[lp4vn]

I think that at some point AI will enter the domain of repetitive physical jobs, like for example an experienced butcher will train an AI arm that eventually will learn to slice meat correctly since that's an activity that normally works with a reasonably standardized product(the animal carcass).

Keeping that in mind, honestly for me it's hard not to be very skeptical that it would make any economic sense in the next many decades to have an expensive humanoid robot performing tasks with a very low aggregated value like moving boxes one by one. Just because it's a robot performing an economic activity it doesn't necessarily mean it will be productive or competitive.

[darkwizard42]

I don't know about your example, animal carcasses vary a bunch and it would have to be great at not leaving any meat wasted. Maybe like hammering a nail or properly turning screws and checking tightness?

I think it will do well at performing economic activities that don't have real hard time constraints, but just require labor. Ex. moving boxes around from one storehouse to another...leave the last bit of sorting and packing to humans, but now you don't need someone doing the boring work.

[lp4vn]

>I don't know about your example, animal carcasses vary a bunch and it would have to be great at not leaving any meat wasted.

Cutting meat in practice is an activity that's more complex than it looks and for sure carcasses vary a lot, but even taking in account all their variation they still follow a pattern. At some point I'm sure we'll have robotic AI arms that will be trainable just like chatGPT is, so after being trained by a profession with thousands and thousands of hours, the hypotetical intelligent robotic arm eventually learns to do the job just fine 99.99% of the time and that's enough.