From AI Slop to Robotics Slop

The success of the first robot housekeeper depends on users accepting robotics slop.

I’m one of those people who found the quote “I want AI to do my dishes and my laundry so that I can do art and writing” a bit silly.

Don’t get me wrong, when I was a kid, I always dreamed of having a robot that would do all the chores, like in The Jetsons. However, as an adult, I understand we don’t need an AI-driven robot for that. We already have home machines that handle those chores effectively.

That said, after watching the video of 1X Technologies’ humanoid robot “NEO” I felt like a kid again.

For a moment, I was fascinated by how far we got. Finally, a robot that can handle all those household chores, I thought.

But then I heard that most actions of the robot shown in the video were controlled by a human operator. I was disappointed at first, but after doing some research, I understood why a human operator on the loop is needed (at certain stages, not at every moment), where the home-robot field is heading, and why the term “robotics slop” will be a thing in the coming years (and I’m not sure that’s necessarily bad).

---

Hi! I’m Frank (aka The Pycoach)

AI & Python instructor. Over the past year, I’ve helped thousands on YouTube and Udemy master AI with beginner-friendly guides and in-depth tutorials

If you’re new here, welcome! Here’s what you might’ve missed:

• How to Go From AI Beginner to Pro in 2025

• ChatGPT Atlas Will Change Web Browsing. Here’s Where It Beats Chrome

• How to Use AI with Excel: From Generating Functions to Complete Reports

• How to quickly optimize your prompts with ChatGPT

NEO The “Autonomous“ Home Robot

NEO is a humanoid domestic robot developed by 1X Technologies for home use. NEO is described by 1X as “the world’s first consumer-ready humanoid robot.” The robot has a head, torso, two arms, and two legs. It’s about 5’6” (167 cm) tall and wrapped in a soft fabric suit that makes it look like an attentive home assistant rather than a cold machine.

NEO can move around on its own, manipulate objects with human-like dexterity, and interact with users through voice and AI-driven conversations. The robot’s brain runs 1X’s custom “Redwood” AI – a generalist vision-and-language model that enables NEO to understand spoken commands, recognize people and objects, and perform tasks such as retrieving items for you, opening doors, or navigating to a specific room.

Here are some specs:

Here are some things NEO can do:

• Walk and navigate around your home autonomously

• Perform chores like picking up items, opening doors, doing laundry, etc

• Talk and interact with people using built-in AI

• See and recognize objects/people with vision sensors

• Learn and adapt to your routines over time

All of that sounds amazing, but here’s the catch: while NEO will be able to autonomously do some chores out of the box when it ships in 2026, it’ll still need to learn a bunch of new chores.

How will it learn? A 1X teleoperator will teach NEO to perform the new task in your home.

Why NEO needs to be teleoperated

Whether you purchase NEO for $20,000 or subscribe for $499 per month, you should expect many chores to be controlled remotely by human operators.

Why? There are two main reasons.

First, home settings are extremely varied (furniture layout, objects, human habits, pets, etc) and present challenging conditions for robots. Initially, it’s necessary to have a real person controlling the robot’s actions to help the robot’s visual, manipulation, and contextual models learn from the real world (in this case, your home). People controlling the robots will be 1X Experts (1X employees physically present in the USA).

Second, the human-in-the-loop approach helps 1X collect valuable training data. NEO has an “Expert Mode“ that helps it learn new skills. For any chore it doesn’t know, you can schedule a 1X Expert to guide it, helping NEO learn while getting the job done.

The training data 1X collects is crucial for the success of the product.

The “home environment” is unstructured and variable. Large volumes of data are needed to understand this environment. Consider self-driving cars: Tesla collects video from its fleet to train its neural networks. 1X Technologies must also collect data—except in this case, it needs data from inside homes to teach its robots multiple chores across different house environments.

In the interview with the WSJ, 1X CEO said, “If you buy this product, it’s because you’re OK with that social contract. If we don’t have your data, we can’t make the product better.”

Will people be okay with letting strangers into their homes to teach a robot to do their chores?

To answer this question, we need a sense of who the early adopters will be

The first NEO units are scheduled to ship to customers in 2026, primarily in the United States. The $20k buyers will receive priority delivery throughout 2026. I think people who can afford this know they could pay for cleaning services for a long time with $20k. They’re not buying it for household chores but for the luxury of having a robot in their homes

In the interview with the WSJ, 1X CEO used the word “robotics slop” to compare the current state of humanoid robots to the early days of AI tools that produced flawed outputs. His view is that society might accept a certain level of sloppiness in a robot, as long as it’s good enough to be useful, and that performance will improve over time.

I don’t think robotic slop in our homes is a good idea.

Should we be ok with robotics slop?

The term AI slop has become common lately to describe low-quality content created with AI tools. Annoying as it is, its harms are mostly informational: spammy content, misinformation, and trust erosion.

Physical-world errors (or “robotics slop”) are far more alarming.

These errors can range from minor inconveniences, like breaking dishes, to life-threatening scenarios, like fetching the wrong medication. When automation goes wrong in the real world, people get hurt and property gets damaged.

Consider these examples of automation gone wrong in the real world:

• GM’s Cruise robotaxi dragged a pedestrian 20 feet following tech failures

• A delivery robot nearly crashed into a man

• An industrial robot crushed a worker after apparently mistaking him for a box

These are just some of the widely reported human-robot incidents so far.

Once consumer-ready humanoid robots enter people’s homes, the number of incidents will inevitably increase.

Naturally, NEO is built to operate safely around humans. It features passive safety measures (soft materials, rounded edges, capped force on motors) and active safety systems (sensors to detect collisions, software limits on speed in tight spaces). Additionally, 1X has built multiple layers of safety systems that ensure NEO won’t perform dangerous actions like picking up something heavy, hot, or sharp.

I don’t doubt 1X’s safety systems, but unfortunately, tech products have shown us that there’s never zero risk of failure, especially when human interaction is involved.

Over the past few years, we’ve been getting used to AI slop, but I don’t think we’re ready for robotics slop—at least, not in our houses! But that’s also bad news for the robotics industry. If people don’t adopt this new technology, there will never be enough real-world data for these products to improve.

Early adopters are crucial for the development of this technology. As Joanna Stern from the WSJ said in the video review, “the next few years won’t be about owning a super useful robot, but about raising one.”

The success of this product will depend on early adopters being willing to share video data of their homes and accepting robotics slop as part of the process.

Let me know what you think about NEO in the comments 👇