Discover quality AI tools
Bring together the world’s leading artificial intelligence tools to help improve work efficiency
Articles published by RyanSmith
OpenAI announced a shift in AI safety governance by involving independent third-party organizations early in the development cycle for systematic risk assessments. This move aims to address growing public and regulatory concerns over AI controllability. The company outlined four core priorities for external evaluations: establishing a strong independent operational mechanism, ensuring scientific rigor, implementing robust daily safety practices, and defining clear safety responsibilities. This proactive approach seeks to build a security defense through collaboration with impartial external forces, responding to the accelerating iteration of cutting-edge AI models and long-standing scrutiny regarding technology safety.
OpenAI announced a shift in AI safety governance by involving independent third-party organizations early in the development cycle for systematic risk assessments. This move aims to address growing public and regulatory concerns over AI controllability. The company outlined four core priorities for external evaluations: establishing a strong independent operational mechanism, ensuring scientific rigor, implementing robust daily safety practices, and defining clear safety responsibilities. This proactive approach seeks to build a security defense through collaboration with impartial external forces, responding to the accelerating iteration of cutting-edge AI models and long-standing scrutiny regarding technology safety.
Cloudflare updated its crawler rules tagging bots as search AI proxy or model training. From Sept 15 2026 AI proxies and training crawlers will be blocked by default on ad-supported sites. Hybrid crawlers are blocked if any purpose violates site policy. A new dashboard adapts to AEO page change monitoring reduces duplicate requests and usage based billing lets owners charge for AI content crawling.
Cloudflare updated its crawler rules tagging bots as search AI proxy or model training. From Sept 15 2026 AI proxies and training crawlers will be blocked by default on ad-supported sites. Hybrid crawlers are blocked if any purpose violates site policy. A new dashboard adapts to AEO page change monitoring reduces duplicate requests and usage based billing lets owners charge for AI content crawling.
Anthropic's Claude models are now fully available on Microsoft Azure Foundry, accessible via existing Azure accounts with integrated identity, security, and billing. The service supports prompt caching, extended thinking, and data residency, powered by NVIDIA Blackwell Ultra for complex reasoning. Enterprises like Bolt and Momentic are already using it.
Anthropic's Claude models are now fully available on Microsoft Azure Foundry, accessible via existing Azure accounts with integrated identity, security, and billing. The service supports prompt caching, extended thinking, and data residency, powered by NVIDIA Blackwell Ultra for complex reasoning. Enterprises like Bolt and Momentic are already using it.
Zhihu reported Q1 2026 revenue of 652 million yuan, down 10.7% YoY, but adjusted net profit surged 147.2% to 17.16 million yuan. Paid content and IP operations accounted for 61.7% of revenue, driven by AI short dramas and manga dramas. The industry shows two models: mass production (Yanyan Story) and premium hits (AiKan Interactive). Despite profit improvement, revenue has declined for three consecutive years, with profitability relying on cost cuts rather than growth.
Zhihu reported Q1 2026 revenue of 652 million yuan, down 10.7% YoY, but adjusted net profit surged 147.2% to 17.16 million yuan. Paid content and IP operations accounted for 61.7% of revenue, driven by AI short dramas and manga dramas. The industry shows two models: mass production (Yanyan Story) and premium hits (AiKan Interactive). Despite profit improvement, revenue has declined for three consecutive years, with profitability relying on cost cuts rather than growth.
