Today’s menu 🍽️👇️ 

Length: 5 minutes ⏰ 

• OpenAI drops ChatGPT Pulse 🌅

• Meta launches Vibes short videos 📱

• GPT-5 shows human-level skills 🛠️

• DeepMind AI gives robots Google skills 🌐

• Convert photos to Figurines + Motion 🪄

• Prompt of the week 💬

• Top AI tools of the week ⚒️

FRESH AI NEWS

OpenAI unveils ChatGPT Pulse 🗞️

OpenAI just launched ChatGPT Pulse, a new feature that sends personalized morning briefs automatically 🤖

• Pulse generates several short, context-aware updates while you sleep. News, agenda items, reminders 🔍

• It draws from your chat transcripts, memory, calendar, and email (if you opt in) 📝

• Updates come as digestible cards to scan or tap for more detail. Designed to stop once the day’s brief is done ⚡

• Pulse is rolling out now to mobile users on the $200/month Pro plan, with plans to expand 💵

Why You Should Care: Pulse moves ChatGPT from reactive to proactive: no longer just waiting for your questions, it starts your day for you. But with that comes more data sharing, so privacy and control are now front and center.

Meta launches Vibes for quick AI clips 🚀

Meta launched Vibes, a new feed inside its Meta AI app and website, full of AI-generated short videos that users can scroll, remix, and share 🎬

• Start from scratch or remix a video you see. Add visuals, music, styles, then post or cross-share to Instagram/Facebook ✂️

• The feed is exclusively AI-generated content. Meta describes it as a space for creators and communities to generate expressive videos 🤖

• You can publish Vibes clips directly, send via DM, or post to Stories/Reels on Instagram and Facebook 🔄

• Early version draws on Meta’s own models and partnerships with Midjourney and Black Forest Labs for AI visuals ⚙️

Why You Should Care: This is Meta placing a bet on video creation being driven by AI rather than humans. But since audiences are already calling it “AI slop,” the challenge will be whether those visuals ever feel meaningful, or just synthetic noise.

GPT-5 competes with humans 🔍

OpenAI introduced a new benchmark (GDPval) showing that GPT-5 matches or outperforms human professionals in many(job-related) tasks across sectors 🧠

• The test spanned nine major industries, from healthcare to finance, assessing how GPT-5 performs on real deliverables vs experts 🏷️

• In the benchmark, “GPT-5-high” was judged equal or better than experts ~40.6% of the time 📊

• The model’s performance has more than doubled from GPT-4o to GPT-5 in those evaluated tasks 🤝

• OpenAI admits GDPval only covers a subset of tasks. True job performance involves social, contextual, and unpredictable elements that models can’t fully capture yet ⚠️

Why You Should Care: This milestone shows that AI is rapidly closing the gap in knowledge work, but human skills like judgment, ethics, and long-term context still matter. It sets a new bar for how we measure AI’s real-world impact.

DeepMind teaches robots to search 🧠

DeepMind’s new models let robots browse the internet mid-task, using search results to inform decisions in real time 🤖

• Gemini Robotics-ER 1.5 handles reasoning and web tool access, while Gemini Robotics 1.5 handles vision, action, and execution 🧠

• Robots can query the web (for local rules, weather, recycling protocols, etc.) and use that information to guide physical tasks 🌐

• Capabilities now include sorting laundry, packing bags based on forecasts, and following local waste rules 🔄

• Skills learned on one robot body (arms, humanoid, etc.) can transfer to others thanks to shared underlying models 🔁

Why You Should Care: By combining physical capability with internet knowledge, robots move closer to true autonomy. This is a key leap toward machines that can adapt, plan, and act in complex real-world settings rather than just follow fixed commands.

ANIMATE PHOTOS IN MINUTES

Turn photos into Figurines + Animations with AI ✅

What if you could turn a simple photo into a collectible figurine, and then bring it to life in video form? With Google Gemini’s Nano Banana tool and Luma Dream Machine, you can design hyper-real figurine concepts and animate them for content that grabs attention on social media, ads, or even product launches ✅

Here’s how to get started 👇️ 

1. Head to Gemini, hit “Create images”, and enable the Nano Banana tool.

2. Experiment with Prompts like: “Make a mini product figurine of this sneaker, placed on a branded stand inside a collector’s box.”

   

3. Upload Your Reference. Provide a product or selfie photo, then let Gemini render your figurine in its scene.
   

4. Animate in Luma Dream Machine. Open Luma Labs, create a New Board, and try prompts like: “The figurine jumps off its base, waves, then lands back on the stand.”

   

Pro Tip: Don’t just describe the figurine. Describe the cinematic mood. Adding directions like soft studio lighting, toy commercial style, or TikTok unboxing vibe makes your animations more scroll-stopping ✨

PROMPT OF THE WEEK

AI-driven branding kit 🧠 

Simply paste this prompt:

“Act as a brand strategist. Design a branding kit for an AI startup, including tagline, elevator pitch, color palette, and tone of voice. Provide 3 variations: professional, playful, and futuristic.”

🔧 Pro Tip: Ask for “logo and visual style suggestions for Canva/MidJourney.”

The best place on the Internet to learn about AI! 🌐 🤖 

I just started my own X page where I provide AI news, guides, hacks and more (for FREE)!

We don’t miss a day, that means you get the most recent AI news, practical workflows, expert guides and premium content everyday.

TOOLS OF THE WEEK

These AI tools are going VIRAL 🐝 

Suno v5 - SOTA music generation model 🎶

Kling 2.5 Turbo - New high-quality video generation model 🎥

Mixboard - Google’s experimental canvas for visualizing ideas 🌎

Sembly - AI Notetaker with a “brain” that unlocks meeting intelligence 🧠