Anthropic Teaches Claude Why, OpenAI Codex Safety, Ring-2.6 | AI Brief May 9

Timeline:
00:00 Opening
00:08 Anthropic | Teaching Claude why
00:28 OpenAI | Accidental CoT grading
00:47 OpenAI | Running Codex safely
01:07 Ring-2.6 | Trillion-parameter reasoning
01:25 AllenAI | EMO mixture model
01:43 Codex | Switch to Codex
02:00 Grok | Platform connectors
02:16 Gemini | Notebook organization
02:33 Suno | Voice-only pop songs
02:47 Cursor | Bugbot pricing update
03:03 Alibaba Cloud | Smart Studio platform
03:22 OpenRouter | Human review tools
03:37 Claude Code | Release 2.1.136
03:55 China MIIT | AI device grading
04:16 Anthropic | Huge funding report
04:36 Runway | Child safety approach
04:56 Google | Small-business ads
05:14 Apple | Privacy AI workshop
05:34 Lemonade | Agentic enterprise security
05:52 Apple | RVPO alignment
06:11 Apple | Velox 4D representation
06:28 Jim Fan | Robotics endgame
06:47 Perplexity | Agent skills handbook
07:06 Douyin | Giant-form effect
07:26 Claude Code | HTML as output
07:43 Claude Code | Agent development kit
08:01 AMD | MedQA on ROCm
08:18 BAIR | Adaptive parallel reasoning
08:34 Closing

Видео Anthropic Teaches Claude Why, OpenAI Codex Safety, Ring-2.6 | AI Brief May 9 канала 綠頭鴨早報