Humanoid robots are coming. Finally?

That is The Stepback, a weekly publication breaking down one important story from the tech world. For extra on robotics and AI, comply with Robert Hart. The Stepback arrives in our subscribers’ inboxes at 8AM ET. Choose in for The Stepback right here.

We’re taking a winter break! The Stepback will likely be again on January eleventh, 2026. In the meantime, you compensate for previous points right here.

I’ve a smooth spot for robotic fail movies. I watch them on a loop, chuckling to myself, as a sort of remedy. Possibly I’m a sadist, possibly I have to get out extra — you possibly can choose me later — however they get me each time. So naturally, I’ve been glued to a clip of Tesla’s Optimus robotic falling like a felled tree on the firm’s Autonomy Visualized occasion in Miami that’s been doing the rounds on social media this week.

Based on the footage, Elon Musk’s vaunted humanoid robotic was handing out water bottles from behind a desk earlier than knocking a bunch over, flailing its arms upward, and collapsing backwards like a puppet whose strings had been severed. Watch carefully and also you’ll see two issues: a small plume of water as its arm crushes a bottle mid-fall (it made me snort), and a movement uncannily much like somebody eradicating a VR headset.

It wouldn’t be the primary time Tesla has faked the autonomous half of its autonomous robotic, which Musk says is central to the corporate’s future. An early demonstration was only a dancer in a skintight bodysuit to point out what the Tesla Bot, now Optimus, might be. Later demonstrations have been revealed to be (fairly clearly, by the sound of it) people in disguise, working the robots remotely with one thing like a VR headset, which we all know Tesla makes use of in improvement.

People have been obsessive about robots for hundreds of years; from historical tales of stone golems and synthetic automatons to fashionable science fiction, driverless automobiles, and Roombas, the thought of animating the inanimate with one thing resembling life has fascinated us. A lot of the present hype for humanoids may be traced on to Musk, so it’s affordable to be greater than a bit bit skeptical when he and others promise they’ll revolutionize the world. Musk, who has vowed to construct a “robotic military” of 1 million humanoids, has a deserved repute as an outlandish and unreliable forecaster, and robotics has ridden various hype waves in its historical past. Up to now, expertise has at all times lagged the keenness of these wanting to deliver AI into the true world, however at present, we’re being informed that tech is lastly able to ship.

So what does able to ship truly appear like in 2025?

There’s positively a gold rush of types for humanoid robots proper now. Each main tech firm has them on their roadmap in some capability, and the likes of Nvidia, Meta, SoftBank, Google, Amazon, Microsoft, Intel, and Tesla (duh) are all throwing severe weight — and money — behind them as the following large frontier in tech. They usually’re not alone: A rising constellation of challengers need in on the motion, together with Apptronik, Boston Dynamics, Determine AI, and 1X.

China needs in, too. Beijing has determined embodied AI — which additionally contains non-humanoids like drones, quadrupeds, and different autonomous machines — is the important thing to future financial progress. In its trademark vogue, it has moved to place itself as a world chief in robotics by means of huge funding, authorities directives, and state subsidies. Everybody from tech giants Ant Group and Baidu to startups like Unitree and AgiBot is piling in.

In the event you go by the demos China and everybody else is pumping out, you’d be forgiven for pondering the humanoid robotic future has already arrived. This summer time, robots competed in dance, fight, observe, and area occasions on the first World Humanoid Robotic Video games in China. An identical occasion was held in Greece — the International Humanoid Olympiad — on the birthplace of the Olympic Video games. Humanoid fights are extra well-liked than I might have imagined and are apparently going down completely in every single place, from organized contests and underground battle golf equipment to sparring with CEOs.

Firms are additionally eager to get bots out of factories and into the house. These sorts of human-centric areas are why advocates say humanoids are value pursuing as a substitute of different robotic physique varieties that could possibly be simpler to supply. Determine stated its new Determine 03 robotic can tackle home chores, releasing a video of the bot doing dishes and folding garments. 1X debuted Neo, which it claims is protected and “the world’s first consumer-ready humanoid robotic.” There are additionally movies of Neo shakily finishing fundamental family duties. I’m unsure I can afford it, however if you happen to’re , 1X is promoting items for $20,000, with supply beginning within the US subsequent yr.

However precise use circumstances stay slim, and spectacular demos aren’t the identical as working merchandise. For all of the progress — and there have been spectacular advances — demos are staged, scripted, and even teleoperated. For instance, whereas it’s technically true to say that Ant Group’s R1 was cooking for audiences at a commerce present this yr, it moved at a tempo so glacial it might’ve thrilled Meryl Streep’s Miranda Priestly. Equally, the thought of getting 1X’s Neo sounds a lot much less interesting when you understand it’s essential to additionally decide to having random people telecommute into your private home to manage it remotely. And likewise, robotic sport occasions are so enjoyable to look at as a result of the contestants are wobbly, unsure, and unpredictable, not as a result of they’re proficient athletes.

All of this raises the apparent query: If the tech clearly doesn’t match the hype, why are so many individuals getting so enthusiastic about humanoids? Properly, for the primary time in a very long time, it’s not that loopy to suppose the tech could be beginning to catch up.

Traditionally, it has been exceptionally tough to get robots to carry out even mundane duties that people do with out pondering, like strolling or selecting up a glass of water. {Hardware} — however largely software program — was a limiting issue, and machines have been confined to extremely structured environments and really slim duties. Advances in AI are altering that and turbocharging robotic improvement.

The extraordinary breakthroughs in AI picture and textual content era from corporations like OpenAI and Google are largely all the way down to giant language fashions, or LLMs, that energy chatbots. These work through the use of advanced networks to acknowledge and reproduce patterns discovered within the huge quantities of information they’re skilled on. Greater AI fashions are skilled on virtually all the things that may be scraped from the web, and the ensuing methods are able to appearing in a generalized means, fairly than following inflexible, hand-coded guidelines. To perform in the true world, robots want this sort of generalization, and roboticists try to make use of the identical strategy to offer machines the versatile understanding of the bodily world they’d have to survive in it.

The problem right here is knowledge. The sort of materials used to coach LLMs was abundantly accessible on-line, however the knowledge wanted to coach robots is just not practically as straightforward to return by. Robots want detailed examples of issues transferring in the true world, which doesn’t actually exist at scale. Firms are working exhausting to alter that, and there’s a big effort to create the sort of knowledge wanted to coach robotic fashions at scale. On the floor, this appears ridiculous — Tesla has employees don cameras and sensors to show Optimus to behave like a human — nevertheless it’s working. It’s additionally why corporations like 1X are deploying less-than-autonomous bots into peoples’ properties, as doing so offers them the chance to assemble the information wanted to make autonomous bots by conducting advanced duties remotely, corresponding to loading the dishwasher or cleansing a tough object.

Humanoids have gotten extra reasonably priced, too — particularly in China — as {hardware} prices sink and economies of scale kick in. Whereas costs range wildly — entry-level items may be as little as $1,400 for Chinese language mannequin Bumi to round $13,500–$20,000 for fashions from Unitree or 1X, whereas industrial fashions can price greater than homes — they’re approaching costs reasonably priced by some customers, that means extra bots are being deployed in the true world. This creates a suggestions loop, as corporations get extra knowledge to work with, they construct higher fashions and higher robots that extra folks would need, presuming they’re prepared to welcome them into their properties.

However even with all this progress, it could nonetheless be hype. In November, China’s main financial planning company warned a humanoid robotic bubble could possibly be brewing, contrasting the sheer variety of corporations and scale of funding with the absence of viable use circumstances. The actual fact is, they’re not likely autonomous but, so, apart from hobbyists and researchers, who of their proper thoughts would purchase one? If I wished to get somebody in to scrub my dwelling, I might rent a cleaner with out the hefty upfront price and save my cash till somebody produces a robotic that may truly do the duty I purchased it for.

Till corporations cease hiding behind glitzy promo movies and teleoperating purportedly autonomous robots, it’s going to be exhausting to inform the place we’re. Possibly the robots are lastly coming, or possibly I’ve many extra fail movies coming my means. Time will inform, I suppose, however I’ll get the popcorn prepared.