- NATURAL 20
- Posts
- Gemini 3: A Creative AI Revolution
Gemini 3: A Creative AI Revolution
PLUS: Perplexity Launches Comet AI Browser on Android, Gemini App Adds Image Verification to Spot AI-Generated Photos and more.
Wes Roth
November 21, 2025
In partnership with
WhatsApp Business Calls, Now in Synthflow
Billions of customers already use WhatsApp to reach businesses they trust. But here’s the gap: 65% still prefer voice for urgent issues, while 40% of calls go unanswered — costing $100–$200 in lost revenue each time. That’s trust and revenue walking out the door.
With Synthflow, Voice AI Agents can now answer WhatsApp calls directly, combining support, booking, routing, and follow-ups in one conversation.
It’s not just answering calls — it’s protecting revenue and trust where your customers already are.
One channel, zero missed calls.
Today:
Gemini 3: A Creative AI Revolution
Google DeepMind Launches Nano Banana Pro with Multilingual Text & Smart Editing
OpenAI Launches Free ChatGPT for U.S. K–12 Educators
Perplexity Launches Comet AI Browser on Android
Gemini App Adds Image Verification to Spot AI-Generated Photos
the world wasn't ready for Gemini 3
Gemini 3 is capturing the imagination of developers and creatives with its stunning ability to build games, interactive visuals, and even scientific simulations with minimal input. Users are “vibe coding” wild experiments like 3D Doom with vegetables, Pac-Man on tiny planets, destructible voxel cities, and real-time black hole visualizations.
It's even beaten pro humans at GeoGuessr and built real-time sign language interpreters. One user recreated his old game Theme Park in hours. Gemini 3 isn’t just smart—it’s inspiring, emotional, and deeply weird. And perhaps most mind-bending of all: it’s making us question what reality even is.
Nano Banana Pro is a major upgrade to Google DeepMind’s image generator. It now creates more accurate visuals with readable text, deep knowledge, and smart editing tools. Users can make infographics, posters, or mockups in multiple languages, adjust lighting, blend up to 14 images, and verify AI origins with SynthID.
KEY POINTS
Smarter Visuals: Generates detailed, realistic images using world knowledge and better reasoning from Gemini 3 Pro.
Text in Images: Renders legible, stylized text in multiple languages, ideal for infographics, branding, and translation.
Creative Control: Lets users edit lighting, focus, layout, and characters—perfect for designers, educators, and marketers.
Why it matters
Nano Banana Pro gives everyday users and professionals powerful tools to turn ideas into clear, beautiful images. It blends design, facts, and storytelling in a way that was once only possible with expert software. Plus, built-in watermarking helps people know what’s made by AI.
ChatGPT for Teachers is a free, private version of ChatGPT designed for U.S. K–12 educators, helping them save time, personalize teaching, and collaborate with peers. It offers GPT‑5.1, admin tools, and school-grade data protections—available at no cost through June 2027.
KEY POINTS
Built for Educators: Offers tailored lesson planning, student-safe data handling, and tool integrations (Google Drive, Canva, etc.).
Free Through 2027: Verified U.S. K–12 teachers can use it at no cost, with advanced features like GPT-5.1 and file uploads.
Admin & Policy Support: District-wide controls, SSO security, and AI literacy resources support safe, scalable adoption.
Why it matters
Teachers are under constant time pressure. This tool gives them trusted AI support to plan lessons, adapt materials, and stay creative without risking student data. By offering it free, OpenAI makes powerful AI more accessible to public education, letting teachers lead the way.
Perplexity launches its AI-powered Comet browser on Android, bringing voice-based search, tab summaries, and built-in assistance to mobile users. Though some desktop features like sync aren’t ready yet, key upgrades like agentic voice mode and password support are coming soon.
KEY POINTS
Mobile AI Browser: Comet now available on Android, featuring Perplexity’s built-in assistant for real-time web interaction.
Voice Mode & Summaries: Users can ask questions aloud and get instant summaries of tab content.
More Features Incoming: Bookmark sync, full voice agents, and a password manager are in development.
Why it matters
AI-native browsers like Comet are reshaping how we explore the web. With mobile access and voice integration, Perplexity gives users a hands-free way to learn, research, and interact with content hinting at a future where browsing becomes a dynamic conversation.
🧠RESEARCH
Kandinsky 5.0 is a new open-source model family for creating high-quality images and 10-second videos. It includes lightweight and powerful versions, trained with advanced techniques to improve speed and quality. The release includes code and checkpoints to help researchers build and experiment with fast, high-resolution visual generation tools.
VR-Bench is a new benchmark that tests how well video models can solve mazes by understanding space and planning steps. The study finds that video models can reason better than other AI systems and improve even more with varied testing. It shows strong promise for visual reasoning through video.
AraLingBench is a benchmark that tests how well AI language models understand Arabic. Using 150 expert-written questions, it checks grammar, spelling, reading, and sentence structure. The results show that while models often perform well on the surface, they struggle with deeper understanding—relying more on pattern matching than real language comprehension.
🛠️TOP TOOLS
Each listing includes a hands-on tutorial so you can get started right away, whether you’re a beginner or a pro.
AI2sql – AI-Powered SQL Query - AI assistant that converts natural‑language prompts into working SQL queries
AICarousels – Free Carousel Maker and Generator - browser-based tool that helps you generate and design swipeable social media carousels
AIChatting: Quick Guide to the Web & Mobile AI Chatbot - GPT‑powered chatbot you can use to ask questions and generate text.
📲SOCIAL MEDIA
🗞️MORE NEWS
Google is adding image verification to the Gemini app, letting users check if a photo was made or edited with Google AI. This helps people spot AI content more easily and trust what they see.
GPT-5 is helping scientists work faster. In early experiments, it’s been used to find hidden immune mechanisms, finish math proofs, and link research across fields. It’s not perfect, but paired with experts, it accelerates discovery.
Trump’s draft executive order would block states from making their own AI laws. It directs lawsuits against states like California and Colorado, threatens to cut federal funding, and pushes for a single national AI policy favoring tech companies.
ChatGPT now supports group chats globally. Up to 20 users can collaborate in shared conversations across all plans. The feature transforms ChatGPT from a solo assistant into a social, team-friendly space for planning, writing, and decision-making.
Google DeepMind has hired ex-Boston Dynamics CTO Aaron Saunders to lead robotics hardware. The goal: turn Gemini into the “Android” of robots—an AI platform that works with all robot types, from humanoids to machines with specialized forms.
ChatGPT Atlas now features Arc-style vertical tabs and lets users set Google as the default search engine. The update enhances tab management, adds drag-and-drop, and improves extension imports—moving the AI browser closer to a full productivity platform.
OLMo 3 is the first fully open “thinking” AI model, exposing its step-by-step logic and full training process. Built by the Allen Institute for AI, it’s efficient, transparent, and rivals larger models in reasoning, coding, and comprehension.
Google’s Gemini AI is now available in Android Auto, replacing Google Assistant in cars. Drivers can use voice commands to message, navigate, check calendars, or just chat—all while keeping hands on the wheel. iOS users excluded.
What'd you think of today's edition? |
Reply