002Fh3>\u003Cp>\u003Cstrong>What happened.\u003C\u002Fstrong> Coverage around Anthropic’s Opus 5 launch emphasized lower pricing, privacy positioning, model tiers, and the broader business implications of timing around a possible IPO.\u003C\u002Fp>\u003Cp>\u003Cstrong>Why it matters.\u003C\u002Fstrong> The market is no longer rewarding only the most capable model; it is rewarding the cheapest credible model with enough trust to sell into real workflows. That pushes rivals to defend margins, and it makes the economics of serving frontier models a first-order strategic issue.\u003C\u002Fp>\u003Cp>\u003Cstrong>Who's affected and next to watch.\u003C\u002Fstrong> Anthropic, OpenAI, and model-hosting partners should be watched for pricing reactions, tier changes, and any shift in how premium models are packaged for enterprise use.\u003C\u002Fp>\u003Ch3>Kimi K3 deepens the split over open weights and China’s model pace\u003C\u002Fh3>\u003Cp>\u003Cstrong>What happened.\u003C\u002Fstrong> Kimi K3’s release triggered sharper debate in U.S. AI circles over open weights, China’s model cadence, and what frontier competition now looks like.\u003C\u002Fp>\u003Cp>\u003Cstrong>Why it matters.\u003C\u002Fstrong> The argument is no longer abstract. Open-weight releases can accelerate adoption, lower switching costs, and force incumbents to justify closed deployment models with more than branding. They also complicate the usual assumption that the most important progress only comes from U.S. labs.\u003C\u002Fp>\u003Cp>\u003Cstrong>Who's affected and next to watch.\u003C\u002Fstrong> U.S. model labs, developers choosing between hosted and self-hosted stacks, and policy watchers should monitor whether Kimi K3 gets broad community uptake or remains mostly a geopolitical talking point.\u003C\u002Fp>\u003Ch3>Grok Build keeps agent work grounded in debugging, not hype\u003C\u002Fh3>\u003Cp>\u003Cstrong>What happened.\u003C\u002Fstrong> Grok Build 0.2.111 shipped with live previews, stronger rewind behavior, and better queue handling for long-running sessions.\u003C\u002Fp>\u003Cp>\u003Cstrong>Why it matters.\u003C\u002Fstrong> These are the kinds of changes that make agents usable by actual teams. Live previews reduce the cost of waiting, rewind reduces the pain of mistakes, and queue handling makes longer workflows less fragile. That is the unglamorous work that turns agent platforms into tools people trust.\u003C\u002Fp>\u003Cp>\u003Cstrong>Who's affected and next to watch.\u003C\u002Fstrong> Builders using Grok Build and similar agent frameworks should watch whether the product adds more inspection, rollback, and state-management features rather than more autonomy slogans.\u003C\u002Fp>\u003Ch3>Prompt engineering is giving way to system engineering\u003C\u002Fh3>\u003Cp>\u003Cstrong>What happened.\u003C\u002Fstrong> A side-by-side look at prompt engineering, loop engineering, and graph engineering framed three layers of AI application design and how each changes control, cost, and orchestration.\u003C\u002Fp>\u003Cp>\u003Cstrong>Why it matters.\u003C\u002Fstrong> This is a useful sign that the market is maturing. Teams are moving away from treating prompts as the main interface and toward designing repeatable control flows, retries, and dependency graphs. That favors developers who can reason about state, not just wording.\u003C\u002Fp>\u003Cp>\u003Cstrong>Who's affected and next to watch.\u003C\u002Fstrong> Product teams building agentic apps, internal AI platform groups, and workflow tooling vendors should watch for more products that expose loops, branching, and graph controls directly in the UI.\u003C\u002Fp>\u003Ch2>Watch Next Week\u003C\u002Fh2>\u003Cul>\u003Cli>Anthropic follow-up on Claude Opus 5 safety methodology and any enterprise-facing clarification.\u003C\u002Fli>\u003Cli>OpenAI and Anthropic pricing moves after the pressure from Opus 5 and Kimi K3.\u003C\u002Fli>\u003Cli>Kimi K3 adoption signals in developer communities and hosted model platforms.\u003C\u002Fli>\u003Cli>Grok Build release notes for further debugging, rollback, or state-management features.\u003C\u002Fli>\u003Cli>Any new infrastructure announcements tied to large-scale campus buildouts or capacity expansion plans.\u003C\u002Fli>\u003C\u002Ful>","Frontier model pricing and safety are splitting the market, while agent tooling keeps getting more practical and infra remains a hard constraint.","oracore.dev","https:\u002F\u002Foracore.dev\u002Fnews\u002Fai-weekly-2026-w32-en",null,"https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1785730828896-l1kg.png","industry","en","e90c712a-4532-416a-8e07-05c01b39196b",[17,18,19,20,21,22,23,24],"AI Weekly","AI news","trend radar","artificial intelligence","Claude Opus 5","Kimi K3","AI agents","model safety",0,"2026-08-03T04:00:39.96111+00:00","2026-08-03T04:00:39.937+00:00",{"tags":29,"relatedLang":30,"relatedPosts":34},[],{"id":15,"slug":31,"title":32,"language":33},"ai-weekly-2026-w32-zh","AI 週報：2026-07-27 ~ 2026-08-03","zh",[35,41,47,53,59,65],{"id":36,"slug":37,"title":38,"cover_image":39,"image_url":39,"created_at":40,"category":13},"e746ff12-ac66-4bc2-9492-a33e2f921677","kimi-k3-test-time-scaling-rules-en","Kimi K3 maps the new rules of test-time scaling","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1785718974890-fnrs.png","2026-08-03T01:02:30.065567+00:00",{"id":42,"slug":43,"title":44,"cover_image":45,"image_url":45,"created_at":46,"category":13},"61ebc42d-aa1c-43c0-9d6f-3ce43f714971","salp-liquidation-ai-trades-revealed-en","What the SALP liquidation reveals about AI trades","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1785717176689-5kgk.png","2026-08-03T00:32:33.409316+00:00",{"id":48,"slug":49,"title":50,"cover_image":51,"image_url":51,"created_at":52,"category":13},"b3199b83-9b2b-45f3-ba99-8ee782d45130","claude-security-test-became-real-breach-en","Claude’s security test became a real breach","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1785715402243-s100.png","2026-08-03T00:02:55.308879+00:00",{"id":54,"slug":55,"title":56,"cover_image":57,"image_url":57,"created_at":58,"category":13},"7c78bbe7-bdbd-461c-b536-90b37dd24ac1","x-posts-let-execs-shape-the-ai-story-en","X posts let execs shape the AI story","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1785697388377-hlfd.png","2026-08-02T19:02:41.347626+00:00",{"id":60,"slug":61,"title":62,"cover_image":63,"image_url":63,"created_at":64,"category":13},"3302d464-a3d5-4550-b328-4b2c7c1a89b7","jensen-huang-agi-definition-lowers-the-bar-en","Jensen Huang’s AGI definition lowers the bar","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1785695567157-abd0.png","2026-08-02T18:32:24.259106+00:00",{"id":66,"slug":67,"title":68,"cover_image":69,"image_url":69,"created_at":70,"category":13},"017693e6-8d4a-409e-a848-5bdf697ee08d","claude-2026-limit-changes-capacity-story-en","Claude’s 2026 limit changes are a capacity story, not a product story","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1785693771298-z5r3.png","2026-08-02T18:02:25.374847+00:00",[72,77,82,87,92,97,102,107,112,117],{"id":73,"slug":74,"title":75,"created_at":76},"d35a1bd9-e709-412e-a2df-392df1dc572a","ai-impact-2026-developments-market-en","AI's Impact in 2026: Key Developments and Market Shifts","2026-03-25T16:20:33.205823+00:00",{"id":78,"slug":79,"title":80,"created_at":81},"5ed27921-5fd6-492e-8c59-78393bf37710","trumps-ai-legislative-framework-en","Trump's AI Legislative Framework: What's Inside?","2026-03-25T16:22:20.005325+00:00",{"id":83,"slug":84,"title":85,"created_at":86},"e454a642-f03c-4794-b185-5f651aebbaca","nvidia-gtc-2026-key-highlights-innovations-en","NVIDIA GTC 2026: Key Highlights and Innovations","2026-03-25T16:22:47.882615+00:00",{"id":88,"slug":89,"title":90,"created_at":91},"0ebb5b16-774a-4922-945d-5f2ce1df5a6d","claude-usage-diversifies-learning-curves-en","Claude Usage Diversifies, Learning Curves Emerge","2026-03-25T16:25:50.770376+00:00",{"id":93,"slug":94,"title":95,"created_at":96},"69934e86-2fc5-4280-8223-7b917a48ace8","openclaw-ai-commoditization-concerns-en","OpenClaw's Rise Raises Concerns of AI Model Commoditization","2026-03-25T16:26:30.582047+00:00",{"id":98,"slug":99,"title":100,"created_at":101},"b4b2575b-2ac8-46b2-b90e-ab1d7c060797","google-gemini-ai-rollout-2026-en","Google's Gemini AI Rollout Extended to 2026","2026-03-25T16:28:14.808842+00:00",{"id":103,"slug":104,"title":105,"created_at":106},"6e18bc65-42ae-4ad0-b564-67d7f66b979e","meta-llama4-fabricated-results-scandal-en","Meta's Llama 4 Scandal: Fabricated AI Test Results Unveiled","2026-03-25T16:29:15.482836+00:00",{"id":108,"slug":109,"title":110,"created_at":111},"bf888e9d-08be-4f47-996c-7b24b5ab3500","accenture-mistral-ai-deployment-en","Accenture and Mistral AI Team Up for AI Deployment","2026-03-25T16:31:01.894655+00:00",{"id":113,"slug":114,"title":115,"created_at":116},"5382b536-fad2-49c6-ac85-9eb2bae49f35","mistral-ai-high-stakes-2026-en","Mistral AI: Facing High Stakes in 2026","2026-03-25T16:31:39.941974+00:00",{"id":118,"slug":119,"title":120,"created_at":121},"9da3d2d6-b669-4971-ba1d-17fdb3548ed5","cursors-meteoric-rise-pressures-en","Cursor's Meteoric Rise Faces Industry Pressures","2026-03-25T16:32:21.899217+00:00"]