https://ai-tldr.dev/releases/two-minute-papers-deepseek-speed-jul7/ AI/TLDR en 2026-07-07T12:00:00+00:00 Two Minute Papers: 'DeepSeek's New AI Speed Hack Is Amazing' https://ai-tldr.dev/releases/wes-roth-claude-conscious-jul7/ AI/TLDR en 2026-07-07T12:00:00+00:00 Wes Roth: 'CLAUDE IS CONSCIOUS' — reacting to Anthropic's global workspace research https://ai-tldr.dev/releases/sam-witteveen-hy3-jul7/ AI/TLDR en 2026-07-07T12:00:00+00:00 Sam Witteveen: 'Hy3 from Tencent - The NEW GLM Competitor' https://ai-tldr.dev/releases/ternlight-v0-1-0/ AI/TLDR en 2026-07-05T12:00:00+00:00 Ternlight — 7 MB embedding model that runs in the browser https://ai-tldr.dev/releases/xai-grok-flagship-voices-jul6/ AI/TLDR en 2026-07-06T12:00:00+00:00 xAI Grok Voice — 21 new flagship voices with speech tags for pacing https://ai-tldr.dev/releases/openai-gpt-realtime-2-1-jul6/ AI/TLDR en 2026-07-06T12:00:00+00:00 OpenAI gpt-realtime-2.1 — voice model gains reasoning-effort dial and a mini variant https://ai-tldr.dev/releases/martinalderson-glm-52-margin-collapse-jul6/ AI/TLDR en 2026-07-06T12:00:00+00:00 Martin Alderson — GLM 5.2 and the coming AI margin collapse https://ai-tldr.dev/releases/maximerivest-riddle-jul6/ AI/TLDR en 2026-07-06T12:00:00+00:00 Riddle — reMarkable Paper Pro becomes Tom Riddle's diary via vision LLM https://ai-tldr.dev/releases/anthropic-alberta-cyber-jul6/ AI/TLDR en 2026-07-06T12:00:00+00:00 Alberta government cybersecurity — Claude Code scans 466M lines in 20 hours https://ai-tldr.dev/releases/anthropic-jacobian-lens/ AI/TLDR en 2026-07-06T12:00:00+00:00 Jacobian Lens — Anthropic reads what Claude thinks but doesn't say https://ai-tldr.dev/releases/tencent-hunyuan-hy3/ AI/TLDR en 2026-07-06T12:00:00+00:00 Tencent Hunyuan Hy3 — 295B/21B MoE open-sourced with 256K context