There are labs training AI robots for these very tasks right now. There are robots that fold laundry and spread peanut butter, right now.
We're maximum 1-2 years away from a general all-around robot the way chatgpt is an all around digital assistant.
Ten years later, I don't think the state of the art in general purpose robotics is that much further along. Even top tier research robots would struggle to climb the stairs or navigate around my kids' messy bedrooms, much less complete the wide array of tasks that my amazing housekeeper cheerfully and proactively handles for a bit over minimum wage.
I'm aware of the 1x Neo, but that has a lot of unknowns— the videos are full of cuts and have significant evidence of teleoperation and no indication of things like lift strength or battery life. What they're purporting to show is simply not aligned with what the state of the art is in perception and manipulation/grasping. They're a long ways from having even the hardware solved, much less the software.
Have you seen the recent side-flipping robot? https://www.youtube.com/watch?v=29xLWhqME2Q
Combine that with Project Astra https://youtu.be/hIIlJt8JERI?t=39
and you are kind of close.
But at the low end, AI is never going to be the janitor or your garbage guy, or your barista (ordering kiosks aside) or your instacart dude. AI won't unclog your toilet or service your car or make your basement not smell funny. It won't tend the garden or cut the grass or install a new air conditioner or re-shingle a roof.
I think the prospect of a world where everyone has their basic needs covered and has to choose to work is scary to people who are used to looking down on certain professions and assuming that no one would do <thing> if they didn't absolutely have to.