ranzware
In a video announcing a collaboration between 1X’s AI team and the Nvidia Gear Lab, the much-touted, still-in-development Neo Gamma humanoid robot gave Nvidia CEO Jensen Huang a studded leather jacket. Depicting the hardware company’s logo on the back in metal studs, the jacket was made by California-based clothing brand ERL—but this collaboration wants to venture beyond the wardrobe and into the home.
1X Technologies is in the business of humanoid robots for industry settings and the home, such as the Neo Gamma. Based partly in California, they’ve been around since at least 2014; originally founded in Norway as Halodi Robotics, they underwent a rebrand in 2023 and have more recently embarked on an endeavour to make friends with heavy hitters in big tech.
Vice President of AI at 1X Technologies Eric Jang said of the collaboration, “[We’re both] super determined to bring general purpose humanoid robots into the world. We think that by swapping notes on how to solve autonomy problems on Neo [Gamma] we can dramatically accelerate the timeline for bringing Neo into people’s homes.”
Questionable fashion choices aside, a recent blog post goes into more detail, sharing, “As a first step, the teams worked together to prepare an autonomy demo for Jensen Huang’s GTC 2025 Keynote, featuring Neo doing a dish loading task autonomously.” The collaboration saw team members take the Neo Gamma into a home environment for one week so it could get some heavily supervised, real-ish world experience of carrying out lightweight domestic tasks. Nice to know that it’s not just me that requires a watchful eye whenever I load the dishwasher—don’t ask me why I keep going through wooden chopping boards and cafetieres at a rate of knots.
The blog post gets into the technical details of the collaboration, explaining 1X’s AI team “created a dataset API for Nvidia to access data collected from 1X offices and employee homes.” Talk about taking work home with you—anyway, this API worked in conjunction with a software dev kit serving up “model predictions at a continuous 5Hz vision-action loop using an onboard Nvidia GPU in Neo’s head or an offboard GPU.”
With this tech infrastructure in place, the teams zeroed in on a specific series of motions—”autonomously [grasping] a cup, [handing] it over to the other hand, and [placing] it in a dishwasher”—to demonstrate how the Neo Gamma could one-day slot into the kitchen of any interested user with sufficiently deep pockets.
Though it’s fascinating to watch the Neo Gamma’s slow, methodical movements and sophisticated articulation, I remain sceptical. 1X tech’s official website shows the robot doing a range of complex domestic tasks like vacuuming, but it’s clear to me that at least this specific example is more of an artistic interpretation rather than an accurate representation of the robot’s current ability. The new promotional video, which shows the team bodily hauling the robot around and the Neo Gamma uncertainly depositing a fragile-looking cup into a dishwasher, seems ever so slightly more honest.
They—speaking very broadly, in an inelegant excuse for a Mean Girls reference—keep trying to make humanoid robots happen. If it’s not Elon Musk’s bartending robots, it’s uncanny ‘companions’. If it’s not strictly speaking human-shaped, then it’s an AI-powered dog that can allegedly learn new tricks and tiny cyber prisons for anime girls.
Sorry, I may have allowed the point to get away from me somewhat there…my point is that while automating domestic drear may appeal, history offers plenty of examples of how automation too often threatens to leave human workers out in the cold, from the historical origin of the word ‘luddite’ in the 1800s, to ChatGPT copying a whole lot of people’s homework in the here and now. Still, given that Neo Gamma still has many thousands of hours of training data to absorb before it can be given free reign over anyone’s chores, maybe I’m getting a little ahead of myself.
Source link
Kurt Indovina
Add comment