Signal over noise, daily.

Signal over noise, daily. A practitioner briefing on what's actually shipping in AI. Every claim linked to its source.

Latest

• Apple opens Siri AI to all developers in iOS 27.3 beta Developer Preview 2026-11-05
• Prime Intellect audit finds broken graders and reward hacks in three public benchmarks 2026-11-05
• Google opens a preview REST API and MCP server for Gemini Notebook 2026-11-05
• Are AI token prices subsidised? A cost breakdown for one open model 2026-11-05
• 'You don't need long context': what the compaction paper means for coding agents 2026-11-04
• Qwen 3.8 27B on a laptop nearly matches GPT-6 Astra at reading model weights 2026-11-04
• SPAR paper: a sales agent that tests its strategies before rewriting memory 2026-11-04
• ParetoSaddler paper: apply cost preference only at the end of agent harness search 2026-11-04
• HarnessRouter pitches Harness Arena for comparing agent harness and model combinations 2026-11-04
• Pipecat Labs releases tiny CPU classifiers for tone and intent in voice agents 2026-11-03
• GitHub Agentic Workflows adds enclaves to cap what agents leak from private repos 2026-11-03
• Claude's Messages API adds message threads to cut resent history in agent loops 2026-11-03
• Google plans to honor a shared robots.txt token for opting out of AI training 2026-11-03
• Red-teamers found 11 ways a misaligned agent could get past Codex Auto-review 2026-11-03
• Cloudflare explains how D1's control plane runs millions of small databases 2026-11-02
• Decoding Anthropic's Skill authoring guide: lean SKILL.md files that actually trigger 2026-11-01
• IBM-attributed JevAct paper gates agent tool calls against policy before they run 2026-11-01
• Why evicting KV cache tokens can free almost no GPU memory on vLLM 2026-11-01
• Claude Code mods: in-session JavaScript hooks for guards, readouts and review views 2026-10-31
• Tessl's skill-eval study: skills help, but agents often don't load them 2026-10-31
• How Uber's AutoClean agents safely delete stale feature-flag code 2026-10-31
• Meta puts LLM-generated bug-catching tests into production 2026-10-31
• Securing MCP beyond the gateway: InfoQ's four control layers, decoded 2026-10-31
• A ten-person game studio describes running its codebase with a fleet of agents 2026-10-30
• NVIDIA releases Nemotron 3.5 Transcribe, an open promptable speech model 2026-10-30
• Vercel CI enters beta, running only the Turborepo tasks a push affects 2026-10-30
• Google opens a preview of one MCP-based path for apps into Gemini and Search 2026-10-30
• Copilot Automations can now learn from outcomes and propose their own prompt changes 2026-10-30
• Glasser lets AI agents buy Ahrefs, Semrush and Apollo data one API call at a time 2026-10-29
• MiniCPM-OCR pairs a 1.3B parser with a 2B model that flags OCR you shouldn't trust 2026-10-29

Topics

• Agentic AI
• AI Coding
• Coding with AI
• Agents & Automation
• Vendor Watch
• Models & Research
• Models & Inference
• AI in Practice
• Safety
• Safety & Security
• RAG & Knowledge
• Industry & Research
• Multimodal
• Anthropic
• Openai
• X
• Mistral
• Google
• Xai
• Deepseek
• Nvidia
• Fine-tuning
• Huggingface
• Perplexity
• Stability
• Cohere
• Openclaw
• Meta