← AM·PM Brief

1100 AI Pros Urge Slowdown; Google Debuts 24/7 AI Agent Gemini Spark

· Evening brief · 10 news · 8:41

Audio in Mandarin Chinese · English transcript below

US AI official urges govt to slow AI dev; Google launches 24/7 Agent; $1B security acquisition accelerates threat response.

The LLM industry is moving from showmanship to substance. Andrew Ng is using AI to reinvent one-on-one education, while DeepSeek-R1 outperforms ChatGPT-4o in thyroid nodule differentiation—taken together, these examples underscore that deep vertical scenario adaptation is the key. Fireworks Nexus cuts coding costs through open-weight routing. Notably, Cyera’s $1 billion acquisition of Oasis Security to strengthen AI Agent protection, coupled with Anthropic’s aggressive tactics drawing backlash in Silicon Valley, signals that security and trust are becoming moats far more formidable than raw parameters.

Today's Top 3 Headlines

  1. AI Regulation

    ⚖️ 1100 AI workers urge government to slow down AI development

    Over 1,100 AI workers signed an open letter urging US support for international governance tools, controlling AI automation development pace, and focusing on recursive self-improvement risks. For developers and investors, this signals a shift from light-touch regulation to active development pace restrictions.

    Source
  2. Others

    🤖 Gemini Spark: Your 24/7 Personal AI Agent

    Google launches Gemini Spark, a 24/7 personal AI Agent for AI Pro users, automating complex tasks. For users, this means focusing on higher-value work, unlocking productivity.

    Source
  3. AI Industry News

    🤖 Cyera acquires Oasis Security for $1B, boosting AI Agent protection

    Cyera agrees to acquire Oasis Security for $1B to enhance protection for surging AI Agents. This marks Cyera's third major acquisition this year, highlighting rapid consolidation in AI security to address new challenges from AI adoption.

    Source

+7 more headlines

  • 🤖 US bans imports of foreign advanced robotics equipment to protect AI safety
  • 🤖 DeepSeek-R1 hits 87.9% sensitivity on thyroid nodules, topping ChatGPT-4o
  • 🤖 Fireworks AI launches Nexus: plug-and-play routing cost control layer, migrating code to open-source models
  • 🤖 Andrew Ng founds LearnVector to create personalized 1-on-1 learning with AI
  • 🤖 Silicon Valley's growing resentment toward Anthropic: Aggressive tactics and safety controversies
  • 🤖 Google AI Overview coverage surges from 15% to 43%
  • 🤖 Open-Weight AI: Innovation Engine or Ethical Challenge?
Unlock all 10 headlines + deep analysis →Free 3-day trial · cancel anytime
Browse all past briefings →