- What's Up in AI
- Posts
- ChatGPT Might Order with it's Own System Soon, DeepMind Launches VaultGemma & xAI’s Grok 4 Fast 4 real
ChatGPT Might Order with it's Own System Soon, DeepMind Launches VaultGemma & xAI’s Grok 4 Fast 4 real
And everything else you need to catch up on that happened this week yet!
What's Up in AI
September 15, 2025
In partnership with
Welcome Back, AI Geeks!
It’s time to stay updated with What’s Up in AI.
OpenAI is turning ChatGPT into a smart ordering system with one-click checkout and parental controls. DeepMind’s VaultGemma brings serious privacy to AI without sacrificing power. xAI’s Grok 4 Fast skips the fluff for lightning-quick answers.
In today’s issue:
🤯 ChatGPT Builds it’s Own Ordering System
🧠 DeepMind Launches VaultGemma
📉 xAI’s New Model Choose Speed over Depth
⚡ 5-4-3-2-1 hits you just can't miss
😂 Your AI-Powered Laugh
Read time: 5 mins
Hear from leaders at Anthropic, Rocket Money, and more at Pioneer
Pioneer is a summit for the brightest minds in AI customer service to connect, learn, and inspire one another, exploring the latest opportunities and challenges transforming service with AI Agents.
Hear directly from leaders at Anthropic, [solidcore], Rocket Money, and more about how their teams customize, test, and continuously improve Fin across every channel. You’ll take away proven best practices and practical playbooks you can put into action immediately.
See how today’s service leaders are cultivating smarter support systems, and why the future of customer service will never be the same.
LATEST IN AI TODAY📎
🔒 VaultGemma is the world's most capable differentially private LLM.
— Google AI Developers (@googleaidevs)
4:06 PM • Sep 12, 2025
OpenAI is developing an Orders Hub to store credit cards, enable one-click checkout, and track purchases across devices.
Voice Mode will be integrated directly into the main app for smoother multitasking and faster workflows.
Parental controls are being added, hinting at future classroom and kid-friendly use cases.
Microsoft’s revenue cut from OpenAI is set to drop from ~20% to 8% by 2030, freeing up $50B+ for training powerful models.
OpenAI is making big moves to own more of its ecosystem, from payments to compute, while shaping the industry’s future like it did during the dot-com era.
VaultGemma is a 1-billion-parameter model trained with differential privacy, meaning it won’t leak sensitive data.
It follows “DP Scaling Laws” to stay accurate while keeping privacy intact.
Performs just as well as non-private models of the same size, making it the first serious privacy-first LLM.
Weights are open for researchers on Hugging Face and Kaggle.
Demis Hassabis says real AGI is still 5-10 years away, and the next big skill won’t be coding, but “learning how to learn.”
Grok 4 Fast beta is built to answer questions up to 10x faster than Grok 4 by skipping deep reasoning steps.
It focuses on simple fact lookups, straightforward code generation, and quick integrations.
Less useful for creative writing or nuanced analysis, but perfect for fast, no-frills tasks.
Could soon replace Grok 3 in the free-tier version.
Signals a shift toward faster, more iterative updates and user transparency with a changelog now available.
BRAIN BOOST 🧠
If speed beats depth, where do you think this tradeoff will matter most? |
QUICK HITS 💥
It’s time for some quick hits that you’ll love. Let’s go in…..4-3-2-1 👇️
5- TOP AI TOOLS OF THE DAY
🌐 ForumScout
Monitor brand mentions across forums, social platforms, and news
🧠 ChatPlayground AI
Compare 40+ AI models side by side for best fit
🗣️ Angel CX
Give websites a human-like voice for 24/7 engagement
🎨 eBookColoring
Turn text prompts into printable coloring pages and ebooks
💘 CrushCheck
Analyze texts to see if your crush is interested
4-AI JOB OPPORTUNITIES
Want us to search for tech related jobs in your country/city ? Post a request 🔍
3-THINGS IN TREND
A user had to trick an airline’s AI chatbot just to reach customer service.
Higgsfield AI has unveiled Ads 2.0, a new tool it says can replace entire production, marketing and creative teams with a Mini App.
One of the first ways attackers exploit AI models is by hijacking prompts. Here are 5 examples.
2-USEFUL LINKS
1-TOP ROBO NEWS
China’s Haiqin ROV just completed its first mission, diving more than 13,500 feet in the South China Sea. It successfully collected biological samples and sediments, boosting China’s push to explore one of Earth’s final frontiers.
WHAT ELSE HAPPENED IN AI 📰
📰 Penske Sues Google
Rolling Stone’s parent Penske Media sued Google over AI Overviews, claiming illegal use of its content and traffic loss.
🍏 Apple AI Exec Departs
Robby Walker, senior AI exec at Apple who worked on Siri and AI search, has left the company.
✂️ xAI Lays Off Tutors
xAI is cutting 500 AI tutors from its data annotation team training Grok AI.
📈 Google Gemini Tops iOS
Gemini overtook ChatGPT as the top U.S. iOS app after Nano Banana image model went viral.
🇨🇳 Tencent Poaches OpenAI Talent
Tencent hired Yao Shunyu, a leading OpenAI researcher, to integrate AI into its services.
Funding 💰️
Alibaba Cloud leads US$140 million funding round for embodied AI start-up X Square Robot
GET AN AI-POWERED LAUGH 😁
Apparently even AI needs rest
Acquire new customers and drive revenue by partnering with us
What’s Up in AI is one of the most engaged & loved newsletter for professionals, AI enthusiasts & business owners with ~45,000+ working at the world’s leading startups and enterprises. Companies like Hubspot, Artisan, Guidde & Gamma etc. feature their products in What’s Up in AI. You can learn more about partnering with us here.
And that’s a wrap!
Do drop a comment to tell us👇️
What do you think of today's issue? |
We’ll be back soon with more spicy takes on What’s Happening in AI, so stay tuned & share our newsletter with a friend using this link!
Stay Tuned! 🤖
Team What’s Up in AI