Welcome, AI Builders!
A single deal now ties NVIDIA’s silicon to OpenAI’s future at a scale that looks more like nation-building than partnership. Meanwhile OpenAI is poaching Apple’s design talent to craft its own devices, Musk’s xAI is driving reasoning costs to near zero, and Google is turning televisions into AI companions. The AI race is spilling into every screen, chip, and circuit.
📌 In today’s Generative AI Newsletter:
NVIDIA and OpenAI plan $100B datacenter buildout
OpenAI hires Apple veterans for hardware push
xAI launches Grok 4 Fast with 98% cost savings
Google brings Gemini AI to living room TVs
Special highlight from our network
Microservices in minutes, not months
Introducing ServiceBricks, the open-source tool that uses AI to accelerate backend development:
Full-stack in seconds: Generate APIs, databases, tests, and integrations, ready for production.
Flexible architecture: Works with SQL Server, PostgreSQL, Cosmos DB, Azure Tables, and more.
Built for developers: Skip the boilerplate and focus on what matters—innovation.
💥 NVIDIA Bets $100B on OpenAI’s Compute Future
Image Credit: OpenAI
NVIDIA and OpenAI have signed a landmark agreement to build 10 gigawatts of AI datacenters, powered by millions of NVIDIA GPUs. The deal would make NVIDIA not just a supplier, but OpenAI’s strategic compute partner, with up to $100B in investment tied to each stage of deployment.
What’s in play:
Scale of ambition: 10 GW is enough capacity to power millions of homes, redirected instead into training and running OpenAI’s next generation of models.
First phase: A gigawatt of NVIDIA systems will go live in 2026 on the company’s new Vera Rubin platform.
The money: NVIDIA plans to invest up to $100B as the buildout advances, cementing itself at the core of OpenAI’s infrastructure.
Shifting alliances: The deal comes as OpenAI diversifies beyond Microsoft, adding NVIDIA to a lineup of partners that includes Oracle, SoftBank, and Stargate.
The figures read like something out of industrial history. A single supplier underwriting 10 gigawatts of compute and $100B of capital is not a partnership, it is empire-building. NVIDIA is positioning itself as the backbone of AI infrastructure, and every breakthrough that follows will run on its silicon.
📱 OpenAI Raids Apple Talent for AI Hardware Push
Image Credit: The Information
OpenAI is building a hardware team stocked with former Apple talent, recruiting interface designers, audio engineers, and manufacturing specialists with packages north of $1M, according to The Information. The effort is led by ex-Apple executive Tang Tan, who is pitching candidates on fewer bureaucratic hurdles and a bold product roadmap.
What’s in motion:
Talent grab: Dozens of Apple veterans have already signed on.
Manufacturing deals: Partnerships with iPhone suppliers Luxshare and Goertek are in place for early production.
Device concepts: A display-less smart speaker is the first candidate, alongside glasses, a pin wearable, and a voice recorder.
Timeline: The inaugural release is targeted for late 2026 or early 2027.
Jony Ive’s design influence looms large, with both hiring strategy and supplier ties echoing Apple’s playbook. What remains unclear is whether OpenAI is building a platform to rival the smartphone, or a set of niche devices that carry the weight of its AI models into the physical world. Either way, the countdown has begun for one of the most closely watched launches in years.
⚡ xAI Unveils Grok 4 Fast, Cutting Costs by 98%
Image credit: X
xAI has introduced Grok 4 Fast, a hyper-efficient reasoning model that matches near-frontier performance while slashing compute requirements. The model uses 40 percent fewer reasoning tokens than Grok 4 and delivers a 98 percent reduction in cost, opening advanced reasoning to all users on grok.com.
Highlights of the release:
Benchmark strength: Scored 85.7% on GPQA Diamond and 92% on AIME 2025, outperforming Claude 4.1 Opus and Gemini 2.5 Pro.
Arena leader: Rose to No. 1 in LMArena’s Search Arena and beat Grok 4 on select coding benchmarks.
Unified model: Combines reasoning and non-reasoning into one architecture, reducing latency and token use.
Extended context: Supports 2M tokens with native tool integration for browsing and code execution.
For developers, Grok 4 Fast means smoother handling of both quick prompts and complex reasoning within a single system. For users, it means top-tier performance at a price point approaching zero. The phrase “intelligence too cheap to meter” has floated through AI circles for years. Grok 4 Fast is one of the first models that makes it sound less like prophecy and more like a bill coming due.
📺 Google Brings Gemini AI to the Television
Image Credit: Google
Google has begun rolling out Gemini-powered assistants to Google TV, starting with TCL’s QM9K and expanding to more sets later this year. The update turns televisions into conversational devices, aiming to make them feel less like screens and more like smart speakers with visuals attached.
What Gemini adds to TV:
Smarter recommendations: Handles vague prompts like “find a hospital drama everyone’s talking about,” returning relevant shows instead of just search lists.
Group awareness: Can weigh multiple viewing preferences at once, surfacing titles a group might agree on.
Beyond entertainment: Responds to general knowledge questions, pulling up both answers and related YouTube videos.
Future scope: Google says shopping, travel content, and purchasing options tied to programs are on the roadmap.
The push shows how Google wants AI assistants to extend beyond phones and speakers into the living room. A TV that listens, responds, and suggests may sound convenient, though it also raises the question whether the centerpiece of home entertainment is becoming one more AI sales channel in disguise.
Believe in AI? ✨ Now there’s extra time to own a piece of it.
The bonus-share deadline has been extended: invest by Oct 17, 2025 and secure up to 25% more shares in GenAI.Works.
GenAI.works has been building the foundation that lets people learn, experiment, and shape what AI can actually do — with solutions in education, orchestration, and distribution. All open, global, and community-led.
It’s already working:
13M+ community
40M+ monthly impressions
200+ companies on board
$3M ARR, bootstrapped
Now turning that base into something bigger:
a GTM AI platform that understands what’s happening in the market and knows exactly when to act.
This is your opportunity to help shape the infrastructure powering the next platform shift.
👉 Don’t just believe in AI. Help build it.
Reg CF offering via DealMaker Securities. Investing is risky.
🚀 Boost your business with us. Advertise where 13M+ AI leaders engage!
🌟 Sign up for the first AI Hub in the world.
