I remain unconvinced of the viability of robots, as much as I love scifi renditions.

Have you seen a modern car manufacturing plant? Many parts of the production pipeline are fully automated. Granted, most of these machines are not ambulatory but they're still considered robots. Or consider modern freight shipping: many ports rely on intelligent automation for container handling. The development path of 3d printing is also leaning more heavily into robotics, featuring freely-moving articulated arms controlled by cameras and sensors.

I'd say robots are entirely viable, and we don't need science fiction to validate them.

	▲	protocolture 6 days ago \| parent \| next [-]
		Yeah so the tradeoff appears to be size vs utility. The problem is that, at best, that means a lot of the world would have to be redesigned to cater to robots. Thats why they excel in auto plants. Space already isnt a concern, so you can make them huge. Huge robots are capable of tremendous strength dexterity and speed. But in an environment built for humans they suck. Redesigning a data center to be 100% robot operated will probably happen, but thats going to be an extraordinarily unfriendly place for a human to be. The amount of space you would lose getting a robot to be able to retrieve a crud rj45 connector, or a stuck sfp module, from any one of 200 racks, at multiple heights, would make the robot massive. So the entire concept of the data centre will have to be rebuilt from the ground up to make it robot friendly. The full tech stack too. Robot friendly connectors etc. Thats a huge capex outlay for something with dubious utility. Imagine ubiquitous robots on the street. Machines capable of tearing humans to shreds. The liability issues are huge on their own. If LLMs are the pinnacle of artificial intelligence, you would probably have a death a week in most cities. Space is worse because the robot has to be launch economical, or built up there. Whats he doing up there without humans. Back to accidents again.
	▲	anonzzzies 6 days ago \| parent \| prev [-]
		... and we are very early on in human-robot development still... We don't know yet if the current push will speed things up or leave it stagnant; I would say it's definitely not a stretch to assume it will speed up...