I'm excited for the time in the near future when a useful robot arm costs roughly as much as a good laptop i.e. ~1k USD.
At that point, it'll become feasible for hobbyists to get one each and start playing around. Which will lead to an explosion in the variety of things tried, and will positively feed back into creating a healthy community, tons of software which makes it easier for the next set of people to get involved, etc. Much like the PC revolution in the 80s/90s.
At the moment, they're kinda like overpriced toys (for a home user / enthusiast point of view) without a killer app.
I feel like people have been saying this is the "near future" for 20 years now.
Odriverobotics already has brought down the barrier to BLDC servomotor control.
Or this thing: https://ieeexplore.ieee.org/document/8613852
2) One of the biggest financial roadblocks I've run across in robotics isn't about the torque available in affordable servos, it's about the ability to get force feedback. Servos/motors are useless for the large set of tasks requiring force feedback, without the additional complexity of building a current sensor and calibrating it. We need a cheap servo with integrated force feedback and an easy means of autocalibration (from a quick search, seems like this feature usually doubles or tripples the cost of an identical servo).
Force feedback is also super important and challenging.
We just need like, synthetic actin :) Disney is actually working on that [2]. yeah it looks super inefficient but so were vacuum tubes before transistors.
1] https://eeepitnl.tksc.jaxa.jp/mews/EN/19th/text/206.pdf&ved=...
2] https://studios.disneyresearch.com/wp-content/uploads/2019/0...
Good 3D printers can now be had for under 1k USD – and if you can tolerate some tinkering and modifications $300-$500 can get you decent results as well.
https://www.trossenrobotics.com/interbotix-turtlebot-2i-mobi...
The arms themselves appears to be in the 400-600 range