Home Humanoid Robots Still Face Cost, Trust, and Dexterity Hurdles

Humanoid robots are being presented as machines that have moved beyond the old image of unstable prototypes taking a few tentative steps before falling over. Cristina Criddle points to a robot named Lightning that finished the Beijing half marathon in 50 minutes and 26 seconds, a result she describes as breaking the human world record. The broader race now includes US companies such as Tesla and Figure, alongside Chinese competitors such as Unitree.

The leap in capability has made the consumer question harder to dismiss. 1X is not just demonstrating a robot; it is offering a home model. Its NEO Home Robot product page showed a standard subscription price of $499 a month, an early-access ownership option of $20,000, a $200 deposit due that day, and the promise to “Automate Your Chores.” Criddle says 1X hopes to ship models to US consumers later this year.

Bernt Børnich says the company’s early demand has already exceeded near-term supply: “We sold more than what we’re going to produce next year of robots in the first five days.” His claim is not that a humanoid will be priced like a small appliance. He explicitly rejects the coffee-machine comparison, calling the robot “a more complicated machine.” But he argues 1X can make it affordable relative to a larger household purchase — “pretty comparable to like a relatively affordable second car” — and says he expects it to add more value to a person’s life than that second car.

The financial forecasts around the category are similarly expansive. Criddle cites McKinsey data saying venture capital investment in robotics has tripled over two years to more than $40bn annually. Barclays Research estimates the humanoid market could reach $200bn by 2035. Bank of America projects humanoid shipments rising from 90,000 in 2026 to more than 1.2mn by 2030.

The consumer applications described are not dramatic. They are chores. Lindon Gao names folding clothes, loading dishwashers and cleaning up a room as the kinds of tasks he expects home robots to handle. When Criddle characterizes these as “all the boring jobs,” Gao agrees: “All the boring jobs that people don’t want to do.”

Criddle notes that economists have compared the current moment in AI and robotics to the arrival of the washing machine: a technology that could free people from domestic work. The robot as consumer product is being sold against the time and effort embedded in the household load.

Gao’s own view is that adoption will not be uniform at the start. He says some people, including himself, would be ready now, while others would not. The threshold changes, in his account, when cost falls and value and performance rise. At that point, he says, bringing a robot into the home would become “a no-brainer.”

The demonstrations still carry a gap between promise and polish. In one exchange, Criddle asks a robot to put its hands on its hips. A synthetic voice replies, “Hands on the hips coming right up,” then says, “There you go, looking sharp now, huh? What do you think?” Criddle’s verdict is only “Almost,” before asking for a wave. The robot can respond verbally and attempt the gesture, but the execution is not yet seamless.

The strongest skepticism comes from what the machines still cannot do well. Ken Goldberg distinguishes between locomotion and manipulation. Robots can walk, run and even do backflips, he says; “we now know that.” The missing capability is dexterity: “they’re not very good at manipulation, at dexterity.”

For a home robot, that limitation is central. Walking through a house is not the same as reliably doing useful household work. Goldberg says dexterity is “gonna be necessary if you want a robot to do something useful in your house.” Criddle also notes that, despite improvements, the machines still cost tens of thousands of dollars and remain “far from perfect.”

The source illustrates that imperfection with a viral X post from Mario Nawfal showing a humanoid robot stumbling on a staircase. The visible caption jokes that the robot is “channeling its inner Michael Jackson,” but that once it stumbles it looks like “someone trying to leave the pub at 2:00 AM.” The clip underscored how conspicuous failures can be to consumers.

That gap between public feats and routine usefulness is the practical test. A robot that runs a race or chases invasive wild boars out of residential neighborhoods, as Criddle says a Unitree humanoid did in a viral Warsaw video, is not automatically a robot that can handle dishes, laundry and clutter around fragile objects.

The case for humanoids depends not only on capability, but on whether the human form is necessary. Criddle raises the question directly: is a human-style robot essential, or could a different design do the job just as well? A promotional video from Syncere showed a robotic arm resembling a green desk lamp folding clothes on a bed — a reminder that household labor may not require a machine with a head, torso and legs.

Bernt Børnich argues the humanoid form is not a gimmick but a general-purpose strategy. In his view, the human body is “the most general machine we know.” He describes the ambition as a machine that can do “any labor across society,” with consumer homes as an important starting point because “that is where the intelligence is.”

Ken Goldberg is unconvinced that this translates into near-term practical adoption. He expects some people will buy humanoids as conversation pieces or novelties. But he doubts that useful, practical and cost-effective humanoids in the home are close: “Some people will get them. But will they be doing something really useful and practical, and will it be cost-effective? I worry that’s not gonna happen in the near future.”

The investment thesis and the engineering caution therefore sit side by side. The forecasts imagine a large and fast-scaling market. The companies are experimenting with consumer pricing and subscriptions. The use cases are obvious because domestic work is repetitive, disliked and valuable to avoid. But the hard part is making humanoids cheap enough, trusted enough and dexterous enough to prove they are more than a gimmick inside the home.