Good evening! 🫡 🔮

What a WILD week in AI! NVIDIA just dropped some absolute bombshells at GTC 2025, and a startup is literally growing living neurons to replace GPUs. The future is here, and it's both exciting and slightly terrifying!

This is NVIDIA’s new "GR00T-N1" robot that stole the show:

Let's dive into what's cooking in the AI world this week ↓

Today we're diving into🤿

🤖 NVIDIA's GR00T-N1: The Humanoid Robot Revolution at GTC 2025
🧠 NVIDIA's Open-Source Reasoning Models Take On DeepSeek
🧫 GPUs Go Biological: Living Neurons Power AI Computing
📰 AI News Roundup - Everything You Need to Know
🛠️ Tools of the Week

So grab your favorite drink, and let's jump in! ↓

_{PRESENTED BY SALESFORGE:}

🔮 Samurai Spotlight - Salesforge

Run Your Sales on Autopilot

Increase the output of your sales team without buying more tools or hiring new SDRs. Onboard Agent Frank, Salesforge’s AI SDR, to fully automate prospecting, personalized outreach and booking meetings while your team focuses on closing deals. Get a personalised email from Agent Frank to see his work in action:

Show me the magic

🤖 NVIDIA's Isaac GR00T-N1: Powering the Humanoid Robot Revolution at GTC 2025

"What an amazing year it was, and we have a lot of incredible things to talk about." - Jensen Huang, Nvidia CEO

At GTC 2025, NVIDIA unveiled Isaac GR00T-N1, the world's first open foundation model for humanoid robots, alongside their groundbreaking Blackwell architecture and "Blue" robot demonstration. CEO Jensen Huang, in his iconic leather jacket without a teleprompter, described the event as the "Super Bowl of AI" and positioned these technologies as solutions to global labor shortages estimated at over 50 million workers by decade's end.

Details: 

GR00T-N1 features a dual-system architecture inspired by human cognition. "System 1" is a fast-thinking action model that mirrors human reflexes, while "System 2" is a slow-thinking model for deliberate decision-making. Powered by a vision language model, System 2 reasons about the environment and instructions to plan actions, which System 1 then translates into precise robot movements.

The model was trained on a pyramid of data: internet-scale web data and human videos at the base, synthetic data generated through NVIDIA Omniverse in the middle, and real robot data at the peak. Using the NVIDIA Isaac GR00T Blueprint, over 750,000 synthetic trajectories were generated in just 11 hours – equivalent to nine continuous months of human demonstration data.

In performance testing, GR00T-N1 significantly outperformed baseline models, achieving a 76.8% success rate on real-world tasks with full training data compared to 46.4% for the baseline Diffusion Policy. Even with just 10% of the training data, GR00T-N1 reached 42.6% success.

GTC 2025 featured several other major announcements:

• The Blackwell architecture delivering a staggering 40x AI performance over Hopper

• New open-source 'Dynamo' framework for AI inference optimization

• "Blue," a Star Wars-inspired robot developed with Google DeepMind and Disney Research

• Newton, an open-source physics engine for robotics simulation

• A detailed roadmap for next-gen Vera Rubin chips through 2027

Leading humanoid developers with early access to GR00T-N1 include Agility Robotics, Boston Dynamics, Mentee Robotics, and NEURA Robotics. In his keynote, Huang demonstrated 1X's humanoid robot autonomously performing domestic tidying tasks using GR00T-N1.

What does this mean? 

NVIDIA is transforming from a GPU manufacturer to an end-to-end AI infrastructure company. The GR00T-N1 model and robotics initiatives represent NVIDIA's expansion beyond traditional AI computing into the physical world.

These technologies could revolutionize industries by handling repetitive or demanding tasks in human workspaces. GR00T-N1 makes it much easier to develop useful humanoid robots by providing a foundation that can be adapted to various robots, tasks, and environments.

The timing is critical for NVIDIA, coming after Chinese startup DeepSeek's R1 model sent their stock tumbling earlier this year. Rather than positioning efficient models as a threat, Huang framed them as driving increased demand for computation—effectively turning a potential weakness into a strength.

As Huang made clear, "The age of generalist robotics is here," and NVIDIA is positioning itself at the center of this revolution.

_{PRESENTED BY ARTISAN:}

🔮 Samurai Spotlight - Artisan

10x Your Outbound With Our AI BDR

Scaling fast but need more support? Our AI BDR Ava enables you to grow your team without increasing headcount.

Ava operates within the Artisan platform, which consolidates every tool you need for outbound:

• 300M+ High-Quality B2B Prospects, including E-Commerce and Local Business Leads

• Automated Lead Enrichment With 10+ Data Sources

• Full Email Deliverability Management

• Multi-Channel Outreach Across Email & LinkedIn

• Human-Level Personalization

Book a demo and supercharge your sales team.

🧠 Nvidia's Open-Source Reasoning Models Take On DeepSeek

In a direct competitive move against DeepSeek R1, Nvidia has unveiled its Llama Nemotron family of open reasoning models specifically designed for agentic AI workloads.

This is a calculated strategic response to the booming market for reasoning-capable AI systems that has emerged in 2025, and particularly to DeepSeek's R1 model that shook Nvidia's stock price earlier this year.

The Llama Nemotron family comes in three sizes:
• Nano: 8B parameters, distilled from Llama 3.1 8B for edge devices
• Super: 49B parameters, distilled from Llama 3.3 70B for data centers
• Ultra: 253B parameters, distilled from Llama 3.1 405B (coming soon)

What makes this big?:

• Outperform DeepSeek R1 on both accuracy and speed 📊
• A unique "reasoning on/off" capability through system prompts 🔄
• Models built on open Llama models with NVIDIA-vetted datasets 🔓
• They're releasing nearly 30 million samples of high-quality training data 📚

"Agents are autonomous software systems designed to reason, plan, act and critique their work," said Kari Briski, VP of Generative AI Software Product Management at Nvidia. "Just like humans, agents need to understand context to breakdown complex requests."

This move positions Nvidia not just as a hardware provider but as a serious player in the foundation model space. By open-sourcing these models, they're creating an ecosystem that naturally drives demand for their hardware while competing directly with specialized reasoning models.

The question remains: Will these models truly match the reasoning capabilities that made DeepSeek R1 so impressive? And how will this affect the broader AI landscape as reasoning becomes a standard feature rather than a premium add-on?

🧫 GPUs Go Biological: BBB Unveils Bionode - Living Neurons Power AI Computing

In what might be the most sci-fi announcement of the year, Baltimore-founded startup Biological Black Box (BBB) has emerged from stealth with its Bionode platform—a revolutionary computing system that integrates living, lab-grown neurons with traditional processors.

Yes, you read that right. They're growing actual neurons to use as computer chips.

"Over the last 20 years, three independent fields—biology, hardware, and computational tools—have advanced to the point where biological computing is now possible," said Alex Ksendzovsky, BBB's co-founder and CEO.

This isn't just theoretical research—BBB claims their neural chips are already powering computer vision and LLMs for customers, offering a low-power alternative to conventional GPUs.

Here's what makes Bionode so revolutionary:

• Uses lab-grown neurons from rat cells and human stem cells 🧫
• Each chip contains hundreds of thousands of neurons integrated with 4,096 electrodes 🔬
• Neurons physically rewire themselves, creating a fundamentally different computing paradigm 🧠
• Significantly reduces energy consumption while improving processing efficiency ⚡

This is huuge, because:
• More efficient computer vision pre-processing
• Accelerated LLM training with on-the-fly adaptation
• Potential for continuous learning without traditional retraining cycles

Interestingly, BBB is a member of Nvidia's Inception incubator and positions itself as complementary to silicon-based AI chips rather than a direct competitor. "Biological computing and silicon computing will coexist," Ksendzovsky noted.

Could living tissue be the key to AI's next evolutionary leap? Or does this raise too many ethical questions to be viable at scale?

📰 Latest AI News

🧰 Tools of the Week

1. ChatLLM by Abacus

Access to all the biggest LLMs.

Artifacts directly in browser, just like with Claude.

Build complex AI Agents.

Saves me $10 each month compared to other regular LLMs.

2. Roblox Cube 3D

Create 3D objects instantly using simple text prompts.

Transform ideas into fully-textured 3D meshes for games and experiences.

Perfect for developers, designers, and creators of all skill levels.

3. Manus AI

Create stunning AI animations in seconds.

Turn static images into dynamic, lifelike motion.

Perfect for marketers, content creators, and social media.

Thank you for reading!

What's on your mind?

Feel free to share your ideas with me at: _{[email protected]} 

If your idea is good, I’ll make sure to bring it to life. Just send them to the above e-mail and I’ll get to work on making them a reality.