Are humanoid robots the future… or a $100B mistake? Over 100 companies from Meta to Amazon are betting big on humanoids. But are we chasing a sci-fi dream that’s not practical or profitable?
In this TechFirst episode, I chat with Bren Pierce, robotics OG and CEO of Kinisi Robots.
We cover:
- Why legs might be overhyped
- How LLMs are transforming robots into agents
- The real cost (and complexity) of robotic hands
- Why warehouse robots work best with wheels
- The geopolitical robot arms race between China, the US, and Europe
- Hot takes, historical context, and a glimpse into the next 10 years of AI + robotics
Watch here … and subscribe to my YouTube channel:
Are humanoid robots a really stupid idea?
Over 100 companies are racing to build humanoid robots. Meta is working on one. Amazon is reportedly building one for last-mile delivery. Billions of dollars are flowing into the space, and every month, a new startup joins the fray.
But is this a tulip craze? Are humanoids a sci-fi fantasy … shiny but impractical … or are they the future of robotics?
To dig in, I spoke with Bren Pierce, CEO of Kinsi. He’s a robotics OG: he helped build humanoid robots decades ago (remember Honda’s Asimo?)
His take: legs are cool, but intelligence is cooler. And maybe wheels are all we need right now.
Here are the big takeaways …
Legs are flashy, but wheels just work
If you’re building robots for warehouses and factories, the floor is flat.
Why do you need legs?
Bren argues that legs are mostly about future potential and fundraising optics. Sure, in ten years they might matter. But today? You can do 80% of the useful stuff with a wheeled mobile manipulator, and do it faster, cheaper, and more reliably.
The real breakthrough isn’t mechanical … it’s AI
What’s actually different today isn’t the motors, servos, or arms. It’s AI.
We’re finally able to build robots that can understand the world:
- Read a screen
- Recognize a tomato
- Follow a vocal instruction
- React in real time
“You used to need 10 engineers and 3 months to write all the logic,” Pierce says. “Now you can just show a robot a screen and have it do the job.”
That’s a massive shift. It means lower costs, faster pilots, and the ability to deploy robots in places you couldn’t before, without hardcoding every step.
Hands are the real challenge
Forget legs. The real frontier in robotics is the last 30 centimeters: manipulating things.
“Our hands are 50% of the complexity of a human. We can grip lightly or with force, and they self-repair. Replicating that is insanely hard.”
BTW, Sanctuary AI founder Geordie Rose would agree.
Some robotic hands cost $15,000–30,000 per hand, with over 100 motors and tendons. The better approach? Use purpose-built grippers: vacuum, pincers, tweezers. Better yet, make them hot-swappable, says Pierce.
This gets the job done for most real-world use cases, and lets you customize based on task, not biology.
The problem isn’t speed … it’s latency
Why are robots still slow? Not because the motors can’t move fast.
It’s compute.
“The robot could go as fast as it wants. But if it has to run five AI models and process HD camera feeds in real time, that’s where the latency comes in.”
You need to process vision, understand intent, map a trajectory, and execute, all without hitting a human. That requires massive onboard compute, and today’s robots often use cloud models, which means there’s a delay baked in.
So speed comes down to how much GPU you can strap to your robot. (And that’s expensive.)
Humanoids are the Ferraris of robotics
Sure, we’ll have humanoids. But not soon. And not everywhere, says Pierce.
“If you’re rich, you’ll have a humanoid in your house. But most people will have a Roomba.”
It’s a cost thing. It’s a use-case thing. It’s a complexity thing.
Humanoids make sense for marketing, fundraising, or tasks that truly require a human shape. But for now, we’ll see more warehouse bots, mobile manipulators, and hybrid formats—like four-legged robots that stand on two legs and roll on wheels.
But … this space is evolving fast
Despite his skepticism about humanoid hype, Bren is optimistic:
“The speed of progress is blowing my mind. The tech is there. The bottleneck might be regulators, not engineers.”
Like with autonomous cars or drones, the real friction may come from safety rules, policy, and bureaucracy, not a lack of innovation.
So maybe humanoids aren’t a stupid idea.
They’re just … a future idea. And in the meantime, we’ve got plenty of work to do with robots on wheels, smarter hands, and AI brains.
I know I’ll want one when they’re available!