Androids don’t dream of electric sheep, they dream of themselves

Welcome to AI Collision 💥,

In today’s collision between AI and our world:

• Thor, Isaac and GR00T
• Computex Keynote 2025
• $20 trillion or bust

If that’s enough to get the androids dreaming, read on…

AI Collision 💥

When Jensen Huange (Nvidia CEO) took the stage at Computex in Taiwan last year it was “peak Nvidia” he was even signing his name on a fan’s chest… like a true rockstar…

This year, not chest signing (that we’re aware of) but his visit to deliver the Computex keynote has now been labelled “Jensanity” due to the hysteria he’s creating in arguably the world’s third most important AI country (behind America and China).

That’s largely in part because Jensen Huang takes the stage, you don’t just get a product launch or major announcement, you get a glimpse of Nvidia’s future. At Computex 2025, that future came as a humanoid robot and massive silicon robot brain.

The big deal (so far) is the release of Jetson Thor, NVIDIA’s new system-on-chip built for humanoid robots and autonomous vehicles. It’s very much the “brains” of Nvidia’s real-world AI ambitions, however, if the headlines are all about Thor, they’re not telling you the real story.

The real story is GR00T.

And if history is any guide, what GR00T could become for robotics is exactly what CUDA became for AI, the platform that turns hardware dominance into an “AI infrastructure” empire.

But let’s start with the Thor. Jetson Thor is a beast of a chip built on NVIDIA’s Blackwell architecture, capable delivering up to 2070 FP4/1035 FP8 TFLOPs 128GB memory. In English? That’s enough juice to power real-time language understanding, motion planning, and sensor activity.

That also translates to all the cognitive load a humanoid robot or a self-driving car needs to survive and adapt in the real world, in real-time.

Think, dodging a dog that springs out from behind a parked car, or grabbing a falling cup from a shelf while opening a cupboard, Jetson Thor is designed to do it on the fly. And it’s optimised to work in tandem with NVIDIA’s high-bandwidth memory and GPU cores.

This then works in conjunction with NVidia’s Isaac GR00T platform.

As Huang said in his keynote,

“Physical AI and robotics will bring about the next industrial revolution,”

“From AI brains for robots to simulated worlds to practice in or AI supercomputers for training foundation models, NVIDIA provides building blocks for every stage of the robotics development journey.”

But here’s where it gets interesting. GR00T isn’t just a product. It’s a platform. And that should make investors take notice, for one VERY big reason…

Back in November 2006, Nvidia made an announcement that should be looked upon in history as one of the biggest ever.

It was the release of their CUDA platform for their GPUs.

NVIDIA Corporation (Nasdaq: NVDA), the worldwide leader in graphics processors, today unveiled NVIDIA CUDA technology, a fundamentally new architecture for computing on NVIDIA graphics processing units (GPUs), and the industry’s first C-compiler development environment for the GPU.

GPU computing with CUDA is a new approach to computing where hundreds of on-chip processor cores simultaneously communicate and cooperate to solve complex computing problems up to 100 times faster than traditional approaches.

Before CUDA, NVIDIA GPUs were just graphics cards. After CUDA, they became the engine room of modern AI.

CUDA gave developers a programming model, a toolkit, and a reason to stay loyal to NVIDIA. Once you built your AI workflows around it, switching to AMD or Intel meant rewriting your codebase, retraining your models, and walking into uncertainty.

I believe that GR00T is to robotics and AI real-world systems what CUDA was to GPUs and AI.

The platform enables access to data from real-world training to human simulation but more importantly synthetic data or “GR00T-Dreams” as Nvidia describes it.

So, there you have it, Androids dream of themselves in simulations doing tasks in the multiverse, not electric sheep…

If NVIDIA wins the race to the brain and body of humanoid robots and autonomous machines, it won’t just sell more chips, it will become the default operating system for the physical robotic and AI world. That’s not an exaggeration.

Very quickly the idea of Nvidia becoming a $10 trillion, $20 trillion company in the next decade isn’t so wild. Or maybe it’s as wild as thinking in 2006 they could become a $3 trillion company.

But that’s where I think we are with their development of Thor and GR00T.

GR00T is CUDA for robotic and AI reality.

By the way, to watch his full keynote at Computex, check it out below.

#AD

New AI 2.0 Will Open a Brief 

“Wealth Window” 

…for regular investors to get in on the $15,700,000,000,000 AI Boom 
 
Re: Your Second Chance – If you missed out on AI 1.0, AI 2.0 could be 10X bigger. 
 
Early movers could change their fortunes with Altucher’s three AI wealth-building strategies  

Capital at risk 

Boomers & Busters 💰

AI and AI-related stocks moving and shaking up the markets this week. (All performance data below over the rolling week).

Boom 📈

• Dell Technologies (NASDAQ:DELL) up 19%
• Arm Holdings (NASDAQ:ARM) up 17%
• Tesla (NASDAQ:TSLA) up 17%

Bust 📉

• Predictive Oncology (NASDAQ:POAI) down 2%
• Brainchip (ASX:BRN) down 2%
• Alibaba (NYSE:BABA) down 1%

From the hive mind 🧠

• Anduril is serious business. And I’ll have more on this I think soon. We’ve covered them a bit over the last year or so, and I think they’re going to be the biggest most valuable defence company in history whenever they IPO (likely soon). But you can check out how serious they are with this video.
• I’m aware of Australian private equity companies (and the Australian government) investing in and backing PSIQuantum due to the fact it was started by Aussies. But now it’s said that Nvidia is investing too. Remember it wasn’t all that long ago, comments from Huang tanked the quantum computing market. And now Nvidia is investing directly…says a lot about what’s coming with quantum I think.
• Elton John weighing in on AI wasn’t on my “things I expect to see this week” list. But… here it is.

Artificial Polltelligence 🗳️

Last week we took a look at Trump’s big Middle Eastern tour to raise money to invest in American AI, and I asked what you thought it would do to the AI market…

And then we considered the idea of drug development, not in space, but on earth with the assistance and development of an AI-enabled FDA.

And while Mondays are usually reserved to look back on the polls for last week, I do want to run another poll today…

Weirdest AI image of the day

From Mark to Gecko

ChatGPT’s random quote of the day

Thanks for reading, and don’t forget to leave comments and questions below,

Sam Volkering

Editor-in-Chief

AI Collision