The 1x NEO Humanoid. Available Now. $20,000. or $500/month

[Alan Timm]

https://www.youtube.com/watch?v=LTYMWadOW7c

https://www.1x.tech/neo

https://www.1x.tech/order

Ok.  Maybe not quite now.  Early 2026.

But I don't see anything about limitations or capabilities.  Did they really crack utility?

They're talking a big talk about scheduling chores, and working the the neo team for corrective training.  But are we really here yet?

[Alan Timm]

Some more information on their AI Model and approach:

https://www.1x.tech/discover/redwood-ai-world-model

https://www.1x.tech/1x-world-model.pdf

[Subhobroto Sinha]

Great share Alan! Although I'm responding to you, I'm curious what and when we think about bots like these will actually be able to do chores around the house like raking and blowing leaves, cleaning/mopping (vacuuming bots are great but at some point, some mopping for stains and sanitation becomes required I think)?

After that, I think the next useful step would be home perimeter security. It's becoming a bit inconvenient, I think in certain neighborhoods where people are breaking in and stealing brazenly in broad daylight. I'm not going to engage in politics or pointificate on growing wealth gaps - none of those are good reasons for burglary in my mind. These bots can speak a short sentence in 20 languages and ask if the people are lost and need help. They will also warn that hitting the bot or going past it authorizes the owner or law enforcement to open fire without any further clarification. 

[Alan Timm]

Well, according to 1x they'll be capable of doing real work around the house by the time they're shipped early next year.

There's so many questions.  Like what happens when they fall down?  Who is responsible for the repairs? (to the robot, and even to whatever they broke on the way down)

If I'm renting/leasing it and it falls and breaks its arm on the way down, who covers it?

Aside from functionality which is what most people are focusing on right now, what about liability?

I've seen all the creative ways the Unitree robots flail and fall when things don't go their way.  They're fast and their dangerous.

Anywhoo, I can't wait to find out!

Alan

[Subhobroto Sinha]

> I've seen all the creative ways the Unitree robots flail and fall when things don't go their way.

That's extremely interesting! Would you, please, be able to share a few videos that show this behavior? Or point me to some phrases to use? There seem to be certain channels that focus on putting together carefully crafted, super short montages of these robots failing, which to me, as a person who loves building them, does them a disservice. Longer length set of clips, that show what led the robot to behave that specific way, providing me the context, would be way more useful.

> Well, according to 1x they'll be capable of doing real work around the house by the time they're shipped early next year.

This could be a real gamechanger. Robots have always been available but at this price, it's now also accessible to most people. As you pointed out, it's very important for us to have agency on the repairs but also, as you pointed out, the liability is an issue. It would also be interesting to see if home insurance companies react to the presence of these robots at a home.

There are fantastic times we live in - there's so much to look forward to the future. Thank you for sharing!

--
You received this message because you are subscribed to the Google Groups "RSSC-List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rssc-list+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/rssc-list/0dc38a63-de99-4b48-afd1-327b47851878n%40googlegroups.com.

[Thomas Messerschmidt]

I just can’t imagine awakening to that face inside my house every morning. 

Thomas Messerschmidt

-  

Need something prototyped, built or coded? I’ve been building prototypes for companies for 15 years. I am now incorporating generative AI into products.

Contact me directly or through LinkedIn:   

https://www.linkedin.com/in/ai-robotics/

On Oct 28, 2025, at 8:38 PM, Alan Timm <gest...@gmail.com> wrote:

Well, according to 1x they'll be capable of doing real work around the house by the time they're shipped early next year.

--

[Alan Timm]

Heh I think I'm more ok with NEO's face than whatever Unitree put on their H2.  (shudder)

The WSJ did a piece on NEO yesterday that gives alot of visibility into how they're planning on making this all work.

https://www.youtube.com/watch?v=f3c4mQty_so

From NotebookLM:

"A Wall Street Journal video about the Neo humanoid home robot by 1X, available for pre-order at $20,000 with a 2026 delivery date. The robot features a lightweight design inspired by biology, using strong, light motors and tendon-like structures to move smoothly and quietly, though its strength and dexterity are comparable to a human's. Currently, Neo's actions are tele-operated by a human "pilot," which generates training data for the AI neural network that is intended to make future models autonomous. This learning process requires early adopters to accept a "social contract" involving privacy trade-offs, as 1X requires data from the home environment to improve the product, despite implementing safety and privacy measures like blurring people and enforcing no-go zones. Ultimately, the current version is likened to raising a toddler, signaling the beginning of physical AI in the home rather than delivering an immediately perfect, super-useful device."

[Thomas Messerschmidt]

Very interesting video! Thanks for posting that Alan. Yet another video of a CEO promising what his robots WILL (someday) do. (My mind painfully wanders back to past broken promises about what robots WILL DO, from Jibo to Topo.)

Yes, this one is puppeted, like every other household humanoid we’ve ever seen. The home  humanoid robot industry still has a long way to go. My guess is that we are still a couple of decades away from a robot like Andrew in Bicentennial Man. We’re moving forward, but not as fast as we are lead to believe. 

So basically, if these are being puppeted, you will need to hire some person to work, what eight hours a day, in your home  as a Maid/Butler? Is that included in the monthly fee? Or would they be tacking on an extra $30-40 an hour for a trainer/ cleaning person? 

And, the CEO said it’s “not perfect, but incredibly useful?” Useful to whom I would ask. I would assume that would mean useful to the robot company. As the robot drops and breaks one glass after another every day, I would say it becomes less and less useful to the consumer, with or without a remote robot driver.

Still, the physical robot itself is quite impressive. Someday, when the software catches up with the hardware, the future will arrive.

Thomas Messerschmidt

-  

Need something prototyped, built or coded? I’ve been building prototypes for companies for 15 years. I am now incorporating generative AI into products.

Contact me directly or through LinkedIn:   

https://www.linkedin.com/in/ai-robotics/

On Oct 29, 2025, at 6:52 AM, Alan Timm <gest...@gmail.com> wrote:

Heh I think I'm more ok with NEO's face than whatever Unitree put on their H2.  (shudder)

The WSJ did a piece on NEO yesterday that gives alot of visibility into how they're planning on making this all work.

https://www.youtube.com/watch?v=f3c4mQty_so

From NotebookLM:

"A Wall Street Journal video about the Neo humanoid home robot by 1X, available for pre-order at $20,000 with a 2026 delivery date. The robot features a lightweight design inspired by biology, using strong, light motors and tendon-like structures to move smoothly and quietly, though its strength and dexterity are comparable to a human's. Currently, Neo's actions are tele-operated by a human "pilot," which generates training data for the AI neural network that is intended to make future models autonomous. This learning process requires early adopters to accept a "social contract" involving privacy trade-offs, as 1X requires data from the home environment to improve the product, despite implementing safety and privacy measures like blurring people and enforcing no-go zones. Ultimately, the current version is likened to raising a toddler, signaling the beginning of physical AI in the home rather than delivering an immediately perfect, super-useful device."

On Tuesday, October 28, 2025 at 9:39:59 PM UTC-7 Thomas Messerschmidt wrote:

I just can’t imagine awakening to that face inside my house every morning. 

<image0.jpeg>

Thomas Messerschmidt

-  

Need something prototyped, built or coded? I’ve been building prototypes for companies for 15 years. I am now incorporating generative AI into products.

Contact me directly or through LinkedIn:   

https://www.linkedin.com/in/ai-robotics/

--
You received this message because you are subscribed to the Google Groups "RSSC-List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rssc-list+...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/rssc-list/682eee84-7245-4750-9ed0-2e2dbf41ab82n%40googlegroups.com.

<image0.jpeg>

[Subhobroto Sinha]

Yes, past failed promises dampen enthusiasm, but I argue that if these robots could only do 3 things, they would be tremendously successful:

1. Wet mopping: Vacuum robots are here but they aren't great in removing stains and bacteria. I hypothesize this is a big deal and a market opportunity.

2. Elderly fall detection and paging: Most of us in the club aren't yet in that age bracket to experience this first hand and don't realize how expensive and risky of a situation this is. Currently the only practical solution right now is to have security cameras blanket the rooms where the elderly move around in and on detecting a fall, page and alert for help. We are talking minutes here between life and death.

3. Perimeter security and alerting: Many burglary cases in Southern California are being won by defendants claiming they didn't understand or read warning signs like trespassing or private property. It's said that ignorance of the law isn't a defense, but due to how regulations are phrased in our state, this loophole exists. Attempts have been made to place speakers that shout warnings (defense claims they are deaf), robot dogs and boxes (which were missed because they were not at eye level). It would be interesting to see if defense could counter a human-height robot that speaks 20 languages and might even perform ASL.

It's well known (atleast to burglars) that houses in the U.S. have virtually no security. Garage doors can be opened by a hangar, in the name of safety double-locking doors are illegal so you break the glass and trivially open the lock from the inside, doors can be kicked in, your house breaker must have an easily accessible breaker for firemen (and burglars) to turn off. The only reason why burglary isn't a daily occurrence yet is the threat of violence - that the homeowner is ready to defend themselves.

Unfortunately, in Southern California, the homeowner assumes tremendous liability when defending against a burglar unless extremely specific clauses are met.

My other hope is that one of these platforms embrace open-source and let people run their software on top of the platform so the users can experiment and contribute to the community. We have some excellent buliders

[Alan Timm]

So that WSJ video made things alot more clear.

Here's how I think this first batch is going down:

• They'll only be made available to homes and offices local to their main office unless you're a VIP and they're using it as a press opportunity
• Their prototype malfunctioned during a fluff wsj piece and they had to cart it out on a mini crane, so I don't think shipping these around is an option for this first round
• The cost of these is heavily subsidized, but they're still charging for them which is surprising because:
• The whole purpose of this exercise it to collect real world data to add capabilities to their world model, which can't do anything right now
• Which means that every time you find something novel for the robot to do, someone's going to teleoperate it to collect the training data

I think it's telling that there wasn't a single activity that wasn't teleoperated in the wsj promo piece.  They're betting that progress will happen quickly once they start collecting real world data outside of their lab.

It's an interesting bet.  It hasn't stopped me from wanting one ever since I saw the announcement.  :-)

[Chris Albertson]

On Oct 29, 2025, at 11:16 AM, Subhobroto Sinha <subho...@gmail.com> wrote:

Yes, past failed promises dampen enthusiasm, but I argue that if these robots could only do 3 things, they would be tremendously successful:

1. Wet mopping: Vacuum robots are here but they aren't great in removing stains and bacteria. I hypothesize this is a big deal and a market opportunity.

Yes, this does seem to be something people want. Look here: https://www.tomsguide.com/best-picks/best-robot-mops

2. Elderly fall detection and paging: Most of us in the club aren't yet in that age bracket to experience this first hand and don't realize how expensive and risky of a situation this is. Currently the only practical solution right now is to have security cameras blanket the rooms where the elderly move around in and on detecting a fall, page and alert for help. We are talking minutes here between life and death.

My Apple watch does fall detection and will notify whoever I put on the list.   I announce to will send the notices in 30 seconds unless I cancel it.      There are other devices that do this too.  You can buy one that goes on a neck lanyard: https://shop.ncoa.org/best-medical-alert-systems-nb-3

In addition, there are $30 sensors that will detect a fall using mmWave Radar, but you have to build these into a larger system.   Here is one from TI: https://www.ti.com/video/6247990238001

3. Perimeter security and alerting: Many burglary cases in Southern California are being won by defendants claiming they didn't understand or read warning signs like trespassing or private property. It's said that ignorance of the law isn't a defense, but due to how regulations are phrased in our state, this loophole exists. Attempts have been made to place speakers that shout warnings (defense claims they are deaf), robot dogs and boxes (which were missed because they were not at eye level). It would be interesting to see if defense could counter a human-height robot that speaks 20 languages and might even perform ASL.

I think if you had a $20,000 robot outside.  It would just be another target.     It’s not good to leave something outdoors  that costs as much as a car but can be tossed into a pickup truck.

[Subhobroto Sinha]

Chris,

much thanks for those links! I am aware of the mmWave Radar (as I use it for intruder detection) but missed the fall detect capability! Thank You for reminding me of that. Last time I checked, getting the calibration done was tricky business (which is why I gave up on it and then forgot about it - if any of you have gone down that rabbit hole, I am all ears)!

The robot being a target is accurate - I had assumed its weight would be a barrier (it's heavy), but nothing that 4 well built gentlemen won't be able to handle.

[Thomas Messerschmidt]

Didn’t they say that the robot opening the door was autonomous?

As for collecting training data, as I remember, Tesla collected over 6 billion miles of training data (over 100 years of continuous driving data) for their cars. How would it be possible for  this company to collect enough information for a household robot that might compare to that? 300 robots working in continuous 8 hour shifts for a year? Although , home service Robots (robot servants?) would probably be idle a good part of the day in most households. Not to mention the time needed to sit idly while charging.) Me thinks that the training will be the hardest part of having a useful home robot servant. 

Thomas Messerschmidt

-  

Need something prototyped, built or coded? I’ve been building prototypes for companies for 15 years. I am now incorporating generative AI into products.

Contact me directly or through LinkedIn:   

https://www.linkedin.com/in/ai-robotics/

On Oct 29, 2025, at 8:17 PM, Alan Timm <gest...@gmail.com> wrote:

So that WSJ video made things alot more clear.

--

You received this message because you are subscribed to the Google Groups "RSSC-List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rssc-list+...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/rssc-list/0ab4d604-77de-41cd-a23b-3a087397b484n%40googlegroups.com.

[Chris Albertson]

On Oct 29, 2025, at 9:01 PM, Subhobroto Sinha <subho...@gmail.com> wrote:

Chris,

much thanks for those links! I am aware of the mmWave Radar (as I use it for intruder detection) but missed the fall detect capability! Thank You for reminding me of that. Last time I checked, getting the calibration done was tricky business (which is why I gave up on it and then forgot about it - if any of you have gone down that rabbit hole, I am all ears)!

I’m using mmWave for human presence detection, mostly to turn the lights on if someone is in the room.   Those IR motion detectors have a problem where they will turn the lights off if a person does not move.  But mmWave can detect breathing and heartbeats, and some of them can detect falls.

No calibration is needed if you buy the complete retail unit.     The problem I have with them is they need about 1/2 Watt of power and can’t be battery-powered.   But a coin cell will power an IR detector for a year.

Here is one that seems good and affordable: https://www.seeedstudio.com/MR60FDA2-60GHz-mmWave-Sensor-Fall-Detection-Module-p-5946.html

They say “no-code” integration in 1 or 2 minutes.    Yes, assuming you already have the infrastructure in place.    This unit uses MQTT over WiFi. 

Using a humanoid robot to move a camera around is not cost-effective.   Just buy a few cameras and an AI-based monitor.    

My plan for a domestic robot is that the house itself is the robot.  Rather than sensors and AI in a humanoid robot, we can place the sensors on the wall and the AI in a closet.  Then we need mobile devices to “do stuff."

The big problem is that our current AI is not good enough and will not become good enough by just incremental development.    Like a locomotive will never fly even if we make the wheels a lot better.   You need a different technology.  Trains do transportation well.  But airplanes fly not trains.   Basically, our AI is like a train.  It works, but we need breakthrough-level advancements.   Gradient descent training over tons of data will only get you about what we have.