The $20K Humanoid Robot That Can’t Fold Your Laundry (Yet)

What if I told you the future of household robots has arrived… but it still needs a human babysitter?

Because that’s exactly what just happened.

You see, California-based 1X Technologies opened pre-orders for NEO, the world’s first “consumer-ready” (even if I doubted it ) humanoid robot for the home. Price tag? $20,000. Promise? Folding laundry, loading dishwashers, fetching items, and helping around the house.

But, the reality? It’s not fully autonomous. Actually, even a slightly complex task is controlled by a remote human operator watching the inside of your home through the robot’s cameras.

Yes. Your “AI robot” currently works because another human is driving it. Welcome to the uncanny middle ground between science fiction and outsourced labor.

A Launch With Impeccable Timing and Awkward Contrast

The timing couldn’t be more ironic.

Just one day earlier, DJI launched its ROMO robot vacuum in Europe : a fully autonomous cleaning robot powered by drone-derived sensors. Price? Around $1,500.

So on one side, you have DJI shipping real autonomy at a fraction of the price.

On the other, you have a $20,000 humanoid that still needs a remote human named “Turing” to fold your sweater.

This isn’t a coincidence. It’s a fork in the road for robotics. If we go back to our prediction about AI robotics in 2025, you’ll see the problem is right there.

The Honest Demo Nobody Expected

When The Wall Street Journal journalist Joanna Stern tested NEO, the results were… refreshingly brutal.

NEO couldn’t complete a single task autonomously.

Not one.

Fetching a water bottle from the fridge?
Over one minute with human teleoperation.

Loading three dishes into a dishwasher?
Five minutes.

Folding one sweater?
Two minutes of careful manipulation, again remotely controlled.

Movements were described as slow. Awkward. Sometimes erratic.
Less “Jetsons future.”
More “prototype learning to exist.”

And here’s the key part: 1X didn’t deny it.

“If We Don’t Have Your Data, We Can’t Make the Product Better”

That’s what 1X CEO Bernt Børnich told the WSJ.

Every NEO buyer expecting delivery in 2026 must accept one thing up front:
human teleoperators will control the robot while watching inside your home.

The company calls them “Turing” operators.
Owners schedule access via an app.
Privacy measures include face blurring, no-go zones, and time windows.

But the tradeoff is clear:
your robot learns by watching humans do your chores through it.

Creepy? For some.
Honest? Absolutely.

The Hardware Is Not the Joke

Here’s where the story gets interesting.

Because while the autonomy isn’t there yet, the engineering absolutely is.

NEO stands 5’6” (167 cm) tall and weighs just 66 pounds (29.9 kg), yet it can lift 154 pounds (69.8 kg) and carry objects up to 55 pounds (25 kg).

That power-to-weight ratio would have sounded absurd five years ago.

The robot uses 1X’s patented Tendon Drive system, relying on ultra-high torque-density motors and tendon-based transmissions to create soft, compliant, human-safe motion.

It’s also quiet, just 22 decibels, softer than a refrigerator.

Its hands feature 22 degrees of freedom, enabling near human-level dexterity.
The body uses custom 3D lattice polymers for softness and safety.
Sensors include microphones, speakers, and dual fisheye cameras.

Connectivity? Wi-Fi. Bluetooth. 5G.
Software updates? Over the air.

This is not a toy.
This is serious robotics.

Redwood AI: The Brain Behind the Puppet Strings

NEO runs on Redwood AI, 1X’s proprietary model a 160-million-parameter transformer trained on real-world teleoperation data which combines vision, touch, and body movement and it also includes a built-in large language model for conversation and memory.

That last part matters.

Because memory means learning over time.
Not scripted demos.
Not canned behaviors.
But adaptation.

And adaptation requires something robots are still starving for: data.

The Pricing Reality Check

NEO pricing is bold. Almost confrontational.

• $20,000 outright purchase
• Or $499/month, six-month minimum
• $200 refundable deposit
• Target delivery: 2026

By comparison:

• Premium robot vacuums top out around $1,400
• DJI’s ROMO launches between $1,300–$2,000
• And they already work autonomously

So why would anyone pay 10× more for something worse?

Because NEO isn’t competing with robot vacuums.

It’s competing with the idea of general-purpose physical intelligence.

Follow the Money (It Knows Something)

1X isn’t building in a vacuum.

The company raised:

• $23.5M Series A2 led by OpenAI in 2023
• $100M Series B in 2024
• Strategic partnership with Nvidia

During an onstage appearance in 2024, Nvidia CEO Jensen Huang even accepted a custom leather jacket… handed to him by NEO.

That wasn’t theater.
That was signaling.

---

Humanoids Everywhere — Autonomy Nowhere (Yet)

NEO is far from alone.

Figure AI recently unveiled Figure 03 a home-focused humanoid designed for domestic tasks.
Mass production target? 100,000 units in four years.

Its architecture mirrors a growing industry standard:

• A slow vision-language model (7–9 Hz) for understanding
• A fast control transformer (~200 Hz) for movement

Sound familiar? Nvidia’s Isaac GR00T uses something similar.

And then there’s Tesla.

Optimus has appeared on factory floors, in Tesla showrooms, and next to celebrities. Elon Musk wants to ship one million humanoids in a decade.

He believes they’ll be bigger than cars. Bigger than phones. Bigger than anything.

That sounds unhinged until you look at the incentives.

Silicon Valley Is All In

According to Morgan Stanley, humanoid and service robots could become a $300–$400 billion market within a decade.

Apple is reportedly exploring the space.
Foxconn is deploying humanoids at Nvidia factories.
Investment across humanoid startups surpassed ¥1 trillion (~$6.7B).

And analysts believe only 80 million humanoids will exist in homes by 2050.

Not because demand is low.
But because the problem is brutally hard.

Why Your Home Is Harder Than a Factory

You see : Factories are controlled environments. Warehouses are predictable. Roads, surprisingly, are simpler than kitchens. But homes are chaos.

Infinite object variations.
Unpredictable layouts.
Fragile items.
Soft materials.
Tasks requiring fine manipulation and context.

This is why roboticists roll their eyes at humanoid hype.
Wheels are easier than legs.
Grippers are easier than hands.
Structured beats unstructured every time.

And yet, humans keep coming back to humanoids.

Because psychologically, we’re wired for them.

This Has Happened Before (And People Forgot)

If NEO feels familiar, it should.

Early Tesla Autopilot was mocked.
Critics called it vaporware.
It required constant supervision.

But every car collected data.
Neural networks improved.
Capabilities compounded.

Now driver-assist systems navigate dense cities that once felt impossible.

The difference?
Homes are exponentially harder than roads.

Which means autonomy will take longer : but the curve will be steeper.

The Exponential Part Everyone Misses

The first NEO takes five minutes to load a dishwasher.
The hundredth learns from all previous attempts.
The thousandth has seen more kitchens than any human.
The ten-thousandth makes autonomy realistic.

This is exactly how AI racing drones evolved — from constant crashes to beating human champions.

Data wins.
Iteration wins.
Transparency wins.

Teleoperation Isn’t a Bug but the Business Model

People keep asking:
“Why is 1X hiring robot operators?” Wrong question.

They’re not hiring operators. They’re hiring AI trainers.

This is how autonomy is built. The same way drone pilots became fleet supervisors and autonomous systems always mature.

The jobs are temporary, but the impact is permanent.

Privacy Is the Price of Progress (For Now)

Yes, having a human watch your home is uncomfortable.
Yes, the concerns are valid.

But 1X isn’t pretending otherwise. Scheduled access. No-go zones. Face blurring.User control.

More importantly, this phase is transitional. Early adopters aren’t buying a finished product. They’re funding the training set.

So… Is This Science Fiction or Reality?

It’s neither. It’s the messy middle.

NEO won’t autonomously run your household in 2026. That part is marketing optimism. And unlike most robotics demos, 1X is actually showing the ugly truth : It’s a half-finished prototype with multiple aesthetic choices, and deep down a robot learning how to exist in your home.

Final Thought

Every transformative technology starts this way : Expensive. Imperfect. A little embarrassing:

The iPhone wasn’t great in 2007.
Autopilot wasn’t safe in 2015.
Drones weren’t reliable in 2010.

But the people who bought them early made the future possible.

Looking forward

NEO isn’t the future of household robotics. It’s merely the beginning of building it. AGI won’t be there in the near term, no matter how people hyped it.

And that awkward teleoperators and all are far more interesting than another robot vacuum that already knows how to clean.

If you enjoy learning and entrepreneurship, you can follow me on my YouTube channel or my newsletter, I’m here to practice my entrepreneurship muscle every week! Stay tuned, and see you in the next one!