ChatGPT Pulse proactively works while you sleep

PLUS: Meta introduces Vibes AI video feed & OpenAI compares AI to human workers. Google DeepMind debuts laundry-sorting robot models, Spotify strengthens AI music protections.

In today’s agenda:

1️⃣ OpenAI rolls out ChatGPT Pulse personalized daily briefings for Pro users with Gmail integration

2️⃣ Meta launches Vibes feed enabling users to create and share AI-generated videos within Meta AI app

3️⃣ OpenAI introduces GDPval benchmark comparing GPT-5, Claude, Gemini against human experts across 44 occupations

  • Google DeepMind unveils Gemini Robotics dual-model system for strategic planning and task execution

  • Spotify announces new AI protection policies including impersonation rules and spam filtering systems

MAIN AI UPDATES / 26th September 2025

🔔 ChatGPT Pulse proactively works while you sleep 🔔
Personalized daily briefings transform ChatGPT from reactive to proactive assistant.

ChatGPT Pulse proactively works while you sleep through OpenAI's new briefing feature, rolling out in preview for Pro users on iOS and Android. The feature integrates with user memory, past conversations, and optional Gmail/Google Calendar connections to provide morning agendas, reminders, and travel notes through swipeable cards. Pulse performs background research each night to prepare customized content, with filters to block sensitive topics. The feature will expand to Plus users before reaching all users, representing OpenAI's move toward proactive AI assistance that anticipates your needs rather than waiting for commands.

🎬 Meta introduces Vibes AI video feed 🎬
Meta pushes into AI-generated content creation with dedicated social sharing platform.

Meta introduces Vibes AI video feed, a new feature within the Meta AI app that enables users to create and share short AI-generated videos. Currently in early preview for select users, Vibes allows creation from scratch, remixing of existing clips, and addition of visuals, music, and styles. Completed videos can be shared in the dedicated Vibes feed, sent directly to friends, or posted to Instagram and Facebook Stories and Reels. This represents Meta's strategic push into AI-generated content creation and social sharing, positioning the company to compete in the growing AI video generation market.

📊 OpenAI compares AI to human workers 📊
Comprehensive benchmark tests AI models against professional experts across multiple industries.

OpenAI compares AI to human workers across 44 occupations has introduced GDPval, a comprehensive benchmark testing AI models against professional-level work. The evaluation compared GPT-5, Claude Opus 4.1, Gemini 2.5, and Grok 4 against industry experts with an average of 14 years of experience across nine fields including healthcare and finance. The benchmark covered 1,320 real-world tasks, with Claude Opus 4.1 achieving the highest performance at 47.6% win rate, particularly excelling in visual presentations. This research provides crucial insights into AI's current capabilities relative to human professional performance across diverse industries.

INTERESTING TO KNOW

🤖 Google DeepMind debuts laundry-sorting robot models 🤖

Google DeepMind has launched two complementary models in its Gemini Robotics family to advance AI-powered robotics. Gemini Robotics-ER 1.5 functions as the strategic planner, while Gemini Robotics 1.5 executes these plans, converting instructions into precise motor commands for physical task completion. This dual-model architecture separates high-level planning from low-level execution, potentially enabling more sophisticated robotic applications.

🎵 Spotify strengthens AI music protections 🎵

Spotify has announced comprehensive new AI protection policies including strengthened impersonation rules that require artist authorization for vocal clones and a new music spam filter launching this fall. The platform removed 75 million spammy tracks in the past year and will implement industry-standard AI disclosure credits through DDEX partnerships with major distributors. These measures aim to protect legitimate artists from AI-generated content farms while preserving creative freedom for responsible AI usage.

📩 Have questions or feedback? Just reply to this email , we’d love to hear from you!

🔗 Stay connected: