- The AI Entrepreneurs
- Posts
- 🇸🇦 Nvidia's Saudi AI Revolution
🇸🇦 Nvidia's Saudi AI Revolution
PLUS: đź§ OpenAI GPT-4.1 2025: Smarter AI Now

Welcome to AI Entrepreneurs
This edition highlights how AI is reshaping global power, from Nvidia fueling Saudi Arabia’s bold AI ambitions to Google’s AlphaEvolve solving mathematical frontiers. Meta pushes molecular and neural boundaries, OpenAI levels up with GPT-4.1, and Stability brings audio generation to your smartphone. Plus: Canva’s no-code AI builds apps in seconds, and AI-enhanced ECG tools bring clinical clarity. Whether you're scaling cloud infra, decoding genomes, or designing apps—AI is the new engine.

NVIDIA
Nvidia and Humain: A New AI Kingdom Rising in Arabia
Nvidia has struck a major AI chip deal with Saudi Arabia’s Humain, a state-backed AI startup. The agreement will provide 18,000 of Nvidia’s most advanced Blackwell chips to Humain, strengthening U.S.-Saudi tech ties and positioning Saudi Arabia as a key player in global AI infrastructure

Source: Ideogram 3.0/TheAI Entrepreneurs
Key Highlights:
Humain, launched by Saudi Crown Prince Mohammed bin Salman, aims to build AI factories powered by Nvidia’s chips.
The deal gives Saudi Arabia a preferential position in AI partnerships with the U.S., ahead of China.
AMD has also signed a $10 billion agreement with Humain to provide 500 megawatts of AI compute capacity.
The partnership could add $1 trillion to the AI market over the next decade.
The U.S. has ended the AI diffusion rule, allowing more AI chip exports to trusted allies while restricting China’s access to advanced semiconductors.
Why It Matters:
This deal marks a watershed moment in AI development, reinforcing Nvidia’s dominance in the industry.It aligns with the Kingdom’s Vision 2030, paving the way for a new era of global AI leadership.

Google DeepMind’s AlphaEvolve 2025: Gemini AI Boosts Algorithm Design
Google DeepMind has launched AlphaEvolve, a cutting-edge AI coding agent driven by Gemini models, designed to evolve advanced algorithms for mathematics, computing, and real-world applications like data center optimization and AI training.
AlphaEvolve, powered by Gemini Flash and Pro, is redefining AI optimization—boosting Google data center efficiency by 0.7% and speeding up Gemini training by 1%. It enhanced matrix ops like FlashAttention by 32.5%, proposed TPU-level code improvements, and even surpassed Strassen’s method with a new 4x4 matrix multiplication algorithm. In math, it solved 20% of 50+ open problems, including a breakthrough in the 11D kissing number.
Why It Matters:
AlphaEvolve’s ability to design and optimize algorithms across computing, AI, and mathematics positions it as a game-changer for industries like tech, sustainability, and drug discovery. With plans for an Early Access Program, Google DeepMind is paving the way for broader AI-driven innovation in 2025 and beyond.

FROM THE SPONSOR
Learn AI in 5 minutes a day
This is the easiest way for a busy person wanting to learn AI in as little time as possible:
Sign up for The Rundown AI newsletter
They send you 5-minute email updates on the latest AI news and how to use it
You learn how to become 2x more productive by leveraging AI

NVIDIA
NVIDIA x Microsoft at Build 2025: Powering the Next AI Leap
Join NVIDIA at Microsoft Build 2025 (May 19–22) as they showcase how their full-stack AI platform on Azure is transforming industries—from RTX AI PCs to DGX Cloud and Omniverse. Dive into hands-on labs, code workshops, and powerful demos designed to supercharge generative AI, accelerate development, and unleash enterprise innovation.

META
Molecules, Minds & Meta Models: Inside FAIR’s AI Revolution
Meta’s FAIR team just dropped a powerful lineup of open-source tools advancing AI across molecular science, generative modeling, and neuroscience. Highlights include OMol25, a massive atomic-scale chemistry dataset, and the Universal Model for Atoms (UMA) trained on over 30B atoms—set to supercharge innovation in drug discovery and climate solutions. FAIR also introduced Adjoint Sampling, a breakthrough algorithm that trains generative models with no data—just a reward signal.
In a landmark neuroscience study with Rothschild Foundation Hospital, Meta mapped how children’s brains develop language and found striking parallels to LLMs like Llama 3.1. By decoding neural activity from 7,000+ electrodes, the study reveals how speech understanding matures—and how AI may one day learn like the human brain. This is more than research—it’s a glimpse into how AI is decoding intelligence itself.

AI BYTES
Nvidia-Backed AI Cloud Leader Exceeds Forecasts
CoreWeave, a Nvidia-backed AI cloud computing startup, surpassed Wall Street expectations in its first public earnings report, posting $981.6 million in Q1 revenue against a $852.9 million forecast, per LSEG data. The New Jersey-based company, which debuted on Nasdaq in March 2025, saw shares jump 11% in extended trading, reflecting a 68% rise since its IPO, driven by strong demand for AI infrastructure like Nvidia chips and data centers.
Gemini Advanced and GitHub Unite: Next-Level AI Coding Power
Google’s Gemini Advanced now integrates with GitHub, making it a game-changing AI coding assistant! Developers can connect directly to repos, generate and modify functions, debug code, and get instant explanations—all in one seamless workflow. With NotebookLM Plus and Colab in the mix, Gemini Advanced is poised to redefine AI-driven development.
Meet GPT-4.1: OpenAI’s Sharpest, Smartest, Speediest Models Yet
OpenAI’s new GPT-4.1 model family—including mini and nano versions—brings massive gains in coding, instruction-following, and long-context comprehension (up to 1 million tokens). Outperforming GPT-4o across key benchmarks, these models offer faster, cheaper, and more capable AI for real-world tasks—from legal document analysis and software engineering to image understanding and intelligent agents.
Stable Audio Open Small: Generative Audio, Now on Your Smartphone
Stability AI and Arm have teamed up to launch Stable Audio Open Small, a lightweight, open-source text-to-audio model optimized for mobile devices. With just 341 million parameters, this stereo model generates up to 11 seconds of high-quality audio on Arm-powered phones in under 8 seconds—no GPU required.

AI HEALTH
AI Tool Enhances ECG Image Interpretation for Doctors
A team from Technion - Israel Institute of Technology developed an AI tool that interprets ECG images with pixel-level accuracy, even from imperfect or photographed scans, and explains its reasoning in medical terms. This breakthrough improves diagnostic precision and builds trust by providing clear, interpretable insights for cardiac care.
Issue #48 Overview : This issue highlights AI’s healthcare revolution, from redesigning proteins to outsmart patents and predicting post-surgery pain to supporting youth mental health and enhancing ECG scans. Vote for your favorite quirky AI health gadget to shape medicine’s future by 2050!
Subscribe → AIHealthTechInsider.com
Interested in AIHealthTech Insider?Are you interested in receiving the AIHealthTech Insider newsletter directly to your inbox? Stay updated on the latest AI-driven healthcare innovations. |

AI CREATIVITY
How to Build a No-Code App with Canva’s “Code for Me”
Canva AI's “Code for Me” tool lets you turn plain English into fully functional apps—no coding required. Whether you're building a to-do list or a goal tracker, here’s how to do it in minutes:
Step-by-Step Guide
Step 1: Log into Canva (free or paid). On the homepage, search “Canva AI” and select Code for Me.
Step 2: In the input box, explain what you want the app to do. Canva will convert it into HTML, CSS, and JavaScript.
Step 3: Review the generated app. You can tweak the layout, colors, and features using simple follow-up instructions.
Step 4: Try moving elements, adding inputs, or clicking through the app. Update it as needed.
Step 5: Add it to a Canva design or share the live link. No backend setup required.
Pro Tip: Use visual keywords like “bright,” “minimal,” “colorful,” or “professional” in your prompt to guide the app's design style right from the start.


Reply