As someone with two small kids at home, I will most assuredly not be an early adopter for a robotic assistant at home. I love the idea of magical cleaning but that's something my kids should learn to do for themselves. I hate the idea of having a humanoid robot in my home all day long, and I am quite sure my wife would veto it on principle. I wonder if this is, like AR/VR headsets, out of touch with what the kinds of physical tech regular people want to buy?
I would also have massive safety concerns, having lived through all the BS around the safety of self-driving cars for the past decade. The main thing I wonder watching this video is how much was hard-coded and manually constrained in what we are watching? Is it a physical world version of the Cognition Labs / Devin demos that were floating around twitter and then quickly debunked last year?
My expectation is the majority (perhaps all?) of what we see isn't hard-coded and is a consequence of Helix driving the robot control and vision. However, what we don't know is how robust it is - we see a bunch of demonstrations, but the history of robotics is littered with examples of things like lighting changes or weight changes or object dynamics or strange actuator properties leading to brittleness. Nonetheless, I do think things have moved forward quite significantly from where they were years ago (when tons more stuff was hard-coded).
With regard to home robots, I'm not sure? There are a few tedious tasks I do these days - vacuuming, swiffering, picking up toys (so many toys!), picking up books, unloading the dishwasher, etc, that I'd rather have someone else do for me (and I feel like hiring a human cleaner is genuinely quite excessive here). I also plan to teach my kid to clean up as well, so I agree with you there.
Really appreciate the way you digest the landscape for us readers. Super valuable and nowadays the only AI newsletter that I always make time for. And the fiction is thoughtful, too!
As someone with two small kids at home, I will most assuredly not be an early adopter for a robotic assistant at home. I love the idea of magical cleaning but that's something my kids should learn to do for themselves. I hate the idea of having a humanoid robot in my home all day long, and I am quite sure my wife would veto it on principle. I wonder if this is, like AR/VR headsets, out of touch with what the kinds of physical tech regular people want to buy?
I would also have massive safety concerns, having lived through all the BS around the safety of self-driving cars for the past decade. The main thing I wonder watching this video is how much was hard-coded and manually constrained in what we are watching? Is it a physical world version of the Cognition Labs / Devin demos that were floating around twitter and then quickly debunked last year?
My expectation is the majority (perhaps all?) of what we see isn't hard-coded and is a consequence of Helix driving the robot control and vision. However, what we don't know is how robust it is - we see a bunch of demonstrations, but the history of robotics is littered with examples of things like lighting changes or weight changes or object dynamics or strange actuator properties leading to brittleness. Nonetheless, I do think things have moved forward quite significantly from where they were years ago (when tons more stuff was hard-coded).
With regard to home robots, I'm not sure? There are a few tedious tasks I do these days - vacuuming, swiffering, picking up toys (so many toys!), picking up books, unloading the dishwasher, etc, that I'd rather have someone else do for me (and I feel like hiring a human cleaner is genuinely quite excessive here). I also plan to teach my kid to clean up as well, so I agree with you there.
We'll see! Thanks for the comment
Really appreciate the way you digest the landscape for us readers. Super valuable and nowadays the only AI newsletter that I always make time for. And the fiction is thoughtful, too!