💻 Claude Opus Codes For Days

Hey folks,

Meta's testing an AI-powered morning brief that pulls from your Facebook social network to create personalized updates called Project Luna. Unlike ChatGPT's generic responses or Google's search-based answers, this taps into your actual social network, the posts your connections share, topics your network discusses and trending conversations in your circles. The strategic bet is social context being the differentiator that ChatGPT can't replicate. While OpenAI and Google build general-purpose AI assistants, Meta's positioning this as the AI that actually knows what matters to you based on who you know and what they're talking about. If it works, Meta's trying to make AI the interface to your social life, not just another chatbot.

Let's dive in..

🏆 Gemini 3 Topped Benchmarks, Then A Memo Leaked

🐝 The Buzz: Gemini 3 Pro has been topping the Leaderboards, outperforming GPT-5.1 across major benchmarks. Then two days later, Sam Altman sent an internal memo warning OpenAI staff to brace for rough vibes and economic headwinds as Google surges ahead. When your competitor's CEO is publicly acknowledging they're winning, you need to play catchup.

• Alphabet stock surged 5% toward $4T valuation: The market responded immediately. Alphabet hit $315.90 per share as investors repriced Google's AI position, pushing market cap to $3.82 trillion. If they cross $4 trillion, they become only the fourth company ever to reach that milestone. Salesforce CEO Marc Benioff called this "the most profound shift since ChatGPT's original debut".

• OpenAI's "Shallotpeat" project reveals deeper problems: Internal sources confirm OpenAI is working on a model codenamed "Shallotpeat", literally a reference to fixing "bad soil" designed to address fundamental bugs in their pre-training process. This is fixing the foundation model instead of looking ahead towards a new model. That takes time that Google is using to compound their lead.

💡 Takeaway: The competitive landscape is open. Google's not catching up anymore and they're ahead on all benchmarks. Anthropic's ahead on coding. OpenAI's rebuilding foundations. If you're making long-term platform bets, test all three on your use cases this week. The "just use GPT" era is over.

Together with Lindy

The Simplest Way to Create and Launch AI Agents and Apps

You know that AI can help you automate your work, but you just don't know how to get started.

With Lindy, you can build AI agents and apps in minutes simply by describing what you want in plain English.

→ "Create a booking platform for my business."
→ "Automate my sales outreach."
→ "Create a weekly summary about each employee's performance and send it as an email."

From inbound lead qualification to AI-powered customer support and full-blown apps, Lindy has hundreds of agents that are ready to work for you 24/7/365.

Stop doing repetitive tasks manually. Let Lindy automate workflows, save time, and grow your business

Get $20 Worth of Free Credits

💻 Claude Opus 4.5 Launched And Codes For Days

🐝 The Buzz: Anthropic dropped Claude Opus 4.5 and the immediate value proposition is simple, multi-day software development projects now finish in hours. The model beati GPT-5.1-Codex-Max (77.9%), Gemini 3 Pro (76.2%), and even Anthropic's own Sonnet 4.5 (77.2%). But the real story is what you can actually build now.

• Infinite conversations limits: Claude now automatically summarizes earlier context as conversations grow, so you never hit the wall mid-project. No more "your context window is full", the conversation just keeps going. For developers working on complex refactors or debugging sessions that span hours, this changes the workflow.

• Usage caps are gone entirely: More critically, Anthropic removed Opus-specific usage caps. The flagship model is now accessible for production workloads and continuous use, not rationed for special occasions. If you've been saving Opus for "important" tasks because of rate limits, that constraint just disappeared.

• It handles ambiguity like a senior engineer: Internal testing at Anthropic showed Opus 4.5 scored higher on engineering take-home exams than any human applicant in company history. Employee testers consistently reported improved judgment and intuition across tasks, the model developing a sense of what actually matters in real-world contexts, not just following instructions literally.

💡 Takeaway: This is removing the constraints that made flagship AI impractical for continuous work. Context limits? Gone. Usage caps? Removed. Price barrier? Cut by two-thirds. If you've been rationing your use of high-capability models or manually restarting sessions when you hit limits, test Opus 4.5 on your actual workflow.

📰 Like newsletters? Here are more our readers enjoy.

💻️ How to Turn Your LinkedIn Profile Into An Infographic

• Go to LinkedIn

• Download your profile as a PDF

• Upload the PDF as a source in NotebookLM

• Click the Infographic pencil 

• Set customization detail as country and western theme

• Generate and Download

🐝 AI buzz bits

 🎨 Nano Banana Pro finally solves text-in-images, and it's free. Every previous model struggled with typography, Midjourney produced nonsense letters, DALL-E gets wonky. Nano Banana Pro generates professional posters and mockups with correct text, connects to Google Search for fact-checking, and outputs at 4K resolution.

⚡ Meta's applying for federal approval to trade electricity because AI data centers need that much power. Meta filed to trade power on wholesale markets, joining Microsoft in seeking authorization that Apple already has. The play: make long-term commitments to new power plants while maintaining resale flexibility to manage risk.

📈 Google's AI infrastructure boss told employees they must double capacity every 6 months, 1000x growth in 4-5 years. That's handling user requests, not just training infrastructure. This is what exponential AI adoption looks like on the infrastructure side.

🔥 AI productivity tools

For a full list of 1000+ AI tools, visit our Supertool Directory

🤘 Wed Vibes