Today’s menu 🍽️👇️ 

Length: 5 minutes ⏰ 

• MathGPT expands to 50+ U.S. colleges 📚

• Microsoft brings Copilot AI to Samsung TVs 📺

• Anthropic makes browsing safer with Claude 🛡️

• OpenAI co-founder calls for shared standards 🏛️

• Nvidia smashes Q2 with $46.7B revenue 📈

• Turn prompts into clips to fully customized video ✨

• Prompt of the week 💬

• Top AI tools of the week ⚒️

FRESH AI NEWS

MathGPT helps professors fight cheating 🎓

MathGPT, an AI-powered tutor that uses Socratic questioning to prevent cheating and assist professors, has expanded from 30 to over 50 U.S. colleges this fall, increasing its presence in higher education 👩‍🎓

• It encourages critical thinking by guiding students with questions and hints, rather than providing direct answers❓

• They can set when the AI is accessible, limit student attempts, require that students upload work for verification, and generate auto-graded assignments 🎓

• After a successful pilot in 30 colleges, MathGPT is expanding to hundreds of classrooms at institutions like Penn State, Tufts, and Liberty University 🏛️

• It integrates with Canvas, Blackboard, and Brightspace, and includes accessibility features like screen reader support and audio mode 🌐

Why You Should Care: It provides a balanced solution to AI-driven cheating by acting as an AI tutor that supports learning and maintains integrity, reflecting a growing trust in educational tools that uphold academic standards.

Copilot Lands on Samsung TVs 📺

Microsoft's Copilot AI is integrated into Samsung's 2025 smart TVs and monitors, offering an animated AI companion accessible via voice or remote control 📺

• Copilot can be summoned via voice command or a click on the remote from the Tizen OS home screen, Samsung Daily+, or Click to Search 🗣

• Copilot is an animated blob that provides movie recommendations, spoiler-free recaps, and answers questions in real time, displayed on visual cards for big screens 🟠

• The assistant enhances group experiences by selecting shows that match everyone's preferences and offering post-watch trivia, making TV a collaborative space 👪

Why You Should Care: Microsoft and Samsung are embedding a conversational AI companion into TVs to transform family interactions into shared, interactive experiences of entertainment, discovery, and assistance.

Claude turns Chrome into an AI assistant 🖥️

Anthropic has launched Claude for Chrome, a browser-embedded agent in a sidebar that allows users to chat and perform web tasks within Chrome as part of a safety-focused pilot 🌐

• Select Max-plan subscribers ($100–$200/month) can install the extension and interact with Claude in a side panel that maintains browsing context and, with permission, automates tasks like clicking, form-filling, and summarizing 🖥️

• Anthropic reduced prompt-injection risks by lowering attack success rates from ~23.6% to 11.2% and browser-specific attacks from 35.7% to 0%. The agent blocks access to sensitive domains by default and requests confirmation for high-risk actions 🛡️

• The agent is currently limited to 1,000 trusted users on the Max plan, with broader access contingent on safety learnings. There's still an open waitlist for interested users ⏳

Why You Should Care: Claude for Chrome represents a significant advancement in AI, transitioning from standalone chatbots to integrated software agents. This positions Anthropic prominently in the AI-powered browser wars, highlighting the need for safety to progress alongside innovation.

OpenAI co-founder calls for safety teamwork 🤝

OpenAI co-founder Wojciech Zaremba urges AI labs, including OpenAI and Anthropic, to collaboratively safety-test each other's models, emphasizing cross-company collaboration as AI enters a crucial development phase 🤝

• Zaremba disclosed that OpenAI and Anthropic temporarily shared limited API access to each other's models, excluding GPT-5, to identify safety blind spots and promote shared standards 🔍

• He acknowledged that despite fierce competition, collaborative safety testing is essential to maintain trustworthy AI as models scale ⚖️

• Zaremba and Anthropic's safety researcher Nicholas Carlini stressed that safety collaborations should become routine to ensure responsible AI progress 🔄

Why You Should Care: In a competitive AI landscape, this push for transparency and joint safety assessments sets a precedent for safety collaboration alongside innovation and rivalry, paving the way for more trustworthy AI deployment.

Nvidia launches $60B stock buyback plan 💰

Nvidia broke Q2 records with $46.7 billion in revenue due to surging AI demand, but cautious investor reactions highlight concerns about sustaining the hype🔋

• Revenue rose 56% YoY to $46.74 billion, exceeding expectations, with net income up 59% to $26.4 billion. Data center sales reached $41.1 billion 📈 

• Despite stellar numbers, shares fell 2–3% in after-hours trading. Investors were concerned about slightly weaker data center revenue, slowing growth momentum, and geopolitical risks ⚠️ 

• Nvidia projected $54 billion in Q3 revenue and announced a $60 billion stock buyback. CEO Jensen Huang estimated $3–4 trillion in global AI infrastructure spending by decade's end, indicating the boom continues 🚀

Why You Should Care: Nvidia dominates the global AI surge, but investor concerns about growth sustainability, particularly regarding China exposure and data center trends, highlight potential risks.

TURN PROMPT INTO VIDEOS

Turn text prompts into high-quality videos with Google Flow 🎬

Video editing is so slow, and almost every software makes it so unnecessarily complicated. What if you could just enter what you'd like it to come out as and have a ready-to-go video in minutes?

You can do exactly this with Google Flow. Using this guide, I’m going to show you how you can convert simple questions into high quality video recordings with a minimum amount of editing stress.

Here’s how to get started 👇️ 

1. Head to Google Flow and choose the Text to Video option (select Veo 3 for audio support).

   

    

   

2. Write a clear, detailed prompt describing your scene, style, and mood.

   Enter a prompt like: “Minimalist coffee commercial, close-up of hot espresso being poured, soft morning sunlight, steam rising, upbeat background mood”

   

3. Generate your video (up to 8 seconds) and preview the result.

4. Use the timeline editor to polish pacing, add transitions, or layer multiple scenes together.

5. Export your final video in high quality, ready to share or edit further.

   

Pro Tip: Try experimenting with a variety of prompts and different creative variations to generate several high-quality clips. Once you have multiple options, carefully arrange and combine them into a polished, consistent, and complete video ✨

Check out Google Flow HERE 👈️

PROMPT OF THE WEEK

Create AI-Powered learning paths 🎯

Simply paste this prompt:

“Build a 6-month upskilling roadmap for a [insert skill] transitioning into AI roles. Include courses, projects, tools to learn, and milestones. Present as a timeline chart.”

🔧Pro tip: Ask for “recommended free vs paid resources” to make it actionable.

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 🐝 

HeyGen Digital Twin - Create realistic AI avatars 🤖 

Wan2.2-S2V - Open-source speech-to-video AI with audio features 💻

Mirage 2 - Generate real-time, playable world engines from text or images ⚙️

MathGPT - Get instant homework help from your on-demand AI math solver 📚