GPT-5: Flop or Flex? 🤖, $10B Cloud Frenemies 💰, AI’s Carbon Cost 🌍

AI Quick Takes ⚡

1. AI’s Carbon Receipt 🌍⚡

After months of doom-scrolling headlines about AI melting the planet, Google finally dropped details on Gemini’s energy diet. Turns out each text prompt sips 0.24 watt-hours, about the juice it takes to watch 9 seconds of TV or zap leftovers in a microwave for one second. Read more

The fine print:

• 0.03 g CO2 (1/150th of a phone charge)

• 0.26 ml of water (5 drops)

• Efficiency improved 33x in the past year

Why it matters: Individual prompts aren’t killing the planet, but scale is the silent killer. Billions of queries add up, and Google left out the heavy stuff (like model training, image/video generation, or Deep Research). Think of it like only counting gas mileage while ignoring the car factory. The real climate bill is still TBD.

2. Anthropic Eyes a $10B Boost 💰

AI startup Anthropic is in late-stage talks to raise up to $10 billion, double its earlier $5 billion target, thanks to surging investor demand. This mega-deal would place Anthropic among the most heavily funded AI companies to date. If finalized, this funding marks a leap from its March Series E round, which raised $3.5 billion at a $61.5 billion valuation. Read more

Bottom Line: Anthropic is positioning itself to scale aggressively, building compute infrastructure, fueling R&D, and battling titans like OpenAI and xAI for dominance in generative AI.

3. Meta + Google: When Frenemies Share the Cloud ☁️ 

Meta just signed a $10 billion, six-year deal with Google Cloud, yes, that Google, to handle its growing AI and platform demands. Rivalry? What rivalry? Meta’s infrastructure needs are so massive that even building their own data centers ($72 billion in capex) wasn’t cutting it. So instead of duking it out solo, they’re leaning into Google’s cloud muscle. Think of it as Uber calling Lyft for a ride. Read more

Why it matters: For Google Cloud, snagging Meta as a client brings serious cred in the battle with AWS and Azure. Competitors are becoming collaborators behind the scenes, and the real winner is whichever cloud can power the AI boom without running out of GPUs.

4. Gemini’s Taking Over Your Smart Home Speaker 🗣️ 🔈️ 

Google is replacing its classic voice assistant with Gemini for Home, a next-gen AI rolling out to Nest speakers and displays this October. It’s smarter, chattier, and finally understands things like “turn off the lights… but not in my room.” Read more

Why it matters: Gemini brings the vibe of a real convo, no more repeating “Hey Google” every five seconds. You can interrupt it, change your mind mid-command, and get responses that make sense in context. Plus, Google teased a sleek new speaker to match (yes, it glows). This is Google’s big leap toward making your smart home feel less like a robot butler and more like a helpful roommate. If Alexa had a glow-up and took improv classes? That’s the vibe.

Try This Week 💡

1. Notion AI 📝

Your built-in power user just leveled up. Notion AI can now run multi-step tasks across your workspace. Think editing multiple pages in one go or updating an entire database without the click-fest. It’s like having your most organized teammate permanently on call.

✨ Example workflows:

• Revise tone across 20 meeting notes in one swoop

• Add deadlines to every project card in a database

• Standardize formatting across multiple team docs

How to try it: Highlight the pages or database you want to update, then prompt Notion AI (look for the sparkle icon).

2. OpenAI launched QuizGPT 💡

OpenAI just dropped QuizGPT inside ChatGPT.

📎 Try At Work: Turn onboarding into a game.
New hire training that doesn’t feel like reading a manual from 2003? Yes, please. Ask it to “use QuizGPT,” feed it your onboarding content, and boom: you’ve got a clickable, trackable quiz for your team. HR will thank you. So will your interns.

📚 Try At Home: Study buddy = upgraded.
Whether your kid has a spelling test or you're brushing up on your Spanish before that Barcelona trip, QuizGPT makes studying actually stick. Think flashcards that respond, score, and don’t end up in a drawer. Bonus: it’s free with ChatGPT.

You Asked 🙋‍♀️

 💬 “Was ChatGPT-5 a massive success or a massive failure? I genuinely can’t tell?”

Short answer: Yes.

Here’s the split-screen reality:

What users felt: Disappointed. People missed the personality of GPT-4o (“charming” is out, “robotic” is in), noticed tighter rate limits, and didn’t experience the promised AGI. Reddit threads were basically group therapy.

What OpenAI built: A powerful model with drastically lower inference costs. GPT-5 is not one giant model, it’s a multitasking hybrid that routes simple tasks to mini-models and saves heavy-lifting for complex queries. That means: faster responses, lower costs, and less strain on GPUs. This is critical for both OpenAI's bottom line and for the wider adoption of AI in enterprise applications.

Behind the scenes win: It’s a massive success in efficiency. Enterprise users are loving it for better coding, fewer hallucinations, and serious AI agent potential.

So… success or failure? Depends on your POV. If you wanted a charming AI bestie, it flopped. If you're OpenAI trying to run a global-scale AI without breaking the internet, or the bank, it crushed.

Final take: GPT-5 didn’t come to dazzle, it came to scale. Think less red carpet, more server room. Glamorous? No. Groundbreaking? Absolutely.

Your Turn 🔁

Got an AI win? Found a tool that made your life 10x easier? Building something cool that deserves a little hype (or know a woman who is)? We want to hear it! You might just see your story in the next edition (with your permission, of course). Because when one of us figures something out, we all get smarter. 🫶 

Resource Roundup 🛠️

🎓 AI Automation Playbook (a.k.a. How ElevenLabs Saved $140K and Slept Better)

What happens when a scrappy growth team ditches overpriced SaaS tools and builds custom AI workflows instead? Magic. Or in this case: $140K saved, faster ops, and zero maintenance headaches.

In this must-watch How I AI episode, ElevenLabs’ Head of Growth Luke Harries shares how they:

• Turned 3-min customer convos into polished case studies and tweet threads using Granola + a custom GPT

• Built a DIY translation system (in a day!) to replace $100K+ in agency spend

• Hooked up WhatsApp to Claude MCP for automated social content and trend analysis

Steal this mindset:
→ Everything is a launch
→ If it’s mostly human coordination, it’s ripe for AI disruption
→ Build over buy when you need better control, brand voice, or speed

TL;DR: This is the playbook for turning marketing chaos into AI-powered calm. Worth a watch, a notepad, and a Monday workflow sprint.

Want more? Quick Links 🔗 

1. xAI exposed over 370,000 Grok chatbot convos via public-share links

2. FieldAI raised $405M at a $2B valuation to develop “robot brains”

3. Seven in ten Americans fear AI will permanently put people out of work

4. Google is rolling out a new update to Docs that generates an audio version of your documents using AI and reads them out loud

What We’re Loving 💖

If you’ve ever Googled “best note app” at 1 am, ToolFinder.co is your new fave.

Why we’re into it:

• 🛠️ 1,000+ curated tools: from AI note-takers to ADHD-friendly planners, all sorted and reviewed.

• 🧠 Real people, real workflows: Tool stacks that show how creators and founders use their tools.

• 📚 Smart guides: Quick reads to help you choose the best app without needing a spreadsheet.

Created by productivity pros, ToolFinder is like your super-organized friend who always has a rec and never sends you down a rabbit hole.

👉 Check it out and find your next fave tool in minutes, not hours.

Her AI Vault 🗂️

Your go-to stash of top resources we’ve created so you never have to dig through old emails to find that one gem.

📘 5 Ways To Get Started With AI At Work: Quick-start guide to delegate the “ugh” tasks and focus on what actually matters.

🔐 AI Privacy Guide: What your AI tools really collect and how to lock down your settings like a pro.

🔍 Her AI Glossary: 40+ essential AI terms, broken down in plain English.

🧠 AI Model Cheat Sheet: Claude, ChatGPT, Gemini... we decoded who does what best, so you can stop guessing and start impressing.

🛠️ Build-Your-Own Productivity Tool: A one-prompt web app for tasks, timers, and focus that is custom-built by Claude and copy-pasted by you.

Before You Go... 💭