So far this week, we’ve talked about why I believe humanoid robotics could become one of the defining investment opportunities of the next decade.
On Monday, we looked at why AI is moving from the digital world into the physical world.
On Tuesday, we explored the “picks and shovels” companies that will supply the entire robotics ecosystem.
Yesterday, we discussed why labor shortages and aging populations make this more than just another exciting technology story.
Today, I want to look at a question that could determine where the biggest fortunes are made.
Who controls the software?
History suggests this is where investors should pay close attention.
Think about the smartphone revolution.
Apple built one of the greatest businesses in history by combining hardware and software into a seamless ecosystem. But Google’s Android operating system now powers billions of smartphones made by dozens of manufacturers around the world.
Google didn’t have to manufacture every phone to become one of the biggest winners.
The same thing happened with personal computers.
Dell, HP, Compaq, Gateway, and dozens of others competed to build PCs.
Microsoft supplied the operating system.
As more computers were sold, Microsoft benefited regardless of which manufacturer won market share.
I believe humanoid robotics could follow a very similar path.
Every robot needs a brain.
But that brain is much more than a semiconductor.
It needs an operating system.
It needs artificial intelligence capable of understanding language, recognizing objects, navigating unfamiliar environments, making decisions, learning from experience, and interacting safely with people.
It needs software. And software tends to scale much faster than hardware.
One area I believe Wall Street is dramatically underestimating is simulation software.
Before a humanoid robot ever takes its first step in the real world, it will likely spend thousands - or even millions - of hours training in a virtual one.
Think of it like a flight simulator.
Pilots don’t learn to fly a $100 million jet by climbing into the cockpit on day one. They spend countless hours practicing in highly realistic digital environments where mistakes are free and every scenario can be repeated over and over again.
Humanoid robots will follow the same path.
Before they’re trusted to work in a warehouse, assist a nurse, or help build a car, they’ll learn how to walk, grasp objects, navigate obstacles, and interact with people inside sophisticated digital simulations.
That creates a massive opportunity for companies building the software that makes these virtual worlds possible.
Nvidia is already leading with its Omniverse and Isaac robotics platforms, while companies like Unity, Autodesk, PTC, Dassault Systèmes, and Siemens provide many of the digital engineering, simulation, and “digital twin” technologies manufacturers rely on today.
The more I study this industry, the more convinced I become that the winners won’t just be the companies building the robots. Some of the biggest beneficiaries could be the companies building the virtual worlds where those robots learn to become smarter before they ever enter the real one.
But simulation is only one piece of the software opportunity.
The operating system itself could prove even more valuable.
Tesla has developed its own AI and software stack for Optimus.
Meta is investing heavily in artificial intelligence and recently expanded its robotics ambitions through acquisitions designed to strengthen its humanoid capabilities.
Nvidia isn’t simply selling chips anymore. Through its robotics software, simulation tools, and AI development platforms, the company is helping developers train and deploy intelligent machines long before they ever enter the real world.
Alphabet is taking a similar approach through DeepMind and its robotics initiatives, combining world-class AI research with technologies designed to bring intelligence into the physical world.
This reminds me of the early days of cloud computing.
Most investors focused on the companies renting server space.
The real winners became the companies building the platforms everyone else relied upon.
Humanoid robotics could evolve the same way.
There may eventually be dozens - even hundreds - of companies manufacturing robots.
But they may all rely on a much smaller group of companies providing the AI models, operating systems, simulation software, and development tools that power those machines.
That’s an incredibly attractive business.
Every new robot shipped could become another customer.
Every software update could make millions of robots smarter overnight.
Every improvement could strengthen the ecosystem and make switching to a competing platform even more difficult.
Those are exactly the kinds of businesses long-term investors love to own.
Of course, it’s still far too early to know who the ultimate winners will be.
Technology revolutions rarely follow a straight line.
Some companies that look like leaders today won’t exist a decade from now.
Others receiving very little attention today may become tomorrow’s household names.
That’s why I believe diversification across the entire humanoid ecosystem makes far more sense than betting everything on a single robot manufacturer.
Tomorrow, we’ll wrap up this week’s series by looking at the bigger picture.
Humanoid robotics isn’t simply another exciting technology trend.
I believe it has the potential to become the next great economic revolution - one that could reshape nearly every industry on the planet while creating investment opportunities far beyond the companies building the robots themselves.
Here’s to the future,
Matt McCall
Editor, Market Insights
