I’ve been watching the humanoid sector explode into the public sphere with promises of making everyone’s lives easier. There are viral videos of robots unpacking groceries, stocking a fridge, and folding laundry. There are bold predictions of robots revolutionizing labor markets and replacing humans on factory production lines. The narrative is exciting, and the vision is intoxicating. But that’s all it is – a vision. Or rather, a fantasy. At least for now.
In reality, the results are dismal, and the robot idea is imploding – or crashing, really. Recently, Russia’s much-touted AIdol humanoid robot took to the stage to the Rocky theme, only to topple over seconds later, scattering parts across the floor. The tumble highlighted not just a technical or mechanical glitch but a broader chasm between robotic promise and performance.
Another viral video shows a robot completely unable to perform the simple task of cooking an egg, instead dropping the hot pan on the floor, making a mess, and ultimately losing its balance, slipping on the spilled eggs, before crashing to the floor.
AIdol isn’t the only robot to face glitching and failure. A Chinese prototype at IROS 2025 glitched mid-dance, and early Optimus trials showed lags in real-time decision-making.
Battery life remains another issue, as recharging downtime makes robots less useful than thought. In addition, robots do well on flat surfaces but less so on uneven areas, requiring vast datasets that generative AI still struggles with.
What we are seeing is a problem of an industry racing ahead of technology. Let’s face it, the engineering, the software and hardware just isn’t quite there yet.
The appeal of humanoid robots is obvious. They are meant to mimic our form and function, engaging in tasks that we humans find repetitive, tedious, or boring. Tesla’s Optimus, unveiled in 2021, was reimagined in 2025 as a factory assistant capable of walking, sorting objects, and integrating with Tesla’s AI vision systems for autonomous navigation. Elon Musk’s idea is to create an “army” of robots working in unison on production lines – a possible major boost for companies in need of better automation.
Boston Dynamics’ Atlas robot does all sorts of acrobatic moves that could be useful for militaries in the field and on difficult terrain. Figure Ai and China’s Unitree are working on robots that could be useful in assisting with elder care and warehouse stocking. The idea is that robots, powered by AI, will replace humans in jobs full of drudgery and repetitiveness.
The World Economic Forum envisions humanoids as everyday fixtures within a decade. While this might displace 85 million jobs worldwide, new jobs will be created as robot oversight becomes necessary.
While possibly useful in clean, “square-based” factory settings, AI robots are not ready for chaotic consumer environments such as messy homes and are still considered unsafe, according to a recent study. For consumers to adopt the idea of a robot in the house, affordability, reliability, and safety remain major factors that current prototypes lack. Industrial applications are here, and within a year or two, we will see thousands of robots deployed in thousands of factories worldwide. Consumer use will take longer, and perhaps in five years we will see the first rollout of useful humanoid robots capable of carrying out safe tasks in a chaotic home environment full of kids, pets, laundry and toys, and other obstacles on the floor, and general clutter.
As we have seen with other technologies in the last 40 years, progress is undeniable. We should remain cautious in the near future and not allow the overhyping of humanoid robots to capture our imagination beyond reason. But there is no question that we are racing – or stumbling – toward a new age that will see humanoid robots everywhere. I don’t know about you, but I’m excited about it.